Merge branch 'master' into topic/robin/dnp3-merge-v3

Conflicts:
	scripts/base/init-default.bro

CHANGES

@@ -1,4 +1,70 @@
 
+2.1-1041 | 2013-08-09 15:32:22 -0700
+
+  * Update coverage baselines for canonical load order of scripts.
+    (Jon Siwek)
+
+2.1-1039 | 2013-08-09 15:30:15 -0700
+
+  * Fix mem leak in DHCP analyzer. (Jon Siwek)
+    
+  * Fix a unit test outdated by recent sumstats changes. (Jon Siwek)
+    
+2.1-1036 | 2013-08-05 17:29:11 -0400
+
+  * Fix the SSL infinite loop I just created. (Seth Hall)
+
+2.1-1035 | 2013-08-05 16:44:50 -0400
+
+  * Change to SSL log delay to cause the log to write even if delay times out. (Seth Hall)
+
+2.1-1034 | 2013-08-03 20:27:43 -0700
+
+  * A set of DHCP extensions. (Vlad Grigorescu)
+
+     - Leases are logged to dhcp.log as they are seen.
+     - scripts/policy/protocols/dhcp/known-devices-and-hostnames.bro
+     - Added DPD sig.
+
+2.1-1027 | 2013-08-03 01:57:37 -0400
+
+  * Fix a major memory issue in the SumStats framework.
+
+2.1-1026 | 2013-08-02 22:35:09 -0400
+
+  * Fix the SumStats top-k plugin and test. (Seth Hall)
+
+  * Rework of SumStats API to reduce high instantaneous memory
+    use on clusters. (Seth Hall)
+
+  * Large update for the SumStats framework.
+
+     - On-demand access to sumstats results through "return from"
+       functions named SumStats::request and Sumstats::request_key.
+       Both functions are tested in standalone and clustered modes.
+
+     - $name field has returned to SumStats which simplifies cluster
+       code and makes the on-demand access stuff possible.
+
+     - Clustered results can only be collected for 1 minute from their
+       time of creation now instead of time of last read.
+
+     - Thresholds use doubles instead of counts everywhere now.
+
+     - Calculation dependency resolution occurs at start up time now
+       instead of doing it at observation time which provide a minor
+       cpu performance improvement.  A new plugin registration mechanism
+       was created to support this change.
+
+     - AppStats now has a minimal doc string and is broken into hook-based
+       plugins.
+
+     - AppStats and traceroute detection added to local.bro (Seth Hall)
+
+2.1-1009 | 2013-08-02 17:19:08 -0700
+
+  * A number of exec module and raw input reader fixes. (Jon Siwek)
+    
 2.1-1007 | 2013-08-01 15:41:54 -0700
 
   * More function documentation. (Bernhard Amann)

VERSION

@@ -1 +1 @@
-2.1-1007
+2.1-1041

aux/binpac

@@ -1 +1 @@
-Subproject commit 314fa8f65fc240e960c23c3bba98623436a72b98
+Subproject commit 00674ed07d702252b00675b5060647d9e811cdd7

aux/bro-aux

@@ -1 +1 @@
-Subproject commit d9963983c0b4d426b24836f8d154d014d5aecbba
+Subproject commit 0e2d74e488195170e4648037e22b51e122dc7b0e

aux/broccoli

@@ -1 +1 @@
-Subproject commit d59c73b6e0966ad63bbc63a35741b5f68263e7b1
+Subproject commit 7ddfa3212d1fd0822588d4a96158f1a30c755afe

aux/broctl

@@ -1 +1 @@
-Subproject commit 52fd91261f41fa1528f7b964837a364d7991889e
+Subproject commit c9293bad3bf4d6fc3e1808a315e791140a632961

cmake

@@ -1 +1 @@
-Subproject commit 026639f8368e56742c0cb5d9fb390ea64e60ec50
+Subproject commit 370f7efc6e144f978e2309ad80b1a10c83feaaa9

doc/scripts/DocSourcesList.cmake

@@ -139,6 +139,9 @@ rest_target(${psd} base/protocols/conn/contents.bro)
 rest_target(${psd} base/protocols/conn/inactivity.bro)
 rest_target(${psd} base/protocols/conn/main.bro)
 rest_target(${psd} base/protocols/conn/polling.bro)
+rest_target(${psd} base/protocols/dhcp/consts.bro)
+rest_target(${psd} base/protocols/dhcp/main.bro)
+rest_target(${psd} base/protocols/dhcp/utils.bro)
 rest_target(${psd} base/protocols/dns/consts.bro)
 rest_target(${psd} base/protocols/dns/main.bro)
 rest_target(${psd} base/protocols/ftp/files.bro)
@@ -206,9 +209,16 @@ rest_target(${psd} policy/frameworks/software/vulnerable.bro)
 rest_target(${psd} policy/integration/barnyard2/main.bro)
 rest_target(${psd} policy/integration/barnyard2/types.bro)
 rest_target(${psd} policy/integration/collective-intel/main.bro)
-rest_target(${psd} policy/misc/app-metrics.bro)
+rest_target(${psd} policy/misc/app-stats/main.bro)
+rest_target(${psd} policy/misc/app-stats/plugins/facebook.bro)
+rest_target(${psd} policy/misc/app-stats/plugins/gmail.bro)
+rest_target(${psd} policy/misc/app-stats/plugins/google.bro)
+rest_target(${psd} policy/misc/app-stats/plugins/netflix.bro)
+rest_target(${psd} policy/misc/app-stats/plugins/pandora.bro)
+rest_target(${psd} policy/misc/app-stats/plugins/youtube.bro)
 rest_target(${psd} policy/misc/capture-loss.bro)
 rest_target(${psd} policy/misc/detect-traceroute/main.bro)
+rest_target(${psd} policy/misc/known-devices.bro)
 rest_target(${psd} policy/misc/load-balancing.bro)
 rest_target(${psd} policy/misc/loaded-scripts.bro)
 rest_target(${psd} policy/misc/profiling.bro)
@@ -218,6 +228,7 @@ rest_target(${psd} policy/misc/trim-trace-file.bro)
 rest_target(${psd} policy/protocols/conn/known-hosts.bro)
 rest_target(${psd} policy/protocols/conn/known-services.bro)
 rest_target(${psd} policy/protocols/conn/weirds.bro)
+rest_target(${psd} policy/protocols/dhcp/known-devices-and-hostnames.bro)
 rest_target(${psd} policy/protocols/dns/auth-addl.bro)
 rest_target(${psd} policy/protocols/dns/detect-external-names.bro)
 rest_target(${psd} policy/protocols/ftp/detect-bruteforcing.bro)

scripts/base/frameworks/sumstats/cluster.bro

@@ -10,10 +10,6 @@
 module SumStats;
 
 export {
-	## Allows a user to decide how large of result groups the workers should transmit
-	## values for cluster stats aggregation.
-	const cluster_send_in_groups_of = 50 &redef;
-
 	## The percent of the full threshold value that needs to be met on a single worker
 	## for that worker to send the value to its manager in order for it to request a
 	## global view for that value.  There is no requirement that the manager requests
@@ -27,45 +23,46 @@ export {
 	## performed.  In practice this should hopefully have a minimal effect.
 	const max_outstanding_global_views = 10 &redef;
 
-	## Intermediate updates can cause overload situations on very large clusters. This
-	## option may help reduce load and correct intermittent problems. The goal for this
-	## option is also meant to be temporary.
-	const enable_intermediate_updates = T &redef;
-
 	## Event sent by the manager in a cluster to initiate the collection of values for
 	## a sumstat.
-	global cluster_ss_request: event(uid: string, ssid: string);
+	global cluster_ss_request: event(uid: string, ss_name: string, cleanup: bool);
 
 	## Event sent by nodes that are collecting sumstats after receiving a request for
 	## the sumstat from the manager.
-	global cluster_ss_response: event(uid: string, ssid: string, data: ResultTable, done: bool);
+	#global cluster_ss_response: event(uid: string, ss_name: string, data: ResultTable, done: bool, cleanup: bool);
 
 	## This event is sent by the manager in a cluster to initiate the collection of
 	## a single key value from a sumstat.  It's typically used to get intermediate
 	## updates before the break interval triggers to speed detection of a value
 	## crossing a threshold.
-	global cluster_key_request: event(uid: string, ssid: string, key: Key);
+	global cluster_get_result: event(uid: string, ss_name: string, key: Key, cleanup: bool);
 
 	## This event is sent by nodes in response to a
-	## :bro:id:`SumStats::cluster_key_request` event.
+	## :bro:id:`SumStats::cluster_get_result` event.
-	global cluster_key_response: event(uid: string, ssid: string, key: Key, result: Result);
+	global cluster_send_result: event(uid: string, ss_name: string, key: Key, result: Result, cleanup: bool);
 
 	## This is sent by workers to indicate that they crossed the percent
 	## of the current threshold by the percentage defined globally in
 	## :bro:id:`SumStats::cluster_request_global_view_percent`
-	global cluster_key_intermediate_response: event(ssid: string, key: SumStats::Key);
+	global cluster_key_intermediate_response: event(ss_name: string, key: SumStats::Key);
 
 	## This event is scheduled internally on workers to send result chunks.
-	global send_data: event(uid: string, ssid: string, data: ResultTable);
+	global send_data: event(uid: string, ss_name: string, data: ResultTable, cleanup: bool);
+
+	global get_a_key: event(uid: string, ss_name: string, cleanup: bool &default=F);
+
+	global send_a_key: event(uid: string, ss_name: string, key: Key);
+	global send_no_key: event(uid: string, ss_name: string);
 
 	## This event is generated when a threshold is crossed.
-	global cluster_threshold_crossed: event(ssid: string, key: SumStats::Key, thold: Thresholding);
+	global cluster_threshold_crossed: event(ss_name: string, key: SumStats::Key, thold_index: count);
 }
 
 # Add events to the cluster framework to make this work.
-redef Cluster::manager2worker_events += /SumStats::cluster_(ss_request|key_request|threshold_crossed)/;
+redef Cluster::manager2worker_events += /SumStats::cluster_(ss_request|get_result|threshold_crossed)/;
-redef Cluster::manager2worker_events += /SumStats::thresholds_reset/;
+redef Cluster::manager2worker_events += /SumStats::(thresholds_reset|get_a_key)/;
-redef Cluster::worker2manager_events += /SumStats::cluster_(ss_response|key_response|key_intermediate_response)/;
+redef Cluster::worker2manager_events += /SumStats::cluster_(ss_response|send_result|key_intermediate_response)/;
+redef Cluster::worker2manager_events += /SumStats::(send_a_key|send_no_key)/;
 
 @if ( Cluster::local_node_type() != Cluster::MANAGER )
 # This variable is maintained to know what keys have recently sent as
@@ -74,12 +71,9 @@ redef Cluster::worker2manager_events += /SumStats::cluster_(ss_response|key_resp
 # an intermediate result has been received.
 global recent_global_view_keys: table[string, Key] of count &create_expire=1min &default=0;
 
-event bro_init() &priority=-100
+# Result tables indexed on a uid that are currently being sent to the
-	{
+# manager.
-	# The manager is the only host allowed to track these.
+global sending_results: table[string] of ResultTable = table() &create_expire=1min;
-	stats_store = table();
-	reducer_store = table();
-	}
 
 # This is done on all non-manager node types in the event that a sumstat is
 # being collected somewhere other than a worker.
@@ -87,95 +81,151 @@ function data_added(ss: SumStat, key: Key, result: Result)
 	{
 	# If an intermediate update for this value was sent recently, don't send
 	# it again.
-	if ( [ss$id, key] in recent_global_view_keys )
+	if ( [ss$name, key] in recent_global_view_keys )
 		return;
 
 	# If val is 5 and global view % is 0.1 (10%), pct_val will be 50.  If that
 	# crosses the full threshold then it's a candidate to send as an
 	# intermediate update.
-	if ( enable_intermediate_updates &&
+	if ( check_thresholds(ss, key, result, cluster_request_global_view_percent) )
-	     check_thresholds(ss, key, result, cluster_request_global_view_percent) )
 		{
 		# kick off intermediate update
-		event SumStats::cluster_key_intermediate_response(ss$id, key);
+		event SumStats::cluster_key_intermediate_response(ss$name, key);
-		++recent_global_view_keys[ss$id, key];
+		++recent_global_view_keys[ss$name, key];
 		}
 	}
 
-event SumStats::send_data(uid: string, ssid: string, data: ResultTable)
+#event SumStats::send_data(uid: string, ss_name: string, cleanup: bool)
-	{
+#	{
-	#print fmt("WORKER %s: sending data for uid %s...", Cluster::node, uid);
+#	#print fmt("WORKER %s: sending data for uid %s...", Cluster::node, uid);
+#
+#	local local_data: ResultTable = table();
+#	local incoming_data: ResultTable = cleanup ? data : copy(data);
+#
+#	local num_added = 0;
+#	for ( key in incoming_data )
+#		{
+#		local_data[key] = incoming_data[key];
+#		delete incoming_data[key];
+#
+#		# Only send cluster_send_in_groups_of at a time.  Queue another
+#		# event to send the next group.
+#		if ( cluster_send_in_groups_of == ++num_added )
+#			break;
+#		}
+#
+#	local done = F;
+#	# If data is empty, this sumstat is done.
+#	if ( |incoming_data| == 0 )
+#		done = T;
+#
+#	# Note: copy is needed to compensate serialization caching issue. This should be
+#	# changed to something else later. 
+#	event SumStats::cluster_ss_response(uid, ss_name, copy(local_data), done, cleanup);
+#	if ( ! done )
+#		schedule 0.01 sec { SumStats::send_data(uid, T) };
+#	}
 
-	local local_data: ResultTable = table();
+event SumStats::get_a_key(uid: string, ss_name: string, cleanup: bool)
-	local num_added = 0;
-	for ( key in data )
 	{
-		local_data[key] = data[key];
+	if ( uid in sending_results )
-		delete data[key];
+		{
-
+		if ( |sending_results[uid]| == 0 )
-		# Only send cluster_send_in_groups_of at a time.  Queue another
+			{
-		# event to send the next group.
+			event SumStats::send_no_key(uid, ss_name);
-		if ( cluster_send_in_groups_of == ++num_added )
+			}
+		else
+			{
+			for ( key in sending_results[uid] )
+				{
+				event SumStats::send_a_key(uid, ss_name, key);
+				# break to only send one.
 				break;
 				}
-
+			}
-	local done = F;
+		}
-	# If data is empty, this sumstat is done.
+	else if ( !cleanup && ss_name in result_store && |result_store[ss_name]| > 0 )
-	if ( |data| == 0 )
+		{
-		done = T;
+		if ( |result_store[ss_name]| == 0 )
-
+			{
-	# Note: copy is needed to compensate serialization caching issue. This should be
+			event SumStats::send_no_key(uid, ss_name);
-	# changed to something else later. 
+			}
-	event SumStats::cluster_ss_response(uid, ssid, copy(local_data), done);
+		else
-	if ( ! done )
+			{
-		schedule 0.01 sec { SumStats::send_data(uid, ssid, data) };
+			for ( key in result_store[ss_name] )
+				{
+				event SumStats::send_a_key(uid, ss_name, key);
+				# break to only send one.
+				break;
+				}
+			}
+		}
+	else
+		{
+		event SumStats::send_no_key(uid, ss_name);
+		}
 	}
 
-event SumStats::cluster_ss_request(uid: string, ssid: string)
+event SumStats::cluster_ss_request(uid: string, ss_name: string, cleanup: bool)
 	{
 	#print fmt("WORKER %s: received the cluster_ss_request event for %s.", Cluster::node, id);
 
-	# Initiate sending all of the data for the requested stats.
+	# Create a back store for the result
-	if ( ssid in result_store )
+	sending_results[uid] = (ss_name in result_store) ? result_store[ss_name] : table();
-		event SumStats::send_data(uid, ssid, result_store[ssid]);
-	else
-		event SumStats::send_data(uid, ssid, table());
 
 	# Lookup the actual sumstats and reset it, the reference to the data
-	# currently stored will be maintained internally by the send_data event.
+	# currently stored will be maintained internally from the 
-	if ( ssid in stats_store )
+	# sending_results table.
-		reset(stats_store[ssid]);
+	if ( cleanup && ss_name in stats_store )
+		reset(stats_store[ss_name]);
 	}
 
-event SumStats::cluster_key_request(uid: string, ssid: string, key: Key)
+event SumStats::cluster_get_result(uid: string, ss_name: string, key: Key, cleanup: bool)
 	{
-	if ( ssid in result_store && key in result_store[ssid] )
+	#print fmt("WORKER %s: received the cluster_get_result event for %s=%s.", Cluster::node, key2str(key), data);
-		{
-		#print fmt("WORKER %s: received the cluster_key_request event for %s=%s.", Cluster::node, key2str(key), data);
 
+	if ( cleanup ) # data will implicitly be in sending_results (i know this isn't great)
+		{
+		if ( uid in sending_results && key in sending_results[uid] )
+			{
 			# Note: copy is needed to compensate serialization caching issue. This should be
 			# changed to something else later. 
-		event SumStats::cluster_key_response(uid, ssid, key, copy(result_store[ssid][key]));
+			event SumStats::cluster_send_result(uid, ss_name, key, copy(sending_results[uid][key]), cleanup);
+			delete sending_results[uid][key];
 			}
 		else
 			{
 			# We need to send an empty response if we don't have the data so that the manager
 			# can know that it heard back from all of the workers.
-		event SumStats::cluster_key_response(uid, ssid, key, table());
+			event SumStats::cluster_send_result(uid, ss_name, key, table(), cleanup);
 			}
 		}
-
+	else 
-event SumStats::cluster_threshold_crossed(ssid: string, key: SumStats::Key, thold: Thresholding)
 		{
-	if ( ssid !in threshold_tracker )
+		if ( ss_name in result_store && key in result_store[ss_name] )
-		threshold_tracker[ssid] = table();
+			{
-
+			event SumStats::cluster_send_result(uid, ss_name, key, copy(result_store[ss_name][key]), cleanup);
-	threshold_tracker[ssid][key] = thold;
+			}
+		else
+			{
+			# We need to send an empty response if we don't have the data so that the manager
+			# can know that it heard back from all of the workers.
+			event SumStats::cluster_send_result(uid, ss_name, key, table(), cleanup);
+			}
+		}
 	}
 
-event SumStats::thresholds_reset(ssid: string)
+event SumStats::cluster_threshold_crossed(ss_name: string, key: SumStats::Key, thold_index: count)
 	{
-	threshold_tracker[ssid] = table();
+	if ( ss_name !in threshold_tracker )
+		threshold_tracker[ss_name] = table();
+
+	threshold_tracker[ss_name][key] = thold_index;
+	}
+
+event SumStats::thresholds_reset(ss_name: string)
+	{
+	delete threshold_tracker[ss_name];
 	}
 
 @endif
@@ -186,7 +236,7 @@ event SumStats::thresholds_reset(ssid: string)
 # This variable is maintained by manager nodes as they collect and aggregate
 # results.
 # Index on a uid.
-global stats_results: table[string] of ResultTable &read_expire=1min;
+global stats_keys: table[string] of set[Key] &create_expire=1min;
 
 # This variable is maintained by manager nodes to track how many "dones" they
 # collected per collection unique id.  Once the number of results for a uid
@@ -194,18 +244,18 @@ global stats_results: table[string] of ResultTable &read_expire=1min;
 # result is written out and deleted from here.
 # Indexed on a uid.
 # TODO: add an &expire_func in case not all results are received.
-global done_with: table[string] of count &read_expire=1min &default=0;
+global done_with: table[string] of count &create_expire=1min &default=0;
 
 # This variable is maintained by managers to track intermediate responses as
 # they are getting a global view for a certain key.
 # Indexed on a uid.
-global key_requests: table[string] of Result &read_expire=1min;
+global key_requests: table[string] of Result &create_expire=1min;
 
 # This variable is maintained by managers to prevent overwhelming communication due
 # to too many intermediate updates.  Each sumstat is tracked separately so that
 # one won't overwhelm and degrade other quieter sumstats.
 # Indexed on a sumstat id.
-global outstanding_global_views: table[string] of count &default=0;
+global outstanding_global_views: table[string] of count &create_expire=1min &default=0;
 
 const zero_time = double_to_time(0.0);
 # Managers handle logging.
@@ -213,15 +263,19 @@ event SumStats::finish_epoch(ss: SumStat)
 	{
 	if ( network_time() > zero_time )
 		{
-		#print fmt("%.6f MANAGER: breaking %s sumstat for %s sumstat", network_time(), ss$name, ss$id);
+		#print fmt("%.6f MANAGER: breaking %s sumstat", network_time(), ss$name);
 		local uid = unique_id("");
 
-		if ( uid in stats_results )
+		if ( uid in stats_keys )
-			delete stats_results[uid];
+			delete stats_keys[uid];
-		stats_results[uid] = table();
+		stats_keys[uid] = set();
 
 		# Request data from peers.
-		event SumStats::cluster_ss_request(uid, ss$id);
+		event SumStats::cluster_ss_request(uid, ss$name, T);
+
+		done_with[uid] = 0;
+		#print fmt("get_key by uid: %s", uid);
+		event SumStats::get_a_key(uid, ss$name, T);
 		}
 
 	# Schedule the next finish_epoch event.
@@ -235,51 +289,160 @@ function data_added(ss: SumStat, key: Key, result: Result)
 	if ( check_thresholds(ss, key, result, 1.0) )
 		{
 		threshold_crossed(ss, key, result);
-		event SumStats::cluster_threshold_crossed(ss$id, key, threshold_tracker[ss$id][key]);
+		event SumStats::cluster_threshold_crossed(ss$name, key, threshold_tracker[ss$name][key]);
 		}
 	}
 
-event SumStats::cluster_key_response(uid: string, ssid: string, key: Key, result: Result)
+function handle_end_of_result_collection(uid: string, ss_name: string, key: Key, cleanup: bool)
 	{
-	#print fmt("%0.6f MANAGER: receiving key data from %s - %s=%s", network_time(), get_event_peer()$descr, key2str(key), result);
-
-	# We only want to try and do a value merge if there are actually measured datapoints
-	# in the Result.
-	if ( uid in key_requests )
-		key_requests[uid] = compose_results(key_requests[uid], result);
-	else
-		key_requests[uid] = result;
-
-	# Mark that a worker is done.
-	++done_with[uid];
-
 	#print fmt("worker_count:%d :: done_with:%d", Cluster::worker_count, done_with[uid]);
-	if ( Cluster::worker_count == done_with[uid] )
+	local ss = stats_store[ss_name];
-		{
-		local ss = stats_store[ssid];
 	local ir = key_requests[uid];
 	if ( check_thresholds(ss, key, ir, 1.0) )
 		{
 		threshold_crossed(ss, key, ir);
-			event SumStats::cluster_threshold_crossed(ss$id, key, threshold_tracker[ss$id][key]);
+		event SumStats::cluster_threshold_crossed(ss_name, key, threshold_tracker[ss_name][key]);
 		}
 
-		delete done_with[uid];
+	if ( cleanup )
-		delete key_requests[uid];
+		{
-		# Check that there is an outstanding view before subtracting.
+		# This is done here because "cleanup" implicitly means
-		if ( outstanding_global_views[ssid] > 0 )
+		# it's the end of an epoch.
-			--outstanding_global_views[ssid];
+		if ( ss?$epoch_result && |ir| > 0 )
+			{
+			local now = network_time();
+			ss$epoch_result(now, key, ir);
 			}
+
+		# Check that there is an outstanding view before subtracting.
+		# Global views only apply to non-dynamic requests. Dynamic
+		# requests must be serviced.
+		if ( outstanding_global_views[ss_name] > 0 )
+			--outstanding_global_views[ss_name];
+		}
+
+	delete key_requests[uid];
+	delete done_with[uid];
+	}
+
+function request_all_current_keys(uid: string, ss_name: string, cleanup: bool)
+	{
+	#print "request_all_current_keys";
+	if ( uid in stats_keys && |stats_keys[uid]| > 0 )
+		{
+		#print fmt("    -- %d remaining keys here", |stats_keys[uid]|);
+		for ( key in stats_keys[uid] )
+			{
+			done_with[uid] = 0;
+			event SumStats::cluster_get_result(uid, ss_name, key, cleanup);
+			when ( uid in done_with && Cluster::worker_count == done_with[uid] )
+				{
+				#print "done getting result";
+				handle_end_of_result_collection(uid, ss_name, key, cleanup);
+				request_all_current_keys(uid, ss_name, cleanup);
+				}
+			delete stats_keys[uid][key];
+			break; # only a single key
+			}
+		}
+	else
+		{
+		# Get more keys!  And this breaks us out of the evented loop.
+		done_with[uid] = 0;
+		#print fmt("get_key by uid: %s", uid);
+		event SumStats::get_a_key(uid, ss_name, cleanup);
+		}
+	}
+
+event SumStats::send_no_key(uid: string, ss_name: string)
+	{
+	#print "send_no_key";
+	++done_with[uid];
+	if ( Cluster::worker_count == done_with[uid] )
+		{
+		delete done_with[uid];
+
+		if ( |stats_keys[uid]| > 0 )
+			{
+			#print "we need more keys!";
+			# Now that we have a key from each worker, lets
+			# grab all of the results.
+			request_all_current_keys(uid, ss_name, T);
+			}
+		else
+			{
+			#print "we're out of keys!";
+			local ss = stats_store[ss_name];
+			if ( ss?$epoch_finished )
+				ss$epoch_finished(network_time());
+			}
+		}
+	}
+
+event SumStats::send_a_key(uid: string, ss_name: string, key: Key)
+	{
+	#print fmt("send_a_key %s", key);
+	if ( uid !in stats_keys )
+		{
+		# no clue what happened here
+		return;
+		}
+
+	if ( key !in stats_keys[uid] )
+		add stats_keys[uid][key];
+
+	++done_with[uid];
+	if ( Cluster::worker_count == done_with[uid] )
+		{
+		delete done_with[uid];
+
+		if ( |stats_keys[uid]| > 0 )
+			{
+			#print "we need more keys!";
+			# Now that we have a key from each worker, lets
+			# grab all of the results.
+			request_all_current_keys(uid, ss_name, T);
+			}
+		else
+			{
+			#print "we're out of keys!";
+			local ss = stats_store[ss_name];
+			if ( ss?$epoch_finished )
+				ss$epoch_finished(network_time());
+			}
+		}
+	}
+
+event SumStats::cluster_send_result(uid: string, ss_name: string, key: Key, result: Result, cleanup: bool)
+	{
+	#print "cluster_send_result";
+	#print fmt("%0.6f MANAGER: receiving key data from %s - %s=%s", network_time(), get_event_peer()$descr, key2str(key), result);
+
+	# We only want to try and do a value merge if there are actually measured datapoints
+	# in the Result.
+	if ( uid !in key_requests || |key_requests[uid]| == 0 )
+		key_requests[uid] = result;
+	else
+		key_requests[uid] = compose_results(key_requests[uid], result);
+
+	# Mark that a worker is done.
+	++done_with[uid];
+
+	#if ( Cluster::worker_count == done_with[uid] )
+	#	{
+	#	print "done";
+	#	handle_end_of_result_collection(uid, ss_name, key, cleanup);
+	#	}
 	}
 
 # Managers handle intermediate updates here.
-event SumStats::cluster_key_intermediate_response(ssid: string, key: Key)
+event SumStats::cluster_key_intermediate_response(ss_name: string, key: Key)
 	{
 	#print fmt("MANAGER: receiving intermediate key data from %s", get_event_peer()$descr);
 	#print fmt("MANAGER: requesting key data for %s", key2str(key));
 
-	if ( ssid in outstanding_global_views &&
+	if ( ss_name in outstanding_global_views &&
-	     |outstanding_global_views[ssid]| > max_outstanding_global_views )
+	     |outstanding_global_views[ss_name]| > max_outstanding_global_views )
 		{
 		# Don't do this intermediate update.  Perhaps at some point in the future
 		# we will queue and randomly select from these ignored intermediate
@@ -287,60 +450,131 @@ event SumStats::cluster_key_intermediate_response(ssid: string, key: Key)
 		return;
 		}
 
-	++outstanding_global_views[ssid];
+	++outstanding_global_views[ss_name];
 
 	local uid = unique_id("");
-	event SumStats::cluster_key_request(uid, ssid, key);
+	done_with[uid] = 0;
+	event SumStats::cluster_get_result(uid, ss_name, key, F);
+	when ( uid in done_with && Cluster::worker_count == done_with[uid] )
+		{
+		handle_end_of_result_collection(uid, ss_name, key, F);
+		}
+	timeout 1.1min
+		{
+		Reporter::warning(fmt("Dynamic SumStat intermediate key request for %s (%s) took longer than 1 minute and was automatically cancelled.", ss_name, key));
 		}
 
-event SumStats::cluster_ss_response(uid: string, ssid: string, data: ResultTable, done: bool)
-	{
-	#print fmt("MANAGER: receiving results from %s", get_event_peer()$descr);
-
-	# Mark another worker as being "done" for this uid.
-	if ( done )
-		++done_with[uid];
-
-	local local_data = stats_results[uid];
-	local ss = stats_store[ssid];
-
-	for ( key in data )
-		{
-		if ( key in local_data )
-			local_data[key] = compose_results(local_data[key], data[key]);
-		else
-			local_data[key] = data[key];
-
-		# If a stat is done being collected, thresholds for each key
-		# need to be checked so we're doing it here to avoid doubly
-		# iterating over each key.
-		if ( Cluster::worker_count == done_with[uid] )
-			{
-			if ( check_thresholds(ss, key, local_data[key], 1.0) )
-				{
-				threshold_crossed(ss, key, local_data[key]);
-				event SumStats::cluster_threshold_crossed(ss$id, key, threshold_tracker[ss$id][key]);
-				}
-			}
 	}
 
-	# If the data has been collected from all peers, we are done and ready to finish.
+#event SumStats::cluster_ss_response(uid: string, ss_name: string, data: ResultTable, done: bool, cleanup: bool)
-	if ( Cluster::worker_count == done_with[uid] )
+#	{
-		{
+#	#print fmt("MANAGER: receiving results from %s", get_event_peer()$descr);
-		if ( ss?$epoch_finished )
+#
-			ss$epoch_finished(local_data);
+#	# Mark another worker as being "done" for this uid.
+#	if ( done )
+#		++done_with[uid];
+#
+#	# We had better only be getting requests for stuff that exists.
+#	if ( ss_name !in stats_store )
+#		return;
+#
+#	if ( uid !in stats_keys )
+#		stats_keys[uid] = table();
+#
+#	local local_data = stats_keys[uid];
+#	local ss = stats_store[ss_name];
+#
+#	for ( key in data )
+#		{
+#		if ( key in local_data )
+#			local_data[key] = compose_results(local_data[key], data[key]);
+#		else
+#			local_data[key] = data[key];
+#
+#		# If a stat is done being collected, thresholds for each key
+#		# need to be checked so we're doing it here to avoid doubly
+#		# iterating over each key.
+#		if ( Cluster::worker_count == done_with[uid] )
+#			{
+#			if ( check_thresholds(ss, key, local_data[key], 1.0) )
+#				{
+#				threshold_crossed(ss, key, local_data[key]);
+#				event SumStats::cluster_threshold_crossed(ss$name, key, threshold_tracker[ss$name][key]);
+#				}
+#			}
+#		}
+#
+#	# If the data has been collected from all peers, we are done and ready to finish.
+#	if ( cleanup && Cluster::worker_count == done_with[uid] )
+#		{
+#		local now = network_time();
+#		if ( ss?$epoch_result )
+#			{
+#			for ( key in local_data )
+#				ss$epoch_result(now, key, local_data[key]);
+#			}
+#
+#		if ( ss?$epoch_finished )
+#			ss$epoch_finished(now);
+#
+#		# Clean up
+#		delete stats_keys[uid];
+#		delete done_with[uid];
+#		reset(ss);
+#		}
+#	}
 
+#function request(ss_name: string): ResultTable
+#	{
+#	# This only needs to be implemented this way for cluster compatibility.
+#	local uid = unique_id("dyn-");
+#	stats_keys[uid] = table();
+#	done_with[uid] = 0;
+#	event SumStats::cluster_ss_request(uid, ss_name, F);
+#
+#	return when ( uid in done_with && Cluster::worker_count == done_with[uid] )
+#		{
+#		if ( uid in stats_keys )
+#			{
+#			local ss_result = stats_keys[uid];
+#			# Clean up
+#			delete stats_keys[uid];
+#			delete done_with[uid];
+#			reset(stats_store[ss_name]);
+#			return ss_result;
+#			}
+#		else
+#			return table();
+#		}
+#	timeout 1.1min
+#		{
+#		Reporter::warning(fmt("Dynamic SumStat request for %s took longer than 1 minute and was automatically cancelled.", ss_name));
+#		return table();
+#		}
+#	}
+
+function request_key(ss_name: string, key: Key): Result
+	{
+	local uid = unique_id("");
+	done_with[uid] = 0;
+	key_requests[uid] = table();
+
+	event SumStats::cluster_get_result(uid, ss_name, key, F);
+	return when ( uid in done_with && Cluster::worker_count == done_with[uid] )
+		{
+		#print "done with request_key";
+		local result = key_requests[uid];
 		# Clean up
-		delete stats_results[uid];
+		delete key_requests[uid];
 		delete done_with[uid];
-		# Not sure I need to reset the sumstat on the manager.
+
-		reset(ss);
+		return result;
+		}
+	timeout 1.1min
+		{
+		Reporter::warning(fmt("Dynamic SumStat key request for %s (%s) took longer than 1 minute and was automatically cancelled.", ss_name, key));
+		return table();
 		}
 	}
 
-event remote_connection_handshake_done(p: event_peer) &priority=5
-	{
-	send_id(p, "SumStats::stats_store");
-	send_id(p, "SumStats::reducer_store");
-	}
 @endif

scripts/base/frameworks/sumstats/main.bro

@@ -74,10 +74,6 @@ export {
 	## Type to store results for multiple reducers.
 	type Result: table[string] of ResultVal;
 
-	## Type to store a table of sumstats results indexed
-	## by keys.
-	type ResultTable: table[Key] of Result;
-
 	## SumStats represent an aggregation of reducers along with
 	## mechanisms to handle various situations like the epoch ending
 	## or thresholds being crossed.
@@ -87,8 +83,12 @@ export {
 	## is no assurance provided as to where the callbacks
 	## will be executed on clusters.
 	type SumStat: record {
+		## An arbitrary name for the sumstat so that it can 
+		## be referred to later.
+		name:               string;
+		
 		## The interval at which this filter should be "broken"
-		## and the '$epoch_finished' callback called.  The
+		## and the '$epoch_result' callback called.  The
 		## results are also reset at this time so any threshold
 		## based detection needs to be set to a
 		## value that should be expected to happen within
@@ -102,22 +102,28 @@ export {
 		## :bro:see:`SumStats::Result` structure which will be used
 		## for thresholding.
 		## This is required if a $threshold value is given.
-		threshold_val:      function(key: SumStats::Key, result: SumStats::Result): count &optional;
+		threshold_val:      function(key: SumStats::Key, result: SumStats::Result): double &optional;
 
 		## The threshold value for calling the
 		## $threshold_crossed callback.
-		threshold:          count             &optional;
+		threshold:          double            &optional;
 
 		## A series of thresholds for calling the
 		## $threshold_crossed callback.
-		threshold_series:   vector of count   &optional;
+		threshold_series:   vector of double  &optional;
 
 		## A callback that is called when a threshold is crossed.
 		threshold_crossed:  function(key: SumStats::Key, result: SumStats::Result) &optional;
 
-		## A callback with the full collection of Results for
+		## A callback that receives each of the results at the
-		## this SumStat.
+		## end of the analysis epoch.  The function will be 
-		epoch_finished:    function(rt: SumStats::ResultTable) &optional;
+		## called once for each key.
+		epoch_result:       function(ts: time, key: SumStats::Key, result: SumStats::Result) &optional;
+	
+		## A callback that will be called when a single collection 
+		## interval is completed.  The ts value will be the time of 
+		## when the collection started.
+		epoch_finished:     function(ts:time) &optional;
 	};
 
 	## Create a summary statistic.
@@ -134,19 +140,23 @@ export {
 	## obs: The data point to send into the stream.
 	global observe: function(id: string, key: SumStats::Key, obs: SumStats::Observation);
 
-	## This record is primarily used for internal threshold tracking.
+	## Dynamically request a sumstat key.  This function should be
-	type Thresholding: record {
+	## used sparingly and not as a replacement for the callbacks 
-		# Internal use only.  Indicates if a simple threshold was already crossed.
+	## from the :bro:see:`SumStat` record.  The function is only
-		is_threshold_crossed: bool &default=F;
+	## available for use within "when" statements as an asynchronous
-
+	## function.
-		# Internal use only.  Current key for threshold series.
+	##
-		threshold_series_index: count &default=0;
+	## ss_name: SumStat name.
-	};
+	##
+	## key: The SumStat key being requested.
+	##
+	## Returns: The result for the requested sumstat key.
+	global request_key: function(ss_name: string, key: Key): Result;
 
 	## This event is generated when thresholds are reset for a SumStat.
 	##
-	## ssid: SumStats ID that thresholds were reset for.
+	## name: SumStats name that thresholds were reset for.
-	global thresholds_reset: event(ssid: string);
+	global thresholds_reset: event(name: string);
 
 	## Helper function to represent a :bro:type:`SumStats::Key` value as
 	## a simple string.
@@ -157,18 +167,49 @@ export {
 	global key2str: function(key: SumStats::Key): string;
 }
 
+# Type to store a table of sumstats results indexed by keys.
+type ResultTable: table[Key] of Result;
+
+# The function prototype for plugins to do calculations.
+type ObserveFunc: function(r: Reducer, val: double, data: Observation, rv: ResultVal);
+
 redef record Reducer += {
-	# Internal use only.  Provides a reference back to the related SumStats by it's ID.
+	# Internal use only.  Provides a reference back to the related SumStats by its name.
-	sid: string &optional;
+	ssname: string &optional;
+
+	calc_funcs: vector of Calculation &optional;
 };
 
 # Internal use only.  For tracking thresholds per sumstat and key.
-global threshold_tracker: table[string] of table[Key] of Thresholding &optional;
+# In the case of a single threshold, 0 means the threshold isn't crossed.
+# In the case of a threshold series, the number tracks the threshold offset.
+global threshold_tracker: table[string] of table[Key] of count;
 
-redef record SumStat += {
+function increment_threshold_tracker(ss_name: string, key: Key)
-	# Internal use only (mostly for cluster coherency).
+	{
-	id: string &optional;
+	if ( ss_name !in threshold_tracker )
-};
+		threshold_tracker[ss_name] = table();
+	if ( key !in threshold_tracker[ss_name] )
+		threshold_tracker[ss_name][key] = 0;
+
+	++threshold_tracker[ss_name][key];
+	}
+
+function get_threshold_index(ss_name: string, key: Key): count
+	{
+	if ( ss_name !in threshold_tracker )
+		return 0;
+	if ( key !in threshold_tracker[ss_name] )
+		return 0;
+
+	return threshold_tracker[ss_name][key];
+	}
+
+# Prototype the hook point for plugins to initialize any result values.
+global init_resultval_hook: hook(r: Reducer, rv: ResultVal);
+
+# Prototype the hook point for plugins to merge Results.
+global compose_resultvals_hook: hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal);
 
 # Store of sumstats indexed on the sumstat id.
 global stats_store: table[string] of SumStat = table();
@@ -182,20 +223,20 @@ global result_store: table[string] of ResultTable = table();
 # Store of threshold information.
 global thresholds_store: table[string, Key] of bool = table();
 
+# Store the calculations.
+global calc_store: table[Calculation] of ObserveFunc = table();
+
+# Store the dependencies for Calculations.
+global calc_deps: table[Calculation] of vector of Calculation = table();
+
+# Hook for registering observation calculation plugins.
+global register_observe_plugins: hook();
+
 # This is called whenever key values are updated and the new val is given as the
 # `val` argument. It's only prototyped here because cluster and non-cluster have
 # separate  implementations.
 global data_added: function(ss: SumStat, key: Key, result: Result);
 
-# Prototype the hook point for plugins to do calculations.
-global observe_hook: hook(r: Reducer, val: double, data: Observation, rv: ResultVal);
-
-# Prototype the hook point for plugins to initialize any result values.
-global init_resultval_hook: hook(r: Reducer, rv: ResultVal);
-
-# Prototype the hook point for plugins to merge Results.
-global compose_resultvals_hook: hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal);
-
 # Event that is used to "finish" measurements and adapt the measurement
 # framework for clustered or non-clustered usage.
 global finish_epoch: event(ss: SumStat);
@@ -210,6 +251,24 @@ function key2str(key: Key): string
 	return fmt("sumstats_key(%s)", out);
 	}
 
+function register_observe_plugin(calc: Calculation, func: ObserveFunc)
+	{
+	calc_store[calc] = func;
+	}
+
+function add_observe_plugin_dependency(calc: Calculation, depends_on: Calculation)
+	{
+	if ( calc !in calc_deps )
+		calc_deps[calc] = vector();
+	calc_deps[calc][|calc_deps[calc]|] = depends_on;
+	}
+
+event bro_init() &priority=100000
+	{
+	# Call all of the plugin registration hooks
+	hook register_observe_plugins();
+	}
+
 function init_resultval(r: Reducer): ResultVal
 	{
 	local rv: ResultVal = [$begin=network_time(), $end=network_time()];
@@ -234,25 +293,17 @@ function compose_results(r1: Result, r2: Result): Result
 	{
 	local result: Result = table();
 
-	if ( |r1| > |r2| )
+	for ( id in r1 )
 		{
-		for ( data_id in r1 )
+		result[id] = r1[id];
-			{
-			if ( data_id in r2 )
-				result[data_id] = compose_resultvals(r1[data_id], r2[data_id]);
-			else
-				result[data_id] = r1[data_id];
 		}
-		}
+
-	else
+	for ( id in r2 )
 		{
-		for ( data_id in r2 )
+		if ( id in r1 )
-			{
+			result[id] = compose_resultvals(r1[id], r2[id]);
-			if ( data_id in r1 )
-				result[data_id] = compose_resultvals(r1[data_id], r2[data_id]);
 		else
-				result[data_id] = r2[data_id];
+			result[id] = r2[id];
-			}
 		}
 
 	return result;
@@ -261,18 +312,43 @@ function compose_results(r1: Result, r2: Result): Result
 
 function reset(ss: SumStat)
 	{
-	if ( ss$id in result_store )
+	if ( ss$name in result_store )
-		delete result_store[ss$id];
+		delete result_store[ss$name];
 
-	result_store[ss$id] = table();
+	result_store[ss$name] = table();
 
-	if ( ss?$threshold || ss?$threshold_series )
+	if ( ss$name in threshold_tracker )
 		{
-		threshold_tracker[ss$id] = table();
+		delete threshold_tracker[ss$name];
-		event SumStats::thresholds_reset(ss$id);
+		threshold_tracker[ss$name] = table();
+		event SumStats::thresholds_reset(ss$name);
 		}
 	}
 
+# This could potentially recurse forever, but plugin authors 
+# should be making sure they aren't causing reflexive dependencies.
+function add_calc_deps(calcs: vector of Calculation, c: Calculation)
+	{
+	#print fmt("Checking for deps for %s", c);
+	for ( i in calc_deps[c] )
+		{
+		local skip_calc=F;
+		for ( j in calcs )
+			{
+			if ( calcs[j] == calc_deps[c][i] )
+				skip_calc=T;
+			}
+		if ( ! skip_calc )
+			{
+			if ( calc_deps[c][i] in calc_deps )
+				add_calc_deps(calcs, calc_deps[c][i]);
+			calcs[|c|] = calc_deps[c][i];
+			#print fmt("add dep for %s [%s] ", c, calc_deps[c][i]);
+			}
+		}
+
+	}
+
 function create(ss: SumStat)
 	{
 	if ( (ss?$threshold || ss?$threshold_series) && ! ss?$threshold_val )
@@ -280,14 +356,34 @@ function create(ss: SumStat)
 		Reporter::error("SumStats given a threshold with no $threshold_val function");
 		}
 
-	if ( ! ss?$id )
+	stats_store[ss$name] = ss;
-		ss$id=unique_id("");
+
-	threshold_tracker[ss$id] = table();
+	if ( ss?$threshold || ss?$threshold_series )
-	stats_store[ss$id] = ss;
+		threshold_tracker[ss$name] = table();
 
 	for ( reducer in ss$reducers )
 		{
-		reducer$sid = ss$id;
+		reducer$ssname = ss$name;
+		reducer$calc_funcs = vector();
+		for ( calc in reducer$apply )
+			{
+			# Add in dependencies recursively.
+			if ( calc in calc_deps )
+				add_calc_deps(reducer$calc_funcs, calc);
+
+			# Don't add this calculation to the vector if 
+			# it was already added by something else as a 
+			# dependency.
+			local skip_calc=F;
+			for ( j in reducer$calc_funcs )
+				{
+				if ( calc == reducer$calc_funcs[j] )
+					skip_calc=T;
+				}
+			if ( ! skip_calc )
+				reducer$calc_funcs[|reducer$calc_funcs|] = calc;
+			}
+
 		if ( reducer$stream !in reducer_store )
 			reducer_store[reducer$stream] = set();
 		add reducer_store[reducer$stream][reducer];
@@ -313,9 +409,9 @@ function observe(id: string, key: Key, obs: Observation)
 		if ( r?$pred && ! r$pred(key, obs) )
 			next;
 
-		local ss = stats_store[r$sid];
+		local ss = stats_store[r$ssname];
 
-		# If there is a threshold and no epoch_finished callback
+		# If there is a threshold and no epoch_result callback
 		# we don't need to continue counting since the data will
 		# never be accessed.  This was leading
 		# to some state management issues when measuring
@@ -323,18 +419,21 @@ function observe(id: string, key: Key, obs: Observation)
 		# NOTE: this optimization could need removed in the
 		#       future if on demand access is provided to the
 		#       SumStats results.
-		if ( ! ss?$epoch_finished &&
+		if ( ! ss?$epoch_result &&
-		     r$sid in threshold_tracker &&
+			 r$ssname in threshold_tracker &&
-		     key in threshold_tracker[r$sid] &&
 		     ( ss?$threshold &&
-		       threshold_tracker[r$sid][key]$is_threshold_crossed ) ||
+		       key in threshold_tracker[r$ssname] &&
+		       threshold_tracker[r$ssname][key] != 0 ) ||
 		     ( ss?$threshold_series &&
-		       threshold_tracker[r$sid][key]$threshold_series_index+1 == |ss$threshold_series| ) )
+		       key in threshold_tracker[r$ssname] &&
+		       threshold_tracker[r$ssname][key] == |ss$threshold_series| ) )
+			{
 			next;
+			}
 
-		if ( r$sid !in result_store )
+		if ( r$ssname !in result_store )
-			result_store[ss$id] = table();
+			result_store[r$ssname] = table();
-		local results = result_store[r$sid];
+		local results = result_store[r$ssname];
 
 		if ( key !in results )
 			results[key] = table();
@@ -350,10 +449,13 @@ function observe(id: string, key: Key, obs: Observation)
 
 		# If a string was given, fall back to 1.0 as the value.
 		local val = 1.0;
-		if ( obs?$num || obs?$dbl )
+		if ( obs?$num )
-			val = obs?$dbl ? obs$dbl : obs$num;
+			val = obs$num;
+		else if ( obs?$dbl )
+			val = obs$dbl;
 
-		hook observe_hook(r, val, obs, result_val);
+		for ( i in r$calc_funcs )
+			calc_store[r$calc_funcs[i]](r, val, obs, result_val);
 		data_added(ss, key, result);
 		}
 	}
@@ -362,10 +464,12 @@ function observe(id: string, key: Key, obs: Observation)
 # mid-break-interval threshold crossing detection for cluster deployments.
 function check_thresholds(ss: SumStat, key: Key, result: Result, modify_pct: double): bool
 	{
-	if ( ! (ss?$threshold || ss?$threshold_series) )
+	if ( ! (ss?$threshold || ss?$threshold_series || ss?$threshold_crossed) )
 		return F;
 
 	# Add in the extra ResultVals to make threshold_vals easier to write.
+	# This length comparison should work because we just need to make 
+	# sure that we have the same number of reducers and results.
 	if ( |ss$reducers| != |result| )
 		{
 		for ( reducer in ss$reducers )
@@ -378,28 +482,21 @@ function check_thresholds(ss: SumStat, key: Key, result: Result, modify_pct: dou
 	local watch = ss$threshold_val(key, result);
 
 	if ( modify_pct < 1.0 && modify_pct > 0.0 )
-		watch = double_to_count(floor(watch/modify_pct));
+		watch = watch/modify_pct;
 
-	if ( ss$id !in threshold_tracker )
+	local t_index = get_threshold_index(ss$name, key);
-		threshold_tracker[ss$id] = table();
-	local t_tracker = threshold_tracker[ss$id];
 
-	if ( key !in t_tracker )
+	if ( ss?$threshold &&
-		{
+	     t_index == 0 && # Check that the threshold hasn't already been crossed.
-		local ttmp: Thresholding;
+	     watch >= ss$threshold )
-		t_tracker[key] = ttmp;
-		}
-	local tt = t_tracker[key];
-
-	if ( ss?$threshold && ! tt$is_threshold_crossed && watch >= ss$threshold )
 		{
 		# Value crossed the threshold.
 		return T;
 		}
 
 	if ( ss?$threshold_series &&
-	     |ss$threshold_series| >= tt$threshold_series_index &&
+	     |ss$threshold_series| > t_index && # Check if there are more thresholds.
-	     watch >= ss$threshold_series[tt$threshold_series_index] )
+	     watch >= ss$threshold_series[t_index] )
 		{
 		# A threshold series was given and the value crossed the next
 		# value in the series.
@@ -415,6 +512,8 @@ function threshold_crossed(ss: SumStat, key: Key, result: Result)
 	if ( ! ss?$threshold_crossed )
 		return;
 
+	increment_threshold_tracker(ss$name,key);
+
 	# Add in the extra ResultVals to make threshold_crossed callbacks easier to write.
 	if ( |ss$reducers| != |result| )
 		{
@@ -426,11 +525,5 @@ function threshold_crossed(ss: SumStat, key: Key, result: Result)
 		}
 
 	ss$threshold_crossed(key, result);
-	local tt = threshold_tracker[ss$id][key];
-	tt$is_threshold_crossed = T;
-
-	# Bump up to the next threshold series index if a threshold series is being used.
-	if ( ss?$threshold_series )
-		++tt$threshold_series_index;
 	}
 

scripts/base/frameworks/sumstats/non-cluster.bro

@@ -4,11 +4,20 @@ module SumStats;
 
 event SumStats::finish_epoch(ss: SumStat)
 	{
-	if ( ss$id in result_store )
+	if ( ss$name in result_store )
 		{
-		local data = result_store[ss$id];
+		local now = network_time();
+
+		if ( ss?$epoch_result )
+			{
+			local data = result_store[ss$name];
+			# TODO: don't block here.
+			for ( key in data )
+				ss$epoch_result(now, key, data[key]);
+			}
+
 		if ( ss?$epoch_finished )
-			ss$epoch_finished(data);
+			ss$epoch_finished(now);
 
 		reset(ss);
 		}
@@ -16,9 +25,32 @@ event SumStats::finish_epoch(ss: SumStat)
 	schedule ss$epoch { SumStats::finish_epoch(ss) };
 	}
 
-
 function data_added(ss: SumStat, key: Key, result: Result)
 	{
 	if ( check_thresholds(ss, key, result, 1.0) )
 		threshold_crossed(ss, key, result);
 	}
+
+function request(ss_name: string): ResultTable
+	{
+	# This only needs to be implemented this way for cluster compatibility.
+	return when ( T )
+		{
+		if ( ss_name in result_store )
+			return result_store[ss_name];
+		else
+			return table();
+		}
+	}
+
+function request_key(ss_name: string, key: Key): Result
+	{
+	# This only needs to be implemented this way for cluster compatibility.
+	return when ( T )
+		{
+		if ( ss_name in result_store && key in result_store[ss_name] )
+			return result_store[ss_name][key];
+		else
+			return table();
+		}
+	}

scripts/base/frameworks/sumstats/plugins/average.bro

@@ -1,4 +1,4 @@
-@load base/frameworks/sumstats/main
+@load ../main
 
 module SumStats;
 
@@ -14,16 +14,17 @@ export {
 	};
 }
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
 	{
-	if ( AVERAGE in r$apply )
+	register_observe_plugin(AVERAGE, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
 		if ( ! rv?$average )
 			rv$average = val;
 		else
 			rv$average += (val - rv$average) / rv$num;
+		});
 	}
-	}
+
 
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)
 	{

scripts/base/frameworks/sumstats/plugins/last.bro

@@ -33,16 +33,20 @@ function get_last(rv: ResultVal): vector of Observation
 	return s;
 	}
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
 	{
-	if ( LAST in r$apply && r$num_last_elements > 0 )
+	register_observe_plugin(LAST, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+		{
+		if ( r$num_last_elements > 0 )
 			{
 			if ( ! rv?$last_elements )
 				rv$last_elements = Queue::init([$max_len=r$num_last_elements]);
 			Queue::put(rv$last_elements, obs);
 			}
+		});
 	}
 
+
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)
 	{
 	# Merge $samples

scripts/base/frameworks/sumstats/plugins/max.bro

@@ -1,4 +1,4 @@
-@load base/frameworks/sumstats/main
+@load ../main
 
 module SumStats;
 
@@ -14,15 +14,15 @@ export {
 	};
 }
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
 	{
-	if ( MAX in r$apply )
+	register_observe_plugin(MAX, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
 		if ( ! rv?$max )
 			rv$max = val;
 		else if ( val > rv$max )
 			rv$max = val;
-		}
+		});
 	}
 
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)

scripts/base/frameworks/sumstats/plugins/min.bro

@@ -1,4 +1,4 @@
-@load base/frameworks/sumstats/main
+@load ../main
 
 module SumStats;
 
@@ -14,16 +14,17 @@ export {
 	};
 }
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
 	{
-	if ( MIN in r$apply )
+	register_observe_plugin(MIN, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
 		if ( ! rv?$min )
 			rv$min = val;
 		else if ( val < rv$min )
 			rv$min = val;
+		});
 	}
-	}
+
 
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)
 	{

scripts/base/frameworks/sumstats/plugins/sample.bro

@@ -47,15 +47,14 @@ function sample_add_sample(obs:Observation, rv: ResultVal)
 		if ( ra < rv$num_samples )
 			rv$samples[ra] = obs;
 		}
-
 	}
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
 	{
-	if ( SAMPLE in r$apply )
+	register_observe_plugin(SAMPLE, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
 		sample_add_sample(obs, rv);
-		}
+		});
 	}
 
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)
@@ -75,7 +74,6 @@ hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)
 		return;
 		}
 
-
 	if ( |rv1$samples| != num_samples && |rv2$samples| < num_samples )
 		{
 		if ( |rv1$samples| != rv1$sample_elements || |rv2$samples| < rv2$sample_elements )

scripts/base/frameworks/sumstats/plugins/std-dev.bro

@@ -1,5 +1,5 @@
-@load base/frameworks/sumstats/main
 @load ./variance
+@load ../main
 
 module SumStats;
 
@@ -21,13 +21,20 @@ function calc_std_dev(rv: ResultVal)
 		rv$std_dev = sqrt(rv$variance);
 	}
 
-# This depends on the variance plugin which uses priority -5
+hook std_dev_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal) &priority=-10
 	{
-	if ( STD_DEV in r$apply )
 	calc_std_dev(rv);
 	}
 
+hook register_observe_plugins() &priority=-10
+	{
+	register_observe_plugin(STD_DEV, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+		{
+		calc_std_dev(rv);
+		});
+	add_observe_plugin_dependency(STD_DEV, VARIANCE);
+	}
+
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal) &priority=-10
 	{
 	calc_std_dev(result);

scripts/base/frameworks/sumstats/plugins/sum.bro

@@ -1,4 +1,4 @@
-@load base/frameworks/sumstats/main
+@load ../main
 
 module SumStats;
 
@@ -14,19 +14,19 @@ export {
 		sum: double &default=0.0;
 	};
 
-	type threshold_function: function(key: SumStats::Key, result: SumStats::Result): count;
+	#type threshold_function: function(key: SumStats::Key, result: SumStats::Result): count;
-	global sum_threshold: function(data_id: string): threshold_function;
+	#global sum_threshold: function(data_id: string): threshold_function;
 }
 
-function sum_threshold(data_id: string): threshold_function
+#function sum_threshold(data_id: string): threshold_function
-	{
+#	{
-	return function(key: SumStats::Key, result: SumStats::Result): count
+#	return function(key: SumStats::Key, result: SumStats::Result): count
-		{
+#		{
-		print fmt("data_id: %s", data_id);
+#		print fmt("data_id: %s", data_id);
-		print result;
+#		print result;
-		return double_to_count(result[data_id]$sum);
+#		return double_to_count(result[data_id]$sum);
-		};
+#		};
-	}
+#	}
 
 hook init_resultval_hook(r: Reducer, rv: ResultVal)
 	{
@@ -34,10 +34,12 @@ hook init_resultval_hook(r: Reducer, rv: ResultVal)
 		rv$sum = 0;
 	}
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
+	{
+	register_observe_plugin(SUM, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
-	if ( SUM in r$apply )
 		rv$sum += val;
+		});
 	}
 
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)

scripts/base/frameworks/sumstats/plugins/topk.bro

@@ -18,18 +18,20 @@ export {
 
 }
 
+hook register_observe_plugins()
+	{
+	register_observe_plugin(TOPK, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+		{
+		topk_add(rv$topk, obs);
+		});
+	}
+
 hook init_resultval_hook(r: Reducer, rv: ResultVal)
 	{
 	if ( TOPK in r$apply && ! rv?$topk )
 		rv$topk = topk_init(r$topk_size);
 	}
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
-	{
-	if ( TOPK in r$apply )
-		topk_add(rv$topk, obs);
-	}
-
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)
 	{
 	if ( rv1?$topk )

scripts/base/frameworks/sumstats/plugins/unique.bro

@@ -1,4 +1,4 @@
-@load base/frameworks/sumstats/main
+@load ../main
 
 module SumStats;
 
@@ -23,15 +23,15 @@ redef record ResultVal += {
 	unique_vals: set[Observation] &optional;
 };
 
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal)
+hook register_observe_plugins()
 	{
-	if ( UNIQUE in r$apply )
+	register_observe_plugin(UNIQUE, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
 		if ( ! rv?$unique_vals )
 			rv$unique_vals=set();
 		add rv$unique_vals[obs];
 		rv$unique = |rv$unique_vals|;
-		}
+		});
 	}
 
 hook compose_resultvals_hook(result: ResultVal, rv1: ResultVal, rv2: ResultVal)

scripts/base/frameworks/sumstats/plugins/variance.bro

@@ -1,5 +1,5 @@
-@load base/frameworks/sumstats/main
 @load ./average
+@load ../main
 
 module SumStats;
 
@@ -28,17 +28,17 @@ function calc_variance(rv: ResultVal)
 	rv$variance = (rv$num > 1) ? rv$var_s/(rv$num-1) : 0.0;
 	}
 
-# Reduced priority since this depends on the average
+hook register_observe_plugins() &priority=-5
-hook observe_hook(r: Reducer, val: double, obs: Observation, rv: ResultVal) &priority=-5
 	{
-	if ( VARIANCE in r$apply )
+	register_observe_plugin(VARIANCE, function(r: Reducer, val: double, obs: Observation, rv: ResultVal)
 		{
 		if ( rv$num > 1 )
 			rv$var_s += ((val - rv$prev_avg) * (val - rv$average));
 
 		calc_variance(rv);
 		rv$prev_avg = rv$average;
-		}
+		});
+	add_observe_plugin_dependency(VARIANCE, AVERAGE);
 	}
 
 # Reduced priority since this depends on the average

scripts/base/init-default.bro

@@ -39,6 +39,7 @@
 @load base/frameworks/tunnels
 
 @load base/protocols/conn
+@load base/protocols/dhcp
 @load base/protocols/dnp3
 @load base/protocols/dns
 @load base/protocols/ftp

scripts/base/protocols/dhcp/__load__.bro

@@ -0,0 +1,4 @@
+@load ./consts
+@load ./main
+
+@load-sigs ./dpd.sig

scripts/base/protocols/dhcp/consts.bro

@@ -0,0 +1,20 @@
+##! Types, errors, and fields for analyzing DHCP data.  A helper file
+##! for DHCP analysis scripts.
+
+module DHCP;
+
+export {
+
+	## Types of DHCP messages. See RFC 1533.
+	const message_types = {
+		[1] = "DHCP_DISCOVER",
+		[2] = "DHCP_OFFER",
+		[3] = "DHCP_REQUEST",
+		[4] = "DHCP_DECLINE",
+		[5] = "DHCP_ACK",
+		[6] = "DHCP_NAK",
+		[7] = "DHCP_RELEASE",
+		[8] = "DHCP_INFORM",
+	} &default = function(n: count): string { return fmt("unknown-message-type-%d", n); };
+
+}

scripts/base/protocols/dhcp/dpd.sig

@@ -0,0 +1,5 @@
+signature dhcp_cookie {
+  ip-proto == udp
+  payload /^.*\x63\x82\x53\x63/
+  enable "dhcp"
+}

scripts/base/protocols/dhcp/main.bro

@@ -0,0 +1,75 @@
+##! Analyzes DHCP traffic in order to log DHCP leases given to clients.
+##! This script ignores large swaths of the protocol, since it is rather
+##! noisy on most networks, and focuses on the end-result: assigned leases.
+##!
+##! If you'd like to track known DHCP devices and to log the hostname
+##! supplied by the client, see policy/protocols/dhcp/known-devices.bro
+
+@load ./utils.bro
+
+module DHCP;
+
+export {
+	redef enum Log::ID += { LOG };
+
+	## The record type which contains the column fields of the DHCP log.
+	type Info: record {
+		## The earliest time at which a DHCP message over the
+		## associated connection is observed.
+		ts:		time		&log;
+		## A unique identifier of the connection over which DHCP is
+		## occuring.
+		uid:		string		&log;
+		## The connection's 4-tuple of endpoint addresses/ports.
+		id:		conn_id		&log;
+		## Client's hardware address.
+		mac:		string		&log &optional;
+		## Client's actual assigned IP address.
+		assigned_ip:	addr		&log &optional;
+		## IP address lease interval.
+		lease_time:	interval	&log &optional;
+		## A random number choosen by the client for this transaction.
+		trans_id:	count		&log;
+	};
+
+	## Event that can be handled to access the DHCP
+	## record as it is sent on to the logging framework.
+	global log_dhcp: event(rec: Info);
+}
+
+# Add the dhcp info to the connection record
+redef record connection += {
+	dhcp: Info &optional;
+};
+
+# 67/udp is the server's port, 68/udp the client.
+const ports = { 67/udp, 68/udp };
+redef likely_server_ports += { 67/udp };
+
+event bro_init()
+	{
+	Log::create_stream(DHCP::LOG, [$columns=Info, $ev=log_dhcp]);
+	Analyzer::register_for_ports(Analyzer::ANALYZER_DHCP, ports);
+	}
+
+event dhcp_ack(c: connection, msg: dhcp_msg, mask: addr, router: dhcp_router_list, lease: interval, serv_addr: addr, host_name: string)
+	{
+	local info: Info;
+	info$ts          = network_time();
+	info$id          = c$id;
+	info$uid         = c$uid;
+	info$lease_time  = lease;
+	info$trans_id    = msg$xid;
+
+	if ( msg$h_addr != "" )
+		info$mac = msg$h_addr;
+
+	if ( reverse_ip(msg$yiaddr) != 0.0.0.0 )
+		info$assigned_ip = reverse_ip(msg$yiaddr);
+	else
+		info$assigned_ip = c$id$orig_h;
+
+	c$dhcp = info;
+
+	Log::write(DHCP::LOG, c$dhcp);
+	}

scripts/base/protocols/dhcp/utils.bro

@@ -0,0 +1,21 @@
+##! Utilities specific for DHCP processing.
+
+@load ./main
+
+module DHCP;
+
+export {
+	## Reverse the octets of an IPv4 IP.
+	##
+	## ip: An :bro:type:`addr` IPv4 address.
+	##
+	## Returns: A reversed addr.
+	global reverse_ip: function(ip: addr): addr;
+}
+
+function reverse_ip(ip: addr): addr
+	{
+	local octets = split(cat(ip), /\./);
+	return to_addr(cat(octets[4], ".", octets[3], ".", octets[2], ".", octets[1]));
+	}
+

scripts/base/protocols/ssl/main.bro

@@ -137,8 +137,9 @@ function log_record(info: Info)
 			}
 		timeout 15secs
 			{
-			Reporter::info(fmt("SSL delay tokens not released in time (%s tokens remaining)",
+			# We are just going to log the record anyway.
-			                   |info$delay_tokens|));
+			delete info$delay_tokens;
+			log_record(info);
 			}
 		}
 	}

scripts/base/utils/dir.bro

@@ -28,7 +28,7 @@ event Dir::monitor_ev(dir: string, last_files: set[string],
                       callback: function(fname: string),
                       poll_interval: interval)
 	{
-	when ( local result = Exec::run([$cmd=fmt("ls -i \"%s/\"", str_shell_escape(dir))]) )
+	when ( local result = Exec::run([$cmd=fmt("ls -i -1 \"%s/\"", str_shell_escape(dir))]) )
 		{
 		if ( result$exit_code != 0 )
 			{

scripts/base/utils/exec.bro

@@ -163,6 +163,7 @@ function run(cmd: Command): Result
 	Input::add_event([$name=cmd$uid,
 	                  $source=fmt("%s |", cmd$cmd),
 	                  $reader=Input::READER_RAW,
+	                  $mode=Input::STREAM,
 	                  $fields=Exec::OneLine,
 	                  $ev=Exec::line,
 	                  $want_record=F,

scripts/policy/misc/app-metrics.bro

@@ -1,109 +0,0 @@
-@load base/protocols/http
-@load base/protocols/ssl
-@load base/frameworks/sumstats
-
-module AppStats;
-
-export {
-	redef enum Log::ID += { LOG };
-
-	type Info: record {
-		## Timestamp when the log line was finished and written.
-		ts:         time   &log;
-		## Time interval that the log line covers.
-		ts_delta:   interval &log;
-		## The name of the "app", like "facebook" or "netflix".
-		app:        string &log;
-		## The number of unique local hosts using the app.
-		uniq_hosts: count  &log;
-		## The number of hits to the app in total.
-		hits:       count  &log;
-		## The total number of bytes received by users of the app.
-		bytes:      count  &log;
-	};
-
-	## The frequency of logging the stats collected by this script.
-	const break_interval = 15mins &redef;
-}
-
-redef record connection += {
-	resp_hostname: string &optional;
-};
-
-event bro_init() &priority=3
-	{
-	Log::create_stream(AppStats::LOG, [$columns=Info]);
-
-	local r1: SumStats::Reducer = [$stream="apps.bytes", $apply=set(SumStats::SUM)];
-	local r2: SumStats::Reducer = [$stream="apps.hits",  $apply=set(SumStats::UNIQUE)];
-	SumStats::create([$epoch=break_interval,
-	                  $reducers=set(r1, r2),
-	                  $epoch_finished(data: SumStats::ResultTable) =
-	                  	{
-	                  	local l: Info;
-	                  	l$ts = network_time();
-	                  	l$ts_delta = break_interval;
-	                  	for ( key in data )
-	                  		{
-	                  		local result = data[key];
-	                  		l$app        = key$str;
-	                  		l$bytes      = double_to_count(floor(result["apps.bytes"]$sum));
-	                  		l$hits       = result["apps.hits"]$num;
-	                  		l$uniq_hosts = result["apps.hits"]$unique;
-	                  		Log::write(LOG, l);
-	                  		}
-	                  	}]);
-	}
-
-function add_sumstats(id: conn_id, hostname: string, size: count)
-	{
-	if ( /\.youtube\.com$/ in hostname && size > 512*1024 )
-		{
-		SumStats::observe("apps.bytes", [$str="youtube"], [$num=size]);
-		SumStats::observe("apps.hits",  [$str="youtube"], [$str=cat(id$orig_h)]);
-		}
-	else if ( /(\.facebook\.com|\.fbcdn\.net)$/ in hostname && size > 20 )
-		{
-		SumStats::observe("apps.bytes", [$str="facebook"], [$num=size]);
-		SumStats::observe("apps.hits",  [$str="facebook"], [$str=cat(id$orig_h)]);
-		}
-	else if ( /\.google\.com$/ in hostname && size > 20 )
-		{
-		SumStats::observe("apps.bytes", [$str="google"], [$num=size]);
-		SumStats::observe("apps.hits",  [$str="google"], [$str=cat(id$orig_h)]);
-		}
-	else if ( /\.nflximg\.com$/ in hostname && size > 200*1024 )
-		{
-		SumStats::observe("apps.bytes", [$str="netflix"], [$num=size]);
-		SumStats::observe("apps.hits",  [$str="netflix"], [$str=cat(id$orig_h)]);
-		}
-	else if ( /\.(pandora|p-cdn)\.com$/ in hostname && size > 512*1024 )
-		{
-		SumStats::observe("apps.bytes", [$str="pandora"], [$num=size]);
-		SumStats::observe("apps.hits",  [$str="pandora"], [$str=cat(id$orig_h)]);
-		}
-	else if ( /\.gmail\.com$/ in hostname && size > 20 )
-		{
-		SumStats::observe("apps.bytes", [$str="gmail"], [$num=size]);
-		SumStats::observe("apps.hits",  [$str="gmail"], [$str=cat(id$orig_h)]);
-		}
-}
-
-
-event ssl_established(c: connection)
-	{
-	if ( c?$ssl && c$ssl?$server_name )
-		c$resp_hostname = c$ssl$server_name;
-	}
-
-event connection_finished(c: connection)
-	{
-	if ( c?$resp_hostname )
-		add_sumstats(c$id, c$resp_hostname, c$resp$size);
-	}
-
-event HTTP::log_http(rec: HTTP::Info)
-	{
-	if( rec?$host )
-		add_sumstats(rec$id, rec$host, rec$response_body_len);
-	}

scripts/policy/misc/app-stats/__load__.bro

@@ -0,0 +1,2 @@
+@load ./main
+@load ./plugins

scripts/policy/misc/app-stats/main.bro

@@ -0,0 +1,77 @@
+#! AppStats collects information about web applications in use
+#! on the network.  
+
+@load base/protocols/http
+@load base/protocols/ssl
+@load base/frameworks/sumstats
+
+module AppStats;
+
+export {
+	redef enum Log::ID += { LOG };
+
+	type Info: record {
+		## Timestamp when the log line was finished and written.
+		ts:         time   &log;
+		## Time interval that the log line covers.
+		ts_delta:   interval &log;
+		## The name of the "app", like "facebook" or "netflix".
+		app:        string &log;
+		## The number of unique local hosts using the app.
+		uniq_hosts: count  &log;
+		## The number of hits to the app in total.
+		hits:       count  &log;
+		## The total number of bytes received by users of the app.
+		bytes:      count  &log;
+	};
+
+	## The frequency of logging the stats collected by this script.
+	const break_interval = 15mins &redef;
+}
+
+redef record connection += {
+	resp_hostname: string &optional;
+};
+
+global add_sumstats: hook(id: conn_id, hostname: string, size: count);
+
+
+event bro_init() &priority=3
+	{
+	Log::create_stream(AppStats::LOG, [$columns=Info]);
+
+	local r1: SumStats::Reducer = [$stream="apps.bytes", $apply=set(SumStats::SUM)];
+	local r2: SumStats::Reducer = [$stream="apps.hits",  $apply=set(SumStats::UNIQUE)];
+	SumStats::create([$name="app-metrics",
+	                  $epoch=break_interval,
+	                  $reducers=set(r1, r2),
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
+	                  	{
+	                  	local l: Info;
+	                  	l$ts         = network_time();
+	                  	l$ts_delta   = break_interval;
+	                  	l$app        = key$str;
+	                  	l$bytes      = double_to_count(floor(result["apps.bytes"]$sum));
+	                  	l$hits       = result["apps.hits"]$num;
+	                  	l$uniq_hosts = result["apps.hits"]$unique;
+	                  	Log::write(LOG, l);
+	                  	}]);
+	}
+
+event ssl_established(c: connection)
+	{
+	if ( c?$ssl && c$ssl?$server_name )
+		c$resp_hostname = c$ssl$server_name;
+	}
+
+event connection_finished(c: connection)
+	{
+	if ( c?$resp_hostname )
+		hook add_sumstats(c$id, c$resp_hostname, c$resp$size);
+	}
+
+event HTTP::log_http(rec: HTTP::Info)
+	{
+	if( rec?$host )
+		hook add_sumstats(rec$id, rec$host, rec$response_body_len);
+	}

scripts/policy/misc/app-stats/plugins/__load__.bro

@@ -0,0 +1,6 @@
+@load ./facebook
+@load ./gmail
+@load ./google
+@load ./netflix
+@load ./pandora
+@load ./youtube

scripts/policy/misc/app-stats/plugins/facebook.bro

@@ -0,0 +1,12 @@
+@load ../main
+
+module AppStats;
+
+hook add_sumstats(id: conn_id, hostname: string, size: count)
+	{
+	if ( /\.(facebook\.com|fbcdn\.net)$/ in hostname && size > 20 )
+		{
+		SumStats::observe("apps.bytes", [$str="facebook"], [$num=size]);
+		SumStats::observe("apps.hits",  [$str="facebook"], [$str=cat(id$orig_h)]);
+		}
+	}

scripts/policy/misc/app-stats/plugins/gmail.bro

@@ -0,0 +1,12 @@
+@load ../main
+
+module AppStats;
+
+hook add_sumstats(id: conn_id, hostname: string, size: count)
+	{
+	if ( /\.gmail\.com$/ in hostname && size > 20 )
+		{
+		SumStats::observe("apps.bytes", [$str="gmail"], [$num=size]);
+		SumStats::observe("apps.hits",  [$str="gmail"], [$str=cat(id$orig_h)]);
+		}
+	}

scripts/policy/misc/app-stats/plugins/google.bro

@@ -0,0 +1,12 @@
+@load ../main
+
+module AppStats;
+
+hook add_sumstats(id: conn_id, hostname: string, size: count)
+	{
+	if ( /\.google\.com$/ in hostname && size > 20 )
+		{
+		SumStats::observe("apps.bytes", [$str="google"], [$num=size]);
+		SumStats::observe("apps.hits",  [$str="google"], [$str=cat(id$orig_h)]);
+		}
+	}

scripts/policy/misc/app-stats/plugins/netflix.bro

@@ -0,0 +1,12 @@
+@load ../main
+
+module AppStats;
+
+hook add_sumstats(id: conn_id, hostname: string, size: count)
+	{
+	if ( /\.nflximg\.com$/ in hostname && size > 200*1024 )
+		{
+		SumStats::observe("apps.bytes", [$str="netflix"], [$num=size]);
+		SumStats::observe("apps.hits",  [$str="netflix"], [$str=cat(id$orig_h)]);
+		}
+	}

scripts/policy/misc/app-stats/plugins/pandora.bro

@@ -0,0 +1,12 @@
+@load ../main
+
+module AppStats;
+
+hook add_sumstats(id: conn_id, hostname: string, size: count)
+	{
+	if ( /\.(pandora|p-cdn)\.com$/ in hostname && size > 512*1024 )
+		{
+		SumStats::observe("apps.bytes", [$str="pandora"], [$num=size]);
+		SumStats::observe("apps.hits",  [$str="pandora"], [$str=cat(id$orig_h)]);
+		}
+	}

scripts/policy/misc/app-stats/plugins/youtube.bro

@@ -0,0 +1,12 @@
+@load ../main
+
+module AppStats;
+
+hook add_sumstats(id: conn_id, hostname: string, size: count)
+	{
+	if ( /\.youtube\.com$/ in hostname && size > 512*1024 )
+		{
+		SumStats::observe("apps.bytes", [$str="youtube"], [$num=size]);
+		SumStats::observe("apps.hits",  [$str="youtube"], [$str=cat(id$orig_h)]);
+		}
+	}

scripts/policy/misc/detect-traceroute/main.bro

@@ -29,7 +29,7 @@ export {
 	## Defines the threshold for ICMP Time Exceeded messages for a src-dst pair.
 	## This threshold only comes into play after a host is found to be
 	## sending low ttl packets.
-	const icmp_time_exceeded_threshold = 3 &redef;
+	const icmp_time_exceeded_threshold: double = 3 &redef;
 
 	## Interval at which to watch for the
 	## :bro:id:`Traceroute::icmp_time_exceeded_threshold` variable to be
@@ -57,16 +57,17 @@ event bro_init() &priority=5
 
 	local r1: SumStats::Reducer = [$stream="traceroute.time_exceeded", $apply=set(SumStats::UNIQUE)];
 	local r2: SumStats::Reducer = [$stream="traceroute.low_ttl_packet", $apply=set(SumStats::SUM)];
-	SumStats::create([$epoch=icmp_time_exceeded_interval,
+	SumStats::create([$name="traceroute-detection",
+	                  $epoch=icmp_time_exceeded_interval,
 	                  $reducers=set(r1, r2),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
 	                  	# Give a threshold value of zero depending on if the host
 	                  	# sends a low ttl packet.
 	                  	if ( require_low_ttl_packets && result["traceroute.low_ttl_packet"]$sum == 0 )
-	                  		return 0;
+	                  		return 0.0;
 	                  	else
-	                  		return result["traceroute.time_exceeded"]$unique;
+	                  		return result["traceroute.time_exceeded"]$unique+0;
 	                  	},
 	                  $threshold=icmp_time_exceeded_threshold,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =

scripts/policy/misc/known-devices.bro

@@ -0,0 +1,41 @@
+##! This script provides infrastructure for logging devices for which Bro has been
+##! able to determine the MAC address, and it logs them once per day (by default).
+##! The log that is output provides an easy way to determine a count of the devices
+##! in use on a network per day.
+##!
+##! ..note::
+##!
+##!     This script will not generate any logs on its own, it needs to be
+##!     supplied with information from elsewhere, such as
+##!     :doc:`policy/protocols/dhcp/known-devices-and-hostnames/scripts/.
+
+module Known;
+
+export {
+	## The known-hosts logging stream identifier.
+	redef enum Log::ID += { DEVICES_LOG };
+
+	## The record type which contains the column fields of the known-devices log.
+	type DevicesInfo: record {
+		## The timestamp at which the host was detected.
+		ts:      time &log;
+		## The MAC address that was detected.
+		mac:     string &log;
+	};
+
+	## The set of all known MAC addresses. It can accessed from other
+        ## to add, and check for, addresses seen in use.
+        ##
+	## We maintain each entry for 24 hours by default so that the existence of
+	## individual addressed is logged each day.
+	global known_devices: set[string] &create_expire=1day &synchronized &redef;
+
+	## An event that can be handled to access the :bro:type:`Known::DevicesInfo`
+	## record as it is sent on to the logging framework.
+	global log_known_devices: event(rec: DevicesInfo);
+}
+
+event bro_init()
+	{
+	Log::create_stream(Known::DEVICES_LOG, [$columns=DevicesInfo, $ev=log_known_devices]);
+	}

scripts/policy/misc/scan.bro

@@ -40,15 +40,11 @@ export {
 
 	## The threshold of a unique number of hosts a scanning host has to have failed
 	## connections with on a single port.
-	const addr_scan_threshold = 25 &redef;
+	const addr_scan_threshold = 25.0 &redef;
 
 	## The threshold of a number of unique ports a scanning host has to have failed
 	## connections with on a single victim host.
-	const port_scan_threshold = 15 &redef;
+	const port_scan_threshold = 15.0 &redef;
-
-	## Custom thresholds based on service for address scan.  This is primarily
-	## useful for setting reduced thresholds for specific ports.
-	const addr_scan_custom_thresholds: table[port] of count &redef;
 
 	global Scan::addr_scan_policy: hook(scanner: addr, victim: addr, scanned_port: port);
 	global Scan::port_scan_policy: hook(scanner: addr, victim: addr, scanned_port: port);
@@ -57,11 +53,12 @@ export {
 event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="scan.addr.fail", $apply=set(SumStats::UNIQUE)];
-	SumStats::create([$epoch=addr_scan_interval,
+	SumStats::create([$name="addr-scan",
+	                  $epoch=addr_scan_interval,
 	                  $reducers=set(r1),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return double_to_count(result["scan.addr.fail"]$unique);
+	                  	return result["scan.addr.fail"]$unique+0.0;
 	                  	},
 	                  #$threshold_func=check_addr_scan_threshold,
 	                  $threshold=addr_scan_threshold,
@@ -81,11 +78,12 @@ event bro_init() &priority=5
 
 	# Note: port scans are tracked similar to: table[src_ip, dst_ip] of set(port);
 	local r2: SumStats::Reducer = [$stream="scan.port.fail", $apply=set(SumStats::UNIQUE)];
-	SumStats::create([$epoch=port_scan_interval,
+	SumStats::create([$name="port-scan",
+	                  $epoch=port_scan_interval,
 	                  $reducers=set(r2),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return double_to_count(result["scan.port.fail"]$unique);
+	                  	return result["scan.port.fail"]$unique+0.0;
 	                  	},
 	                  $threshold=port_scan_threshold,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =

scripts/policy/protocols/dhcp/known-devices-and-hostnames.bro

@@ -0,0 +1,37 @@
+##! Tracks MAC address with hostnames seen in DHCP traffic. They are logged into
+##! ``devices.log``.
+
+@load policy/misc/known-devices
+
+module Known;
+
+export {
+	redef record DevicesInfo += {
+		## The value of the DHCP host name option, if seen
+		dhcp_host_name:	string		&log &optional;
+	};
+}
+
+event dhcp_request(c: connection, msg: dhcp_msg, req_addr: addr, serv_addr: addr, host_name: string)
+	{
+	if ( msg$h_addr == "" )
+		return;
+
+	if ( msg$h_addr !in known_devices )
+		{
+		add known_devices[msg$h_addr];
+		Log::write(Known::DEVICES_LOG, [$ts=network_time(), $mac=msg$h_addr, $dhcp_host_name=host_name]);
+		}
+	}
+
+event dhcp_inform(c: connection, msg: dhcp_msg, host_name: string)
+	{
+	if ( msg$h_addr == "" )
+		return;
+
+	if ( msg$h_addr !in known_devices )
+		{
+		add known_devices[msg$h_addr];
+		Log::write(Known::DEVICES_LOG, [$ts=network_time(), $mac=msg$h_addr, $dhcp_host_name=host_name]);
+		}
+	}

scripts/policy/protocols/ftp/detect-bruteforcing.bro

@@ -17,7 +17,7 @@ export {
 
 	## How many rejected usernames or passwords are required before being
 	## considered to be bruteforcing.
-	const bruteforce_threshold = 20 &redef;
+	const bruteforce_threshold: double = 20 &redef;
 
 	## The time period in which the threshold needs to be crossed before
 	## being reset.
@@ -28,11 +28,12 @@ export {
 event bro_init()
 	{
 	local r1: SumStats::Reducer = [$stream="ftp.failed_auth", $apply=set(SumStats::UNIQUE)];
-	SumStats::create([$epoch=bruteforce_measurement_interval,
+	SumStats::create([$name="ftp-detect-bruteforcing",
+	                  $epoch=bruteforce_measurement_interval,
 	                  $reducers=set(r1),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return result["ftp.failed_auth"]$num;
+	                  	return result["ftp.failed_auth"]$num+0.0;
 	                  	},
 	                  $threshold=bruteforce_threshold,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =

scripts/policy/protocols/http/detect-sqli.bro

@@ -28,7 +28,7 @@ export {
 	## Defines the threshold that determines if an SQL injection attack
 	## is ongoing based on the number of requests that appear to be SQL
 	## injection attacks.
-	const sqli_requests_threshold = 50 &redef;
+	const sqli_requests_threshold: double = 50.0 &redef;
 
 	## Interval at which to watch for the
 	## :bro:id:`HTTP::sqli_requests_threshold` variable to be crossed.
@@ -64,11 +64,12 @@ event bro_init() &priority=3
 	# determine when it looks like an actual attack and how to respond when
 	# thresholds are crossed.
 	local r1: SumStats::Reducer = [$stream="http.sqli.attacker", $apply=set(SumStats::SUM, SumStats::SAMPLE), $num_samples=collect_SQLi_samples];
-	SumStats::create([$epoch=sqli_requests_interval,
+	SumStats::create([$name="detect-sqli-attackers",
+	                  $epoch=sqli_requests_interval,
 	                  $reducers=set(r1),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return double_to_count(result["http.sqli.attacker"]$sum);
+	                  	return result["http.sqli.attacker"]$sum;
 	                  	},
 	                  $threshold=sqli_requests_threshold,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =
@@ -82,11 +83,12 @@ event bro_init() &priority=3
 	                  	}]);
 
 	local r2: SumStats::Reducer = [$stream="http.sqli.victim", $apply=set(SumStats::SUM, SumStats::SAMPLE), $num_samples=collect_SQLi_samples];
-	SumStats::create([$epoch=sqli_requests_interval,
+	SumStats::create([$name="detect-sqli-victims",
+	                  $epoch=sqli_requests_interval,
 	                  $reducers=set(r2),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return double_to_count(result["http.sqli.victim"]$sum);
+	                  	return result["http.sqli.victim"]$sum;
 	                  	},
 	                  $threshold=sqli_requests_threshold,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =

scripts/policy/protocols/ssh/detect-bruteforcing.bro

@@ -27,7 +27,7 @@ export {
 
 	## The number of failed SSH connections before a host is designated as
 	## guessing passwords.
-	const password_guesses_limit = 30 &redef;
+	const password_guesses_limit: double = 30 &redef;
 
 	## The amount of time to remember presumed non-successful logins to build
 	## model of a password guesser.
@@ -42,20 +42,29 @@ export {
 
 event bro_init()
 	{
-	local r1: SumStats::Reducer = [$stream="ssh.login.failure", $apply=set(SumStats::SUM)];
+	local r1: SumStats::Reducer = [$stream="ssh.login.failure", $apply=set(SumStats::SUM, SumStats::SAMPLE), $num_samples=5];
-	SumStats::create([$epoch=guessing_timeout,
+	SumStats::create([$name="detect-ssh-bruteforcing",
+	                  $epoch=guessing_timeout,
 	                  $reducers=set(r1),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return double_to_count(result["ssh.login.failure"]$sum);
+	                  	return result["ssh.login.failure"]$sum;
 	                  	},
 	                  $threshold=password_guesses_limit,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
 	                  	local r = result["ssh.login.failure"];
+	                  	local sub_msg = fmt("Sampled servers: ");
+	                  	local samples = r$samples;
+	                  	for ( i in samples )
+	                  		{
+	                  		if ( samples[i]?$str )
+	                  			sub_msg = fmt("%s%s %s", sub_msg, i==0 ? "":",", samples[i]$str);
+	                  		}
 	                  	# Generate the notice.
 	                  	NOTICE([$note=Password_Guessing,
 	                  	        $msg=fmt("%s appears to be guessing SSH passwords (seen in %d connections).", key$host, r$num),
+	                  	        $sub=sub_msg,
 	                  	        $src=key$host,
 	                  	        $identifier=cat(key$host)]);
 	                  	}]);
@@ -78,5 +87,5 @@ event SSH::heuristic_failed_login(c: connection)
 	# be ignored.
 	if ( ! (id$orig_h in ignore_guessers &&
 	        id$resp_h in ignore_guessers[id$orig_h]) )
-		SumStats::observe("ssh.login.failure", [$host=id$orig_h], [$num=1]);
+		SumStats::observe("ssh.login.failure", [$host=id$orig_h], [$str=cat(id$resp_h)]);
 	}

scripts/site/local.bro

@@ -11,6 +11,13 @@
 # Load the scan detection script.
 @load misc/scan
 
+# Log some information about web applications being used by users 
+# on your network.
+@load misc/app-stats
+
+# Detect traceroute being run on the network.  
+@load misc/detect-traceroute
+
 # Generate notices when vulnerable versions of software are discovered.
 # The default is to only monitor software found in the address space defined
 # as "local".  Refer to the software framework's documentation for more 

scripts/test-all-policy.bro

@@ -35,10 +35,19 @@
 @load integration/barnyard2/types.bro
 @load integration/collective-intel/__load__.bro
 @load integration/collective-intel/main.bro
-@load misc/app-metrics.bro
+@load misc/app-stats/__load__.bro
+@load misc/app-stats/main.bro
+@load misc/app-stats/plugins/__load__.bro
+@load misc/app-stats/plugins/facebook.bro
+@load misc/app-stats/plugins/gmail.bro
+@load misc/app-stats/plugins/google.bro
+@load misc/app-stats/plugins/netflix.bro
+@load misc/app-stats/plugins/pandora.bro
+@load misc/app-stats/plugins/youtube.bro
 @load misc/capture-loss.bro
 @load misc/detect-traceroute/__load__.bro
 @load misc/detect-traceroute/main.bro
+@load misc/known-devices.bro
 @load misc/load-balancing.bro
 @load misc/loaded-scripts.bro
 @load misc/profiling.bro
@@ -48,6 +57,7 @@
 @load protocols/conn/known-hosts.bro
 @load protocols/conn/known-services.bro
 @load protocols/conn/weirds.bro
+@load protocols/dhcp/known-devices-and-hostnames.bro
 @load protocols/dns/auth-addl.bro
 @load protocols/dns/detect-external-names.bro
 @load protocols/ftp/detect-bruteforcing.bro

src/analyzer/protocol/dhcp/DHCP.cc

@@ -1,4 +1,3 @@
-
 #include "DHCP.h"
 
 #include "events.bif.h"

src/analyzer/protocol/dhcp/dhcp-analyzer.pac

@@ -8,12 +8,10 @@ flow DHCP_Flow(is_orig: bool) {
 
 	%member{
 		BroVal dhcp_msg_val_;
-		BroAnalyzer interp;
 	%}
 
 	%init{
 		dhcp_msg_val_ = 0;
-		interp = connection->bro_analyzer();
 	%}
 
 	%cleanup{
@@ -45,7 +43,7 @@ flow DHCP_Flow(is_orig: bool) {
 			}
 
 		if ( type == 0 )
-			interp->Weird("DHCP_no_type_option");
+			connection()->bro_analyzer()->ProtocolViolation("no DHCP message type option");
 
 		return type;
 		%}
@@ -56,11 +54,12 @@ flow DHCP_Flow(is_orig: bool) {
 
 		// Requested IP address to the server.
 		::uint32 req_addr = 0, serv_addr = 0;
+		StringVal* host_name = 0;
 
-		for ( ptr = options->begin();
+		for ( ptr = options->begin();  ptr != options->end() && ! (*ptr)->last(); ++ptr )
-		       ptr != options->end() && ! (*ptr)->last(); ++ptr )
+			{
+			switch ( (*ptr)->code() )
 				{
-			switch ( (*ptr)->code() ) {
 				case REQ_IP_OPTION:
 					req_addr = htonl((*ptr)->info()->req_addr());
 					break;
@@ -68,40 +67,48 @@ flow DHCP_Flow(is_orig: bool) {
 				case SERV_ID_OPTION:
 					serv_addr = htonl((*ptr)->info()->serv_addr());
 					break;
+
+				case HOST_NAME_OPTION:
+					host_name = new StringVal((*ptr)->info()->host_name().length(),
+								  (const char*) (*ptr)->info()->host_name().begin());
+					break;
 				}
 			}
 
+		if ( host_name == 0 )
+			host_name = new StringVal("");
+
 		switch ( type )
 			{
 			case DHCPDISCOVER:
 				BifEvent::generate_dhcp_discover(connection()->bro_analyzer(),
 								 connection()->bro_analyzer()->Conn(),
-					dhcp_msg_val_->Ref(), new AddrVal(req_addr));
+								 dhcp_msg_val_->Ref(), new AddrVal(req_addr), host_name);
 				break;
 
 			case DHCPREQUEST:
 				BifEvent::generate_dhcp_request(connection()->bro_analyzer(),
 								connection()->bro_analyzer()->Conn(),
 								dhcp_msg_val_->Ref(), new AddrVal(req_addr),
-				new AddrVal(serv_addr));
+								new AddrVal(serv_addr), host_name);
 				break;
 
 			case DHCPDECLINE:
 				BifEvent::generate_dhcp_decline(connection()->bro_analyzer(),
 								connection()->bro_analyzer()->Conn(),
-					dhcp_msg_val_->Ref());
+								dhcp_msg_val_->Ref(), host_name);
 				break;
 
 			case DHCPRELEASE:
 				BifEvent::generate_dhcp_release(connection()->bro_analyzer(),
 								connection()->bro_analyzer()->Conn(),
-					dhcp_msg_val_->Ref());
+								dhcp_msg_val_->Ref(), host_name);
 				break;
 
 			case DHCPINFORM:
 				BifEvent::generate_dhcp_inform(connection()->bro_analyzer(),
 							       connection()->bro_analyzer()->Conn(),
-					dhcp_msg_val_->Ref());
+							       dhcp_msg_val_->Ref(), host_name);
 				break;
 			}
 
@@ -118,11 +125,13 @@ flow DHCP_Flow(is_orig: bool) {
 		::uint32 subnet_mask = 0, serv_addr = 0;
 
 		uint32 lease = 0;
+		StringVal* host_name = 0;
 
 		for ( ptr = options->begin();
 		      ptr != options->end() && ! (*ptr)->last(); ++ptr )
 			{
-			switch ( (*ptr)->code() ) {
+			switch ( (*ptr)->code() )
+				{
 				case SUBNET_OPTION:
 					subnet_mask = htonl((*ptr)->info()->mask());
 					break;
@@ -134,16 +143,16 @@ flow DHCP_Flow(is_orig: bool) {
 					router_list = new TableVal(dhcp_router_list);
 
 						{
-				int num_routers =
+						int num_routers = (*ptr)->info()->router_list()->size();
-					(*ptr)->info()->router_list()->size();
 
 						for ( int i = 0; i < num_routers; ++i )
 							{
-					vector<uint32>* rlist =
+							vector<uint32>* rlist = (*ptr)->info()->router_list();
-						(*ptr)->info()->router_list();
+
 							uint32 raddr = (*rlist)[i];
 							::uint32 tmp_addr;
 							tmp_addr = htonl(raddr);
+
 							// index starting from 1
 							Val* index = new Val(i + 1, TYPE_COUNT);
 							router_list->Assign(index, new AddrVal(tmp_addr));
@@ -159,28 +168,37 @@ flow DHCP_Flow(is_orig: bool) {
 				case SERV_ID_OPTION:
 					serv_addr = htonl((*ptr)->info()->serv_addr());
 					break;
+
+				case HOST_NAME_OPTION:
+					host_name = new StringVal((*ptr)->info()->host_name().length(),
+								  (const char*) (*ptr)->info()->host_name().begin());
+					break;
 				}
 			}
 
-		switch ( type ) {
+			if ( host_name == 0 )
+				host_name = new StringVal("");
+
+		switch ( type )
+			{
 			case DHCPOFFER:
 				BifEvent::generate_dhcp_offer(connection()->bro_analyzer(),
 							      connection()->bro_analyzer()->Conn(),
 							      dhcp_msg_val_->Ref(), new AddrVal(subnet_mask),
-					router_list, lease, new AddrVal(serv_addr));
+							      router_list, lease, new AddrVal(serv_addr), host_name);
 				break;
 
 			case DHCPACK:
 				BifEvent::generate_dhcp_ack(connection()->bro_analyzer(),
 							    connection()->bro_analyzer()->Conn(),
 							    dhcp_msg_val_->Ref(), new AddrVal(subnet_mask),
-					router_list, lease, new AddrVal(serv_addr));
+							    router_list, lease, new AddrVal(serv_addr), host_name);
 				break;
 
 			case DHCPNAK:
 				BifEvent::generate_dhcp_nak(connection()->bro_analyzer(),
 							    connection()->bro_analyzer()->Conn(),
-					dhcp_msg_val_->Ref());
+							    dhcp_msg_val_->Ref(), host_name);
 				break;
 
 			}
@@ -195,25 +213,30 @@ flow DHCP_Flow(is_orig: bool) {
 		// DHCP or BOOTP.  If not, we are unable to interpret
 		// the message options.
 		if ( ${msg.cookie} != 0x63825363 )
+			{
+			connection()->bro_analyzer()->ProtocolViolation(fmt("bad cookie (%d)", ${msg.cookie}));
 			return false;
+			}
 
 		Unref(dhcp_msg_val_);
-		RecordVal* r = new RecordVal(dhcp_msg);
 
+		const char* mac_str = fmt_mac(${msg.chaddr}.data(), ${msg.chaddr}.length());
+
+		RecordVal* r = new RecordVal(dhcp_msg);
 		r->Assign(0, new Val(${msg.op}, TYPE_COUNT));
 		r->Assign(1, new Val(${msg.type}, TYPE_COUNT));
 		r->Assign(2, new Val(${msg.xid}, TYPE_COUNT));
-
+		r->Assign(3, new StringVal(mac_str));
-		// We want only 6 bytes for Ethernet address.
-		r->Assign(3, new StringVal(6, (const char*) ${msg.chaddr}.begin()));
-
 		r->Assign(4, new AddrVal(${msg.ciaddr}));
 		r->Assign(5, new AddrVal(${msg.yiaddr}));
 
+		delete [] mac_str;
+
 		dhcp_msg_val_ = r;
 
-		switch ( ${msg.op} ) {
+		switch ( ${msg.op} )
-		case BOOTREQUEST:	// presumablye from client to server
+			{
+			case BOOTREQUEST:	// presumably from client to server
 				if ( ${msg.type} == DHCPDISCOVER ||
 			     	     ${msg.type} == DHCPREQUEST ||
 			     	     ${msg.type} == DHCPDECLINE ||
@@ -221,22 +244,28 @@ flow DHCP_Flow(is_orig: bool) {
 			     	     ${msg.type} == DHCPINFORM )
 					parse_request(${msg.options}, ${msg.type});
 				else
-				interp->Weird("DHCP_wrong_msg_type");
+					connection()->bro_analyzer()->ProtocolViolation(fmt("unknown DHCP message type option for BOOTREQUEST (%d)",
+											    ${msg.type}));
 				break;
 
 			case BOOTREPLY:		// presumably from server to client
 				if ( ${msg.type} == DHCPOFFER ||
-			     ${msg.type} == DHCPACK || ${msg.type} == DHCPNAK )
+			     	     ${msg.type} == DHCPACK ||
+				     ${msg.type} == DHCPNAK )
 					parse_reply(${msg.options}, ${msg.type});
 				else
-				interp->Weird("DHCP_wrong_msg_type");
+					connection()->bro_analyzer()->ProtocolViolation(fmt("unknown DHCP message type option for BOOTREPLY (%d)",
+											    ${msg.type}));
+
 				break;
 
 			default:
-			interp->Weird("DHCP_wrong_op_type");
+				connection()->bro_analyzer()->ProtocolViolation(fmt("unknown DHCP message op code (%d). Known codes: 1=BOOTREQUEST, 2=BOOTREPLY",
+										${msg.op}));
 				break;
 			}
 
+		connection()->bro_analyzer()->ProtocolConfirmation();
 		return true;
 		%}
 };

src/analyzer/protocol/dhcp/dhcp-protocol.pac

@@ -12,6 +12,7 @@ enum OP_type {
 enum OPTION_type {
 	SUBNET_OPTION	 = 1,
 	ROUTER_OPTION	 = 3,
+	HOST_NAME_OPTION = 12,
 	REQ_IP_OPTION	 = 50,
 	LEASE_OPTION	 = 51,
 	MSG_TYPE_OPTION  = 53,
@@ -35,12 +36,13 @@ type Option_Info(code: uint8)  = record {
 	length		: uint8;
 	value		: case code of {
 		SUBNET_OPTION	-> mask	: uint32;
-		ROUTER_OPTION   -> router_list: uint32[length/4];
+		ROUTER_OPTION	-> router_list : uint32[length/4];
 		REQ_IP_OPTION	-> req_addr	: uint32;
 		LEASE_OPTION	-> lease	: uint32;
 		MSG_TYPE_OPTION	-> msg_type	: uint8;
-		SERV_ID_OPTION  -> serv_addr: uint32;
+		SERV_ID_OPTION	-> serv_addr	: uint32;
-		default		-> other: bytestring &length = length;
+		HOST_NAME_OPTION-> host_name	: bytestring &length = length;
+		default		-> other	: bytestring &length = length;
 	};
 };
 

src/analyzer/protocol/dhcp/dhcp.pac

@@ -1,3 +1,4 @@
+%include binpac.pac
 %include bro.pac
 
 %extern{

src/analyzer/protocol/dhcp/events.bif

@@ -1,8 +1,5 @@
-## Generated for DHCP messages of type *discover*.
+## Generated for DHCP messages of type *DHCPDISCOVER* (client broadcast to locate
-##
+## available servers).
-## See `Wikipedia
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
@@ -10,33 +7,23 @@
 ##
 ## req_addr: The specific address requested by the client.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
+##
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_request dhcp_decline dhcp_ack dhcp_nak
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
+##              dhcp_release dhcp_inform
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request dns_max_queries dns_session_timeout
-##    dns_skip_addl dns_skip_all_addl dns_skip_all_auth dns_skip_auth
 ##
 ## .. note:: Bro does not support broadcast packets (as used by the DHCP
 ##    protocol). It treats broadcast addresses just like any other and
 ##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_discover%(c: connection, msg: dhcp_msg, req_addr: addr, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_discover%(c: connection, msg: dhcp_msg, req_addr: addr%);
 
-## Generated for DHCP messages of type *offer*.
+## Generated for DHCP messages of type *DHCPOFFER* (server to client in response to
-##
+## DHCPDISCOVER with offer of configuration parameters).
-## See `Wikipedia
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
-## msg: TODO.
+## msg: The parsed type-independent part of the DHCP message.
 ##
 ## mask: The subnet mask specified by the message.
 ##
@@ -46,28 +33,21 @@ event dhcp_discover%(c: connection, msg: dhcp_msg, req_addr: addr%);
 ##
 ## serv_addr: The server address specified by the message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
+##
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
+## .. bro:see:: dhcp_discover dhcp_request dhcp_decline dhcp_ack dhcp_nak
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
+##              dhcp_release dhcp_inform
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
 ## .. note:: Bro does not support broadcast packets (as used by the DHCP
 ##    protocol). It treats broadcast addresses just like any other and
 ##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_offer%(c: connection, msg: dhcp_msg, mask: addr, router: dhcp_router_list, lease: interval, serv_addr: addr, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_offer%(c: connection, msg: dhcp_msg, mask: addr, router: dhcp_router_list, lease: interval, serv_addr: addr%);
 
-## Generated for DHCP messages of type *request*.
+## Generated for DHCP messages of type *DHCPREQUEST* (Client message to servers either
-##
+## (a) requesting offered parameters from one server and implicitly declining offers
-## See `Wikipedia
+## from all others, (b) confirming correctness of previously allocated address after,
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
+## e.g., system reboot, or (c) extending the lease on a particular network address.)
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
@@ -77,55 +57,37 @@ event dhcp_offer%(c: connection, msg: dhcp_msg, mask: addr, router: dhcp_router_
 ##
 ## serv_addr: The server address specified by the message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
+##
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_decline dhcp_ack dhcp_nak
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
+##              dhcp_release dhcp_inform
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
 ## .. note:: Bro does not support broadcast packets (as used by the DHCP
 ##    protocol). It treats broadcast addresses just like any other and
 ##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_request%(c: connection, msg: dhcp_msg, req_addr: addr, serv_addr: addr, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_request%(c: connection, msg: dhcp_msg, req_addr: addr, serv_addr: addr%);
 
-## Generated for DHCP messages of type *decline*.
+## Generated for DHCP messages of type *DHCPDECLINE* (Client to server indicating
-##
+## network address is already in use).
-## See `Wikipedia
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
 ## msg: The parsed type-independent part of the DHCP message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
+##
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_request dhcp_ack dhcp_nak
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
+##              dhcp_release dhcp_inform
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
 ## .. note:: Bro does not support broadcast packets (as used by the DHCP
 ##    protocol). It treats broadcast addresses just like any other and
 ##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_decline%(c: connection, msg: dhcp_msg, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_decline%(c: connection, msg: dhcp_msg%);
 
-## Generated for DHCP messages of type *acknowledgment*.
+## Generated for DHCP messages of type *DHCPACK* (Server to client with configuration
-##
+## parameters, including committed network address).
-## See `Wikipedia
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
@@ -139,101 +101,62 @@ event dhcp_decline%(c: connection, msg: dhcp_msg%);
 ##
 ## serv_addr: The server address specified by the message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
-## .. note:: Bro does not support broadcast packets (as used by the DHCP
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_request dhcp_decline dhcp_nak
-##    protocol). It treats broadcast addresses just like any other and
+##              dhcp_release dhcp_inform
-##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_ack%(c: connection, msg: dhcp_msg, mask: addr, router: dhcp_router_list, lease: interval, serv_addr: addr, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_ack%(c: connection, msg: dhcp_msg, mask: addr, router: dhcp_router_list, lease: interval, serv_addr: addr%);
 
-## Generated for DHCP messages of type *negative acknowledgment*.
+## Generated for DHCP messages of type *DHCPNAK* (Server to client indicating client's
-##
+## notion of network address is incorrect (e.g., client has moved to new subnet) or
-## See `Wikipedia
+## client's lease has expired).
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
 ## msg: The parsed type-independent part of the DHCP message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
+##
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_request dhcp_decline dhcp_ack dhcp_release
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
+##              dhcp_inform
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
 ## .. note:: Bro does not support broadcast packets (as used by the DHCP
 ##    protocol). It treats broadcast addresses just like any other and
 ##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_nak%(c: connection, msg: dhcp_msg, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_nak%(c: connection, msg: dhcp_msg%);
 
-## Generated for DHCP messages of type *release*.
+## Generated for DHCP messages of type *DHCPRELEASE* (Client to server relinquishing
-##
+## network address and cancelling remaining lease).
-## See `Wikipedia
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
 ## msg: The parsed type-independent part of the DHCP message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
-## .. note:: Bro does not support broadcast packets (as used by the DHCP
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_request dhcp_decline dhcp_ack dhcp_nak
-##    protocol). It treats broadcast addresses just like any other and
+##              dhcp_inform
-##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_release%(c: connection, msg: dhcp_msg, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_release%(c: connection, msg: dhcp_msg%);
 
-## Generated for DHCP messages of type *inform*.
+## Generated for DHCP messages of type *DHCPINFORM* (Client to server, asking only for
-##
+## local configuration parameters; client already has externally configured network
-## See `Wikipedia
+## address).
-## <http://en.wikipedia.org/wiki/Dynamic_Host_Configuration_Protocol>`__ for
-## more information about the DHCP protocol.
 ##
 ## c: The connection record describing the underlying UDP flow.
 ##
 ## msg: The parsed type-independent part of the DHCP message.
 ##
-## .. bro:see:: dns_AAAA_reply dns_A_reply dns_CNAME_reply dns_EDNS_addl
+## host_name: The value of the host name option, if specified by the client.
-##    dns_HINFO_reply dns_MX_reply dns_NS_reply dns_PTR_reply dns_SOA_reply
+##
-##    dns_SRV_reply dns_TSIG_addl dns_TXT_reply dns_WKS_reply dns_end
+## .. bro:see:: dhcp_discover dhcp_offer dhcp_request dhcp_decline dhcp_ack dhcp_nak
-##    dns_full_request dns_mapping_altered dns_mapping_lost_name dns_mapping_new_name
+##              dhcp_release
-##    dns_mapping_unverified dns_mapping_valid dns_message dns_query_reply
-##    dns_rejected dns_request non_dns_request
 ##
 ## .. note:: Bro does not support broadcast packets (as used by the DHCP
 ##    protocol). It treats broadcast addresses just like any other and
 ##    associates packets into transport-level flows in the same way as usual.
 ##
-## .. todo:: Bro's current default configuration does not activate the protocol
+event dhcp_inform%(c: connection, msg: dhcp_msg, host_name: string%);
-##    analyzer that generates this event; the corresponding script has not yet
-##    been ported to Bro 2.x. To still enable this event, one needs to
-##    register a port for it or add a DPD payload signature.
-event dhcp_inform%(c: connection, msg: dhcp_msg%);
 

src/input/readers/Raw.cc

@@ -95,29 +95,32 @@ bool Raw::Execute()
 	else if ( childpid == 0 )
 		{
 		// we are the child.
-		close(pipes[stdout_in]);
+		safe_close(pipes[stdout_in]);
-		dup2(pipes[stdout_out], stdout_fileno);
+		if ( dup2(pipes[stdout_out], stdout_fileno) == -1 )
+			Error(Fmt("Error on dup2 stdout_out: %d", errno));
 
 		if ( stdin_towrite )
 			{
-			close(pipes[stdin_out]);
+			safe_close(pipes[stdin_out]);
-			dup2(pipes[stdin_in], stdin_fileno);
+			if ( dup2(pipes[stdin_in], stdin_fileno) == -1 )
+				Error(Fmt("Error on dup2 stdin_in: %d", errno));
 			}
 
 		if ( use_stderr )
 			{
-			close(pipes[stderr_in]);
+			safe_close(pipes[stderr_in]);
-			dup2(pipes[stderr_out], stderr_fileno);
+			if ( dup2(pipes[stderr_out], stderr_fileno) == -1 )
+				Error(Fmt("Error on dup2 stderr_out: %d", errno));
 			}
 
-		execl("/bin/sh", "sh", "-c", fname.c_str(), NULL);
+		execl("/bin/sh", "sh", "-c", fname.c_str(), (char*) NULL);
 		fprintf(stderr, "Exec failed :(......\n");
 		exit(255);
 		}
 	else
 		{
 		// we are the parent
-		close(pipes[stdout_out]);
+		safe_close(pipes[stdout_out]);
 		pipes[stdout_out] = -1;
 
 		if ( Info().mode == MODE_STREAM )
@@ -125,7 +128,7 @@ bool Raw::Execute()
 
 		if ( stdin_towrite )
 			{
-			close(pipes[stdin_in]);
+			safe_close(pipes[stdin_in]);
 			pipes[stdin_in] = -1;
 			fcntl(pipes[stdin_out], F_SETFL, O_NONBLOCK); // ya, just always set this to nonblocking. we do not want to block on a program receiving data.
 			// note that there is a small gotcha with it. More data is queued when more data is read from the program output. Hence, when having
@@ -134,7 +137,7 @@ bool Raw::Execute()
 
 		if ( use_stderr )
 			{
-			close(pipes[stderr_out]);
+			safe_close(pipes[stderr_out]);
 			pipes[stderr_out] = -1;
 			fcntl(pipes[stderr_in], F_SETFL, O_NONBLOCK); // true for this too.
 			}
@@ -195,7 +198,10 @@ bool Raw::CloseInput()
 		{
 		for ( int i = 0; i < 6; i ++ )
 			if ( pipes[i] != -1 )
-				close(pipes[i]);
+				{
+				safe_close(pipes[i]);
+				pipes[i] = -1;
+				}
 		}
 
 	file = 0;
@@ -393,11 +399,13 @@ void Raw::WriteToStdin()
 		{
 		Error(Fmt("Writing to child process stdin failed: %d. Stopping writing at position %d", errno, pos));
 		stdin_towrite = 0;
-		close(pipes[stdin_out]);
 		}
 
 	if ( stdin_towrite == 0 ) // send EOF when we are done.
-		close(pipes[stdin_out]);
+		{
+		safe_close(pipes[stdin_out]);
+		pipes[stdin_out] = -1;
+		}
 
 	if ( Info().mode == MODE_MANUAL && stdin_towrite != 0 )
 		{
@@ -528,6 +536,7 @@ bool Raw::DoUpdate()
 	if ( childpid != -1 && waitpid(childpid, &return_code, WNOHANG) != 0 )
 		{
 		// child died
+		childpid = -1;
 		bool signal = false;
 		int code = 0;
 		if ( WIFEXITED(return_code) )
@@ -539,7 +548,7 @@ bool Raw::DoUpdate()
 
 		else if ( WIFSIGNALED(return_code) )
 			{
-			signal = false;
+			signal = true;
 			code = WTERMSIG(return_code);
 			Error(Fmt("Child process exited due to signal %d", code));
 			}

src/net_util.cc

@@ -148,6 +148,26 @@ const char* fmt_conn_id(const uint32* src_addr, uint32 src_port,
 	return fmt_conn_id(src, src_port, dst, dst_port);
 	}
 
+char* fmt_mac(const unsigned char* m, int len)
+	{
+	char* buf = new char[25];
+
+	if ( len < 8 )
+		{
+		*buf = '\0';
+		return buf;
+		}
+
+	if ( m[6] == 0 && m[7] == 0 ) // EUI-48
+		snprintf(buf, 19, "%02x:%02x:%02x:%02x:%02x:%02x",
+			 m[0], m[1], m[2], m[3], m[4], m[5]);
+	else
+		snprintf(buf, 25, "%02x:%02x:%02x:%02x:%02x:%02x:%02x:%02x",
+			 m[0], m[1], m[2], m[3], m[4], m[5], m[6], m[7]);
+
+	return buf;
+	}
+
 uint32 extract_uint32(const u_char* data)
 	{
 	uint32 val;

src/net_util.h

@@ -156,6 +156,18 @@ extern const char* fmt_conn_id(const IPAddr& src_addr, uint32 src_port,
 extern const char* fmt_conn_id(const uint32* src_addr, uint32 src_port,
 				const uint32* dst_addr, uint32 dst_port);
 
+/**
+* Given a MAC address, formats it in hex as 00:de:ad:be:ef.
+* Supports both EUI-48 and EUI-64. If it's neither, returns
+* an empty string.
+*
+* @param m EUI-48 or EUI-64 MAC address to format, as a char array
+* @param len Number of bytes valid starting at *n*. This must be at
+*            least 8 for a valid address.
+* @return A string of the formatted MAC. Passes ownership to caller.
+*/
+extern char* fmt_mac(const unsigned char* m, int len);
+
 // Read 4 bytes from data and return in network order.
 extern uint32 extract_uint32(const u_char* data);
 

testing/btest/Baseline/core.print-bpf-filters/output2

@@ -26,6 +26,8 @@
 1 6667
 1 6668
 1 6669
+1 67
+1 68
 1 80
 1 8000
 1 8080
@@ -36,8 +38,8 @@
 1 992
 1 993
 1 995
-40 and
+42 and
-39 or
+41 or
-40 port
+42 port
 31 tcp
-9 udp
+11 udp

testing/btest/Baseline/core.tunnels.teredo/conn.log

@@ -3,7 +3,7 @@
 #empty_field	(empty)
 #unset_field	-
 #path	conn
-#open	2008-05-16-15-50-57
+#open	2013-08-04-03-28-45
 #fields	ts	uid	id.orig_h	id.orig_p	id.resp_h	id.resp_p	proto	service	duration	orig_bytes	resp_bytes	conn_state	local_orig	missed_bytes	history	orig_pkts	orig_ip_bytes	resp_pkts	resp_ip_bytes	tunnel_parents
 #types	time	string	addr	port	addr	port	enum	string	interval	count	count	string	bool	count	string	count	count	count	count	table[string]
 1210953047.736921	arKYeMETxOg	192.168.2.16	1576	75.126.130.163	80	tcp	-	0.000357	0	0	SHR	-	0	fA	1	40	1	40	(empty)
@@ -21,10 +21,10 @@
 1210953074.570439	c4Zw9TmAE05	192.168.2.16	1580	67.228.110.120	80	tcp	http	0.466677	469	3916	SF	-	0	ShADadFf	7	757	6	4164	(empty)
 1210953052.202579	nQcgTWjvg4c	192.168.2.16	3797	65.55.158.80	3544	udp	teredo	8.928880	129	48	SF	-	0	Dd	2	185	1	76	(empty)
 1210953060.829233	GSxOnSLghOa	192.168.2.16	3797	83.170.1.38	32900	udp	teredo	13.293994	2359	11243	SF	-	0	Dd	12	2695	13	11607	(empty)
-1210953058.933954	iE6yhOq3SF	0.0.0.0	68	255.255.255.255	67	udp	-	-	-	-	S0	-	0	D	1	328	0	0	(empty)
+1210953058.933954	iE6yhOq3SF	0.0.0.0	68	255.255.255.255	67	udp	dhcp	-	-	-	S0	-	0	D	1	328	0	0	(empty)
 1210953052.324629	TEfuqmmG4bh	192.168.2.16	3797	65.55.158.81	3544	udp	-	-	-	-	SHR	-	0	d	0	0	1	137	(empty)
 1210953046.591933	UWkUyAuUGXf	192.168.2.16	138	192.168.2.255	138	udp	-	28.448321	416	0	S0	-	0	D	2	472	0	0	(empty)
 1210953052.324629	FrJExwHcSal	fe80::8000:f227:bec8:61af	134	fe80::8000:ffff:ffff:fffd	133	icmp	-	-	-	-	OTH	-	0	-	1	88	0	0	TEfuqmmG4bh
 1210953060.829303	qCaWGmzFtM5	2001:0:4137:9e50:8000:f12a:b9c8:2815	128	2001:4860:0:2001::68	129	icmp	-	0.463615	4	4	OTH	-	0	-	1	52	1	52	GSxOnSLghOa,nQcgTWjvg4c
 1210953052.202579	j4u32Pc5bif	fe80::8000:ffff:ffff:fffd	133	ff02::2	134	icmp	-	-	-	-	OTH	-	0	-	1	64	0	0	nQcgTWjvg4c
-#close	2008-05-16-15-51-16
+#close	2013-08-04-03-28-45

testing/btest/Baseline/coverage.bare-load-baseline/canonified_loaded_scripts.log

@@ -3,7 +3,7 @@
 #empty_field	(empty)
 #unset_field	-
 #path	loaded_scripts
-#open	2013-07-29-22-37-52
+#open	2013-08-09-16-13-58
 #fields	name
 #types	string
 scripts/base/init-bare.bro
@@ -23,28 +23,28 @@ scripts/base/init-bare.bro
     build/scripts/base/bif/plugins/Bro_DCE_RPC.events.bif.bro
     build/scripts/base/bif/plugins/Bro_DHCP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_DNS.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_FTP.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_FTP.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_File.events.bif.bro
     build/scripts/base/bif/plugins/Bro_FileHash.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Finger.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_GTPv1.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_FTP.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_FTP.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_Gnutella.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_GTPv1.events.bif.bro
     build/scripts/base/bif/plugins/Bro_HTTP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_HTTP.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_ICMP.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_IRC.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Ident.events.bif.bro
     build/scripts/base/bif/plugins/Bro_InterConn.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_IRC.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Login.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Login.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_MIME.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Modbus.events.bif.bro
     build/scripts/base/bif/plugins/Bro_NCP.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_NTP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_NetBIOS.events.bif.bro
     build/scripts/base/bif/plugins/Bro_NetBIOS.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_NetFlow.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_NTP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_PIA.events.bif.bro
     build/scripts/base/bif/plugins/Bro_POP3.events.bif.bro
     build/scripts/base/bif/plugins/Bro_RPC.events.bif.bro
@@ -91,6 +91,7 @@ scripts/base/init-bare.bro
       scripts/base/utils/site.bro
         scripts/base/utils/patterns.bro
   build/scripts/base/bif/__load__.bro
+    build/scripts/base/bif/top-k.bif.bro
 scripts/policy/misc/loaded-scripts.bro
   scripts/base/utils/paths.bro
-#close	2013-07-29-22-37-52
+#close	2013-08-09-16-13-58

testing/btest/Baseline/coverage.default-load-baseline/canonified_loaded_scripts.log

@@ -3,7 +3,7 @@
 #empty_field	(empty)
 #unset_field	-
 #path	loaded_scripts
-#open	2013-07-29-22-37-53
+#open	2013-08-09-16-13-37
 #fields	name
 #types	string
 scripts/base/init-bare.bro
@@ -23,28 +23,28 @@ scripts/base/init-bare.bro
     build/scripts/base/bif/plugins/Bro_DCE_RPC.events.bif.bro
     build/scripts/base/bif/plugins/Bro_DHCP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_DNS.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_FTP.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_FTP.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_File.events.bif.bro
     build/scripts/base/bif/plugins/Bro_FileHash.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Finger.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_GTPv1.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_FTP.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_FTP.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_Gnutella.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_GTPv1.events.bif.bro
     build/scripts/base/bif/plugins/Bro_HTTP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_HTTP.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_ICMP.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_IRC.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Ident.events.bif.bro
     build/scripts/base/bif/plugins/Bro_InterConn.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_IRC.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Login.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Login.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_MIME.events.bif.bro
     build/scripts/base/bif/plugins/Bro_Modbus.events.bif.bro
     build/scripts/base/bif/plugins/Bro_NCP.events.bif.bro
-    build/scripts/base/bif/plugins/Bro_NTP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_NetBIOS.events.bif.bro
     build/scripts/base/bif/plugins/Bro_NetBIOS.functions.bif.bro
     build/scripts/base/bif/plugins/Bro_NetFlow.events.bif.bro
+    build/scripts/base/bif/plugins/Bro_NTP.events.bif.bro
     build/scripts/base/bif/plugins/Bro_PIA.events.bif.bro
     build/scripts/base/bif/plugins/Bro_POP3.events.bif.bro
     build/scripts/base/bif/plugins/Bro_RPC.events.bif.bro
@@ -91,6 +91,7 @@ scripts/base/init-bare.bro
       scripts/base/utils/site.bro
         scripts/base/utils/patterns.bro
   build/scripts/base/bif/__load__.bro
+    build/scripts/base/bif/top-k.bif.bro
 scripts/base/init-default.bro
   scripts/base/utils/active-http.bro
     scripts/base/utils/exec.bro
@@ -147,6 +148,7 @@ scripts/base/init-default.bro
       scripts/base/frameworks/sumstats/plugins/std-dev.bro
         scripts/base/frameworks/sumstats/plugins/variance.bro
       scripts/base/frameworks/sumstats/plugins/sum.bro
+      scripts/base/frameworks/sumstats/plugins/topk.bro
       scripts/base/frameworks/sumstats/plugins/unique.bro
     scripts/base/frameworks/sumstats/non-cluster.bro
   scripts/base/frameworks/tunnels/__load__.bro
@@ -156,6 +158,10 @@ scripts/base/init-default.bro
     scripts/base/protocols/conn/contents.bro
     scripts/base/protocols/conn/inactivity.bro
     scripts/base/protocols/conn/polling.bro
+  scripts/base/protocols/dhcp/__load__.bro
+    scripts/base/protocols/dhcp/consts.bro
+    scripts/base/protocols/dhcp/main.bro
+      scripts/base/protocols/dhcp/utils.bro
   scripts/base/protocols/dns/__load__.bro
     scripts/base/protocols/dns/consts.bro
     scripts/base/protocols/dns/main.bro
@@ -202,4 +208,4 @@ scripts/base/init-default.bro
     scripts/base/files/extract/main.bro
   scripts/base/misc/find-checksum-offloading.bro
 scripts/policy/misc/loaded-scripts.bro
-#close	2013-07-29-22-37-53
+#close	2013-08-09-16-13-37

testing/btest/Baseline/scripts.base.frameworks.sumstats.cluster-intermediate-update/manager-1..stdout

@@ -1,3 +1,3 @@
 A test metric threshold was crossed with a value of: 101.0
-End of epoch handler was called
 101.0
+End of epoch handler was called

testing/btest/Baseline/scripts.base.frameworks.sumstats.on-demand-cluster/manager-1..stdout

@@ -0,0 +1,2 @@
+SumStat key request
+    Host: 7.2.1.5 -> 145

testing/btest/Baseline/scripts.base.frameworks.sumstats.on-demand/.stdout

@@ -0,0 +1,2 @@
+Key request for 1.2.3.4
+    Host: 1.2.3.4 -> 42

testing/btest/Baseline/scripts.base.frameworks.sumstats.sample-cluster/manager-1..stdout

@@ -0,0 +1,8 @@
+Host: 6.5.4.3  Sampled observations: 2
+    [2, 5]
+Host: 10.10.10.10  Sampled observations: 1
+    [5]
+Host: 1.2.3.4  Sampled observations: 34
+    [5, 22, 52, 91, 94]
+Host: 7.2.1.5  Sampled observations: 2
+    [1, 91]

testing/btest/Baseline/scripts.base.frameworks.sumstats.sample-cluster/out

@@ -1,18 +0,0 @@
-1
-1.2.3.4
-10.10.10.10
-2
-2
-34
-6.5.4.3
-7.2.1.5
-[num=1, dbl=<uninitialized>, str=<uninitialized>]
-[num=2, dbl=<uninitialized>, str=<uninitialized>]
-[num=22, dbl=<uninitialized>, str=<uninitialized>]
-[num=5, dbl=<uninitialized>, str=<uninitialized>]
-[num=5, dbl=<uninitialized>, str=<uninitialized>]
-[num=5, dbl=<uninitialized>, str=<uninitialized>]
-[num=52, dbl=<uninitialized>, str=<uninitialized>]
-[num=91, dbl=<uninitialized>, str=<uninitialized>]
-[num=91, dbl=<uninitialized>, str=<uninitialized>]
-[num=94, dbl=<uninitialized>, str=<uninitialized>]

testing/btest/Baseline/scripts.base.protocols.dhcp.dhcp-all-msg-types/dhcp.log

@@ -0,0 +1,10 @@
+#separator \x09
+#set_separator	,
+#empty_field	(empty)
+#unset_field	-
+#path	dhcp
+#open	2013-07-31-21-00-49
+#fields	ts	uid	id.orig_h	id.orig_p	id.resp_h	id.resp_p	mac	assigned_ip	lease_time	trans_id
+#types	time	string	addr	port	addr	port	string	addr	interval	count
+1370200444.371332	nQcgTWjvg4c	128.2.6.189	68	128.2.6.152	67	90:b1:1c:99:49:29	128.2.6.189	900.000000	1984
+#close	2013-07-31-21-00-50

testing/btest/Baseline/scripts.base.protocols.dhcp.inform/dhcp.log

@@ -0,0 +1,10 @@
+#separator \x09
+#set_separator	,
+#empty_field	(empty)
+#unset_field	-
+#path	dhcp
+#open	2013-08-03-01-18-52
+#fields	ts	uid	id.orig_h	id.orig_p	id.resp_h	id.resp_p	mac	assigned_ip	lease_time	trans_id
+#types	time	string	addr	port	addr	port	string	addr	interval	count
+1374432420.191205	UWkUyAuUGXf	128.2.6.122	68	128.2.6.152	67	90:b1:1c:99:49:29	128.2.6.122	0.000000	2754407505
+#close	2013-08-03-01-18-52

testing/btest/Baseline/scripts.base.utils.exec/bro..stdout

@@ -3,5 +3,4 @@ test1, [exit_code=0, signal_exit=F, stdout=[done, exit, stop], stderr=<uninitial
 [out2] = [insert more text here, and there]
 }]
 test2, [exit_code=1, signal_exit=F, stdout=[here's something on stdout, some more stdout, last stdout], stderr=[and some stderr, more stderr, last stderr], files=<uninitialized>]
-test3, [exit_code=9, signal_exit=F, stdout=[FML], stderr=<uninitialized>, files=<uninitialized>]
 test4, [exit_code=0, signal_exit=F, stdout=[hibye], stderr=<uninitialized>, files=<uninitialized>]

testing/btest/Baseline/scripts.policy.protocols.dhcp.known-devices-and-hostnames.basic/known_devices.log

@@ -0,0 +1,11 @@
+#separator \x09
+#set_separator	,
+#empty_field	(empty)
+#unset_field	-
+#path	known_devices
+#open	2013-07-31-21-27-41
+#fields	ts	mac	dhcp_host_name
+#types	time	string	string
+1370200443.344965	90:b1:1c:99:49:29	btest.is.cool
+1374432420.186878	90:b1:1c:99:49:29	(empty)
+#close	2013-07-31-21-27-41

testing/btest/core/leaks/basic-cluster.bro

@@ -26,16 +26,16 @@ global n = 0;
 event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test", $apply=set(SumStats::SUM, SumStats::MIN, SumStats::MAX, SumStats::AVERAGE, SumStats::STD_DEV, SumStats::VARIANCE, SumStats::UNIQUE)];
-	SumStats::create([$epoch=5secs,
+	SumStats::create([$name="test",
+	                  $epoch=5secs,
 	                  $reducers=set(r1),
-	                  $epoch_finished(rt: SumStats::ResultTable) =
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	for ( key in rt )
+	                  	local r = result["test"];
-	                  		{
-	                  		local r = rt[key]["test"];
 	                  	print fmt("Host: %s - num:%d - sum:%.1f - avg:%.1f - max:%.1f - min:%.1f - var:%.1f - std_dev:%.1f - unique:%d", key$host, r$num, r$sum, r$average, r$max, r$min, r$variance, r$std_dev, r$unique);
-	                  		}
+	                  	},
-
+	                  $epoch_finished(ts: time) =
+	                  	{
 	                  	terminate();
 	                  	}]);
 	}

testing/btest/scripts/base/frameworks/input/raw/executestdin.bro

@@ -39,6 +39,5 @@ event bro_init()
 	try = 0;
 	outfile = open("../out");
 	Input::add_event([$source="cat > ../test.txt |", $reader=Input::READER_RAW, $mode=Input::STREAM, $name="input", $fields=Val, $ev=line, $want_record=F, $config=config_strings]);
-	Input::remove("input");
 	Input::add_event([$source="cat |", $reader=Input::READER_RAW, $mode=Input::STREAM, $name="input2", $fields=Val, $ev=line, $want_record=F, $config=config_strings]);
 	}

testing/btest/scripts/base/frameworks/sumstats/basic-cluster.bro

@@ -23,16 +23,16 @@ global n = 0;
 event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test", $apply=set(SumStats::SUM, SumStats::MIN, SumStats::MAX, SumStats::AVERAGE, SumStats::STD_DEV, SumStats::VARIANCE, SumStats::UNIQUE)];
-	SumStats::create([$epoch=5secs,
+	SumStats::create([$name="test",
+	                  $epoch=5secs,
 	                  $reducers=set(r1),
-	                  $epoch_finished(rt: SumStats::ResultTable) =
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	for ( key in rt )
+	                  	local r = result["test"];
-	                  		{
-	                  		local r = rt[key]["test"];
 	                  	print fmt("Host: %s - num:%d - sum:%.1f - avg:%.1f - max:%.1f - min:%.1f - var:%.1f - std_dev:%.1f - unique:%d", key$host, r$num, r$sum, r$average, r$max, r$min, r$variance, r$std_dev, r$unique);
-	                  		}
+	                  	},
-
+	                  $epoch_finished(ts: time) =
+	                  	{
 	                  	terminate();
 	                  	}]);
 	}

testing/btest/scripts/base/frameworks/sumstats/basic.bro

@@ -11,16 +11,14 @@ event bro_init() &priority=5
 	                                          SumStats::MIN, 
 	                                          SumStats::STD_DEV,
 	                                          SumStats::UNIQUE)];
-	SumStats::create([$epoch=3secs,
+	SumStats::create([$name="test",
+	                  $epoch=3secs,
 	                  $reducers=set(r1),
-	                     $epoch_finished(data: SumStats::ResultTable) = 
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                     	for ( key in data )
+	                  	local r = result["test.metric"];
-	                     		{
-	                     		local r = data[key]["test.metric"];
 	                  	print fmt("Host: %s - num:%d - sum:%.1f - var:%.1f - avg:%.1f - max:%.1f - min:%.1f - std_dev:%.1f - unique:%d", key$host, r$num, r$sum, r$variance, r$average, r$max, r$min, r$std_dev, r$unique);
 	                  	}
-	                     	}
 		 ]);
 
 	SumStats::observe("test.metric", [$host=1.2.3.4], [$num=5]);

testing/btest/scripts/base/frameworks/sumstats/cluster-intermediate-update.bro

@@ -20,20 +20,23 @@ redef Log::default_rotation_interval = 0secs;
 event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test.metric", $apply=set(SumStats::SUM)];
-	SumStats::create([$epoch=10secs,
+	SumStats::create([$name="test",
+	                  $epoch=10secs,
 	                  $reducers=set(r1),
-	                  $epoch_finished(data: SumStats::ResultTable) = 
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) = 
+	                  	{
+	                  	print result["test.metric"]$sum;
+	                  	},
+	                  $epoch_finished(ts: time) = 
 	                  	{
 	                  	print "End of epoch handler was called";
-				  for ( res in data ) 
-				  	print data[res]["test.metric"]$sum;
 	                  	terminate();
 	                  	},
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	return double_to_count(result["test.metric"]$sum);
+	                  	return result["test.metric"]$sum;
 	                  	},
-	                  $threshold=100,
+	                  $threshold=100.0,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
 	                  	print fmt("A test metric threshold was crossed with a value of: %.1f", result["test.metric"]$sum);

testing/btest/scripts/base/frameworks/sumstats/on-demand-cluster.bro

@@ -0,0 +1,96 @@
+# @TEST-SERIALIZE: comm
+#
+# @TEST-EXEC: btest-bg-run manager-1 BROPATH=$BROPATH:.. CLUSTER_NODE=manager-1 bro %INPUT
+# @TEST-EXEC: sleep 1
+# @TEST-EXEC: btest-bg-run worker-1  BROPATH=$BROPATH:.. CLUSTER_NODE=worker-1 bro %INPUT
+# @TEST-EXEC: btest-bg-run worker-2  BROPATH=$BROPATH:.. CLUSTER_NODE=worker-2 bro %INPUT
+# @TEST-EXEC: btest-bg-wait 15
+
+# @TEST-EXEC: btest-diff manager-1/.stdout
+
+@TEST-START-FILE cluster-layout.bro
+redef Cluster::nodes = {
+	["manager-1"] = [$node_type=Cluster::MANAGER, $ip=127.0.0.1, $p=37757/tcp, $workers=set("worker-1", "worker-2")],
+	["worker-1"]  = [$node_type=Cluster::WORKER,  $ip=127.0.0.1, $p=37760/tcp, $manager="manager-1", $interface="eth0"],
+	["worker-2"]  = [$node_type=Cluster::WORKER,  $ip=127.0.0.1, $p=37761/tcp, $manager="manager-1", $interface="eth1"],
+};
+@TEST-END-FILE
+
+redef Log::default_rotation_interval = 0secs;
+
+global n = 0;
+
+event bro_init() &priority=5
+	{
+	local r1 = SumStats::Reducer($stream="test", $apply=set(SumStats::SUM, SumStats::MIN, SumStats::MAX, SumStats::AVERAGE, SumStats::STD_DEV, SumStats::VARIANCE, SumStats::UNIQUE));
+	SumStats::create([$name="test sumstat",
+	                  $epoch=1hr,
+	                  $reducers=set(r1)]);
+	}
+
+event remote_connection_closed(p: event_peer)
+	{
+	terminate();
+	}
+
+global ready_for_data: event();
+redef Cluster::manager2worker_events += /^ready_for_data$/;
+
+event ready_for_data()
+	{
+	if ( Cluster::node == "worker-1" )
+		{
+		SumStats::observe("test", [$host=1.2.3.4], [$num=34]);
+		SumStats::observe("test", [$host=1.2.3.4], [$num=30]);
+		SumStats::observe("test", [$host=6.5.4.3], [$num=1]);
+		SumStats::observe("test", [$host=7.2.1.5], [$num=54]);
+		}
+	if ( Cluster::node == "worker-2" )
+		{
+		SumStats::observe("test", [$host=1.2.3.4], [$num=75]);
+		SumStats::observe("test", [$host=1.2.3.4], [$num=30]);
+		SumStats::observe("test", [$host=7.2.1.5], [$num=91]);
+		SumStats::observe("test", [$host=10.10.10.10], [$num=5]);
+		}
+	}
+
+
+event on_demand2()
+	{
+	local host = 7.2.1.5;
+	when ( local result = SumStats::request_key("test sumstat", [$host=host]) )
+		{
+		print "SumStat key request";
+		if ( "test" in result )
+			print fmt("    Host: %s -> %.0f", host, result["test"]$sum);
+		terminate();
+		}
+	}
+
+event on_demand()
+	{
+	#when ( local results = SumStats::request("test sumstat") )
+	#	{
+	#	print "Complete SumStat request";
+	#	print fmt("    Host: %s -> %.0f", 6.5.4.3, results[[$host=6.5.4.3]]["test"]$sum);
+	#	print fmt("    Host: %s -> %.0f", 10.10.10.10, results[[$host=10.10.10.10]]["test"]$sum);
+	#	print fmt("    Host: %s -> %.0f", 1.2.3.4, results[[$host=1.2.3.4]]["test"]$sum);
+	#	print fmt("    Host: %s -> %.0f", 7.2.1.5, results[[$host=7.2.1.5]]["test"]$sum);
+
+		event on_demand2();
+	#	}
+	}
+
+global peer_count = 0;
+event remote_connection_handshake_done(p: event_peer) &priority=-5
+	{
+	++peer_count;
+	if ( peer_count == 2 )
+		{
+		if ( Cluster::local_node_type() == Cluster::MANAGER )
+			event ready_for_data();
+
+		schedule 1sec { on_demand() };
+		}
+	}
+

testing/btest/scripts/base/frameworks/sumstats/on-demand.bro

@@ -0,0 +1,46 @@
+# @TEST-EXEC: bro %INPUT
+# @TEST-EXEC: btest-diff .stdout
+
+redef exit_only_after_terminate=T;
+
+
+## Requesting a full sumstats resulttable is not supported yet.
+#event on_demand()
+#	{
+#	when ( local results = SumStats::request("test") )
+#		{
+#		print "Complete SumStat request";
+#		for ( key in results )
+#			{
+#			print fmt("    Host: %s -> %.0f", key$host, results[key]["test.reducer"]$sum);
+#			}
+#		}
+#	}
+
+event on_demand_key()
+	{
+	local host = 1.2.3.4;
+	when ( local result = SumStats::request_key("test", [$host=host]) )
+		{
+		print fmt("Key request for %s", host);
+		print fmt("    Host: %s -> %.0f", host, result["test.reducer"]$sum);
+		terminate();
+		}
+	}
+
+event bro_init() &priority=5
+	{
+	local r1: SumStats::Reducer = [$stream="test.reducer", 
+	                               $apply=set(SumStats::SUM)];
+	SumStats::create([$name="test",
+	                  $epoch=1hr,
+	                  $reducers=set(r1)]);
+
+	# Seed some data but notice there are no callbacks defined in the sumstat!
+	SumStats::observe("test.reducer", [$host=1.2.3.4], [$num=42]);
+	SumStats::observe("test.reducer", [$host=4.3.2.1], [$num=7]);
+
+	#schedule 0.1 secs { on_demand() };
+	schedule 1 secs { on_demand_key() };
+	}
+

testing/btest/scripts/base/frameworks/sumstats/sample-cluster.bro

@@ -5,8 +5,7 @@
 # @TEST-EXEC: btest-bg-run worker-1  BROPATH=$BROPATH:.. CLUSTER_NODE=worker-1 bro %INPUT
 # @TEST-EXEC: btest-bg-run worker-2  BROPATH=$BROPATH:.. CLUSTER_NODE=worker-2 bro %INPUT
 # @TEST-EXEC: btest-bg-wait 15
-# @TEST-EXEC: cat manager-1/.stdout | sort > out
+# @TEST-EXEC: btest-diff manager-1/.stdout
-# @TEST-EXEC: btest-diff out
 
 @TEST-START-FILE cluster-layout.bro
 redef Cluster::nodes = {
@@ -18,25 +17,24 @@ redef Cluster::nodes = {
 
 redef Log::default_rotation_interval = 0secs;
 
-global n = 0;
-
 event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test", $apply=set(SumStats::SAMPLE), $num_samples=5];
-	SumStats::create([$epoch=5secs,
+	SumStats::create([$name="test",
+	                  $epoch=5secs,
 	                  $reducers=set(r1),
-	                  $epoch_finished(rt: SumStats::ResultTable) =
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                  	for ( key in rt )
+	                  	local r = result["test"];
-	                  		{
+	                  	print fmt("Host: %s  Sampled observations: %d", key$host, r$sample_elements);
-					print key$host;
+	                  	local sample_nums: vector of count = vector();
-	                     		local r = rt[key]["test"];
+	                  	for ( sample in r$samples ) 
-					for ( sample in r$samples ) {
+	                  		sample_nums[|sample_nums|] =r$samples[sample]$num;
-						print r$samples[sample];
-					}
-					print r$sample_elements;
-	                  		}
 
+	                  	print fmt("    %s", sort(sample_nums));
+	                  	},
+                      $epoch_finished(ts: time) =
+                      	{
 	                  	terminate();
 	                  	}]);
 	}

testing/btest/scripts/base/frameworks/sumstats/sample.bro

@@ -5,19 +5,16 @@ event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test.metric", 
 	                               $apply=set(SumStats::SAMPLE), $num_samples=2];
-	SumStats::create([$epoch=3secs,
+	SumStats::create([$name="test",
+	                  $epoch=3secs,
 	                  $reducers=set(r1),
-	                     $epoch_finished(data: SumStats::ResultTable) = 
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
-	                     	{
-	                     	for ( key in data )
 	                  	{
 	                  	print key$host;
-	                     		local r = data[key]["test.metric"];
+	                  	local r = result["test.metric"];
 	                  	print r$samples;
 	                  	print r$sample_elements;
-	                     		}
+	                  	}]);
-	                     	}
-		 ]);
 
 	SumStats::observe("test.metric", [$host=1.2.3.4], [$num=5]);
 	SumStats::observe("test.metric", [$host=1.2.3.4], [$num=22]);

testing/btest/scripts/base/frameworks/sumstats/thresholding.bro

@@ -8,14 +8,15 @@ redef enum Notice::Type += {
 event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test.metric", $apply=set(SumStats::SUM)];
-	SumStats::create([$epoch=3secs,
+	SumStats::create([$name="test1",
+	                  $epoch=3secs,
 	                  $reducers=set(r1),
 	                  #$threshold_val = SumStats::sum_threshold("test.metric"),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{ 
-	                  	return double_to_count(result["test.metric"]$sum);
+	                  	return result["test.metric"]$sum;
 	                  	},
-	                  $threshold=5,
+	                  $threshold=5.0,
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) = 
 	                  	{
 	                  	local r = result["test.metric"];
@@ -24,14 +25,15 @@ event bro_init() &priority=5
 	                  ]);
 
 	local r2: SumStats::Reducer = [$stream="test.metric", $apply=set(SumStats::SUM)];
-	SumStats::create([$epoch=3secs,
+	SumStats::create([$name="test2",
+	                  $epoch=3secs,
 	                  $reducers=set(r2),
 	                  #$threshold_val = SumStats::sum_threshold("test.metric"),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{ 
-	                  	return double_to_count(result["test.metric"]$sum); 
+	                  	return result["test.metric"]$sum;
 	                  	},
-	                  $threshold_series=vector(3,6,800),
+	                  $threshold_series=vector(3.0,6.0,800.0),
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) = 
 	                  	{
 	                  	local r = result["test.metric"];
@@ -41,19 +43,20 @@ event bro_init() &priority=5
 
 	local r3: SumStats::Reducer = [$stream="test.metric", $apply=set(SumStats::SUM)];
 	local r4: SumStats::Reducer = [$stream="test.metric2", $apply=set(SumStats::SUM)];
-	SumStats::create([$epoch=3secs,
+	SumStats::create([$name="test3",
+	                  $epoch=3secs,
 	                  $reducers=set(r3, r4),
 	                  $threshold_val(key: SumStats::Key, result: SumStats::Result) =
 	                  	{ 
 	                  	# Calculate a ratio between sums of two reducers.
 	                  	if ( "test.metric2" in result && "test.metric" in result &&
 	                  	     result["test.metric"]$sum > 0 )
-	                  		return double_to_count(result["test.metric2"]$sum / result["test.metric"]$sum);
+	                  		return result["test.metric2"]$sum / result["test.metric"]$sum;
 	                  	else
-	                  		return 0;
+	                  		return 0.0;
 	                  	},
 	                  # Looking for metric2 sum to be 5 times the sum of metric
-	                  $threshold=5, 
+	                  $threshold=5.0, 
 	                  $threshold_crossed(key: SumStats::Key, result: SumStats::Result) =
 	                  	{
 	                  	local thold = result["test.metric2"]$sum / result["test.metric"]$sum;

testing/btest/scripts/base/frameworks/sumstats/topk-cluster.bro

@@ -23,27 +23,24 @@ event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test.metric", 
 	                               $apply=set(SumStats::TOPK)];
-	SumStats::create([$epoch=5secs,
+	SumStats::create([$name="topk-test", 
+	                  $epoch=5secs,
 	                  $reducers=set(r1),
-	                     $epoch_finished(data: SumStats::ResultTable) = 
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                     	for ( key in data )
+	                  	local r = result["test.metric"];
-	                     		{
-	                     		local r = data[key]["test.metric"];
-					
 	                  	local s: vector of SumStats::Observation;
 	                  	s = topk_get_top(r$topk, 5);
-				
 	                  	print fmt("Top entries for key %s", key$str);
 	                  	for ( element in s ) 
 	                  		{
 	                  		print fmt("Num: %d, count: %d, epsilon: %d", s[element]$num, topk_count(r$topk, s[element]), topk_epsilon(r$topk, s[element]));
 	                  		}
-
+	                  	},
+	                  $epoch_finished(ts: time) = 
+	                  	{
 	                  	terminate();
-	                     		}
+	                  	}]);
-	                     	}
-		 ]);
 
 
 	}

testing/btest/scripts/base/frameworks/sumstats/topk.bro

@@ -5,14 +5,12 @@ event bro_init() &priority=5
 	{
 	local r1: SumStats::Reducer = [$stream="test.metric", 
 	                               $apply=set(SumStats::TOPK)];
-	SumStats::create([$epoch=3secs,
+	SumStats::create([$name="topk-test",
+	                  $epoch=3secs,
 	                  $reducers=set(r1),
-	                     $epoch_finished(data: SumStats::ResultTable) = 
+	                  $epoch_result(ts: time, key: SumStats::Key, result: SumStats::Result) =
 	                  	{
-	                     	for ( key in data )
+	                  	local r = result["test.metric"];
-	                     		{
-	                     		local r = data[key]["test.metric"];
-					
 	                  	local s: vector of SumStats::Observation;
 	                  	s = topk_get_top(r$topk, 5);
 	                  	
@@ -21,10 +19,7 @@ event bro_init() &priority=5
 	                  		{
 	                  		print fmt("Num: %d, count: %d, epsilon: %d", s[element]$num, topk_count(r$topk, s[element]), topk_epsilon(r$topk, s[element]));
 	                  		}
-
+	                  	}]);
-	                     		}
-	                     	}
-		 ]);
 
 
 	const loop_v: vector of count = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100};

testing/btest/scripts/base/protocols/dhcp/dhcp-all-msg-types.btest

@@ -0,0 +1,6 @@
+# This tests that DHCP leases are logged in dhcp.log
+# The trace has a message of each DHCP message type,
+# but only one lease should show up in the logs.
+
+# @TEST-EXEC: bro -r $TRACES/dhcp/dhcp.trace %INPUT
+# @TEST-EXEC: btest-diff dhcp.log

testing/btest/scripts/base/protocols/dhcp/inform.test

@@ -0,0 +1,5 @@
+# DHCPINFORM leases are special-cased in the code.
+# This tests that those leases are correctly logged.
+
+# @TEST-EXEC: bro -r $TRACES/dhcp/dhcp_inform.trace %INPUT
+# @TEST-EXEC: btest-diff dhcp.log

testing/btest/scripts/base/utils/active-http.test

@@ -1,4 +1,3 @@
-# @TEST-REQUIRES: which httpd
 # @TEST-REQUIRES: which python
 #
 # @TEST-EXEC: btest-bg-run httpd python $SCRIPTS/httpd.py --max 1
@@ -8,7 +7,7 @@
 # @TEST-EXEC: btest-diff bro/.stdout
 
 @load base/utils/active-http
-
+@load base/frameworks/communication # let network-time run. otherwise there are no heartbeats...
 redef exit_only_after_terminate = T;
 
 event bro_init()

testing/btest/scripts/base/utils/dir.test

@@ -1,11 +1,11 @@
 # @TEST-EXEC: btest-bg-run bro bro -b ../dirtest.bro
-# @TEST-EXEC: btest-bg-wait 10
+# @TEST-EXEC: btest-bg-wait 15
 # @TEST-EXEC: TEST_DIFF_CANONIFIER=$SCRIPTS/diff-sort btest-diff bro/.stdout
 
 @TEST-START-FILE dirtest.bro
 
 @load base/utils/dir
-
+@load base/frameworks/communication # let network-time run. otherwise there are no heartbeats...
 redef exit_only_after_terminate = T;
 
 global c: count = 0;
@@ -33,14 +33,20 @@ function new_file2(fname: string)
 event change_things()
 	{
 	system("touch ../testdir/newone");
-	system("rm ../testdir/bye && touch ../testdir/bye");
+	system("rm ../testdir/bye");
+	}
+
+event change_things2()
+	{
+	system("touch ../testdir/bye");
 	}
 
 event bro_init()
 	{
 	Dir::monitor("../testdir", new_file1, .5sec);
 	Dir::monitor("../testdir", new_file2, 1sec);
-	schedule 1sec { change_things() };
+	schedule 3sec { change_things() };
+	schedule 6sec { change_things2() };
 	}
 
 @TEST-END-FILE

testing/btest/scripts/base/utils/exec.test

@@ -1,11 +1,11 @@
 # @TEST-EXEC: btest-bg-run bro bro -b ../exectest.bro
-# @TEST-EXEC: btest-bg-wait 10
+# @TEST-EXEC: btest-bg-wait 15
 # @TEST-EXEC: TEST_DIFF_CANONIFIER=$SCRIPTS/diff-sort btest-diff bro/.stdout
 
 @TEST-START-FILE exectest.bro
 
 @load base/utils/exec
-
+@load base/frameworks/communication # let network-time run. otherwise there are no heartbeats...
 redef exit_only_after_terminate = T;
 
 global c: count = 0;
@@ -14,7 +14,7 @@ function check_exit_condition()
 	{
 	c += 1;
 
-	if ( c == 4 )
+	if ( c == 3 )
 		terminate();
 	}
 
@@ -32,7 +32,8 @@ event bro_init()
 	test_cmd("test1", [$cmd="bash ../somescript.sh",
 	                   $read_files=set("out1", "out2")]);
 	test_cmd("test2", [$cmd="bash ../nofiles.sh"]);
-	test_cmd("test3", [$cmd="bash ../suicide.sh"]);
+	# Not sure of a portable way to test signals yet.
+	#test_cmd("test3", [$cmd="bash ../suicide.sh"]);
 	test_cmd("test4", [$cmd="bash ../stdin.sh", $stdin="hibye"]);
 	}
 

testing/btest/scripts/policy/protocols/dhcp/known-devices-and-hostnames/basic.test

@@ -0,0 +1,8 @@
+# This tests that the known_devices log is created,
+# that devices are logged by MAC address, and that
+# the DHCP hostname is added, if available.
+
+# @TEST-EXEC: bro -r $TRACES/dhcp/dhcp.trace -r $TRACES/dhcp/dhcp_inform.trace %INPUT
+# @TEST-EXEC: btest-diff known_devices.log
+
+@load policy/protocols/dhcp/known-devices-and-hostnames