Merge remote-tracking branch 'origin/topic/jsiwek/empty-lines'

* origin/topic/jsiwek/empty-lines:
  Add 'smtp_excessive_pending_cmds' weird
  Fix SMTP command string comparisons
  Improve handling of empty lines in several text protocol analyzers
  Add rate-limiting sampling mechanism for weird events
  Teach timestamp canonifier about timestamps before ~2001

NEWS

@@ -409,6 +409,7 @@ Changed Functionality
 - The string_to_pattern() built-in (and the now-deprecated merge_pattern()
   built-in) is no longer restricted to only be called at initialization time.
 
+
 - GeoIP Legacy Database support has been replaced with GeoIP2 MaxMind DB
   format support.
 
@@ -418,6 +419,21 @@ Changed Functionality
     after January 2, 2019.  It's also noted that all GeoIP Legacy databases
     may be discontinued as they are superseded by GeoIP2.
 
+- "Weird" events are now generally suppressed/sampled by default according to
+  some tunable parameters:
+
+  - Weird::sampling_whitelist
+  - Weird::sampling_threshold
+  - Weird::sampling_rate
+  - Weird::sampling_duration
+
+  Those options can be changed if one needs the previous behavior of
+  a "net_weird", "flow_weird", or "conn_weird" event being raised for
+  every single event.  Otherwise, there is a new weird_stats.log which
+  contains concise summaries of weird counts per type per time period
+  and the original weird.log may not differ much either, except in
+  the cases where a particular weird type exceeds the sampling threshold.
+
 Removed Functionality
 ---------------------
 

doc/script-reference/log-files.rst

@@ -152,6 +152,8 @@ Miscellaneous
 +----------------------------+---------------------------------------+---------------------------------+
 | weird.log                  | Unexpected network-level activity     | :bro:type:`Weird::Info`         |
 +----------------------------+---------------------------------------+---------------------------------+
+| weird-stats.log            | Statistics about unexpected activity  | :bro:type:`WeirdStats::Info`    |
++----------------------------+---------------------------------------+---------------------------------+
 
 Bro Diagnostics
 ---------------

scripts/base/init-bare.bro

@@ -82,6 +82,13 @@ type addr_vec: vector of addr;
 ##    directly and then remove this alias.
 type table_string_of_string: table[string] of string;
 
+## A table of counts indexed by strings.
+##
+## .. todo:: We need this type definition only for declaring builtin functions
+##    via ``bifcl``. We should extend ``bifcl`` to understand composite types
+##    directly and then remove this alias.
+type table_string_of_count: table[string] of count;
+
 ## A set of file analyzer tags.
 ##
 ## .. todo:: We need this type definition only for declaring builtin functions
@@ -626,6 +633,17 @@ type BrokerStats: record {
 	num_ids_outgoing: count;
 };
 
+## Statistics about reporter messages and weirds.
+##
+## .. bro:see:: get_reporter_stats
+type ReporterStats: record {
+	## Number of total weirds encountered, before any rate-limiting.
+	weirds: count;
+	## Number of times each individual weird is encountered, before any
+	## rate-limiting is applied.
+	weirds_by_type:	table[string] of count;
+};
+
 ## Deprecated.
 ##
 ## .. todo:: Remove. It's still declared internally but doesn't seem  used anywhere
@@ -4826,6 +4844,35 @@ export {
 	type Cluster::Pool: record {};
 }
 
+module Weird;
+export {
+	## Prevents rate-limiting sampling of any weirds named in the table.
+	const sampling_whitelist: set[string] &redef;
+
+	## How many weirds of a given type to tolerate before sampling begins.
+	## i.e. this many consecutive weirds of a given type will be allowed to
+	## raise events for script-layer handling before being rate-limited.
+	const sampling_threshold = 25 &redef;
+
+	## The rate-limiting sampling rate.  One out of every of this number of
+	## rate-limited weirds of a given type will be allowed to raise events
+	## for further script-layer handling.
+	const sampling_rate = 1000 &redef;
+
+	## How long a weird of a given type is allowed to keep state/counters in
+	## memory.  For "net" weirds an expiration timer starts per weird name when
+	## first initializing its counter.  For "flow" weirds an expiration timer
+	## starts once per src/dst IP pair for the first weird of any name.  For
+	## "conn" weirds, counters and expiration timers are kept for the duration
+	## of the connection for each named weird and reset when necessary.  e.g.
+	## if a "conn" weird by the name of "foo" is seen more than
+	## :bro:see:`Weird::sampling_threshold` times, then an expiration timer
+	## begins for "foo" and upon triggering will reset the counter for "foo"
+	## and unthrottle its rate-limiting until it once again exceeds the
+	## threshold.
+	const sampling_duration = 10min &redef;
+}
+
 module GLOBAL;
 
 ## Seed for hashes computed internally for probabilistic data structures. Using

scripts/base/protocols/http/main.bro

@@ -199,7 +199,7 @@ event http_request(c: connection, method: string, original_URI: string,
 	c$http$uri = unescaped_URI;
 
 	if ( method !in http_methods )
-		event conn_weird("unknown_HTTP_method", c, method);
+		Reporter::conn_weird("unknown_HTTP_method", c, method);
 	}
 
 event http_reply(c: connection, version: string, code: count, reason: string) &priority=5

scripts/base/protocols/sip/main.bro

@@ -168,7 +168,7 @@ event sip_request(c: connection, method: string, original_URI: string, version:
 	c$sip$uri = original_URI;
 
 	if ( method !in sip_methods )
-		event conn_weird("unknown_SIP_method", c, method);
+		Reporter::conn_weird("unknown_SIP_method", c, method);
 	}
 
 event sip_reply(c: connection, version: string, code: count, reason: string) &priority=5

scripts/base/protocols/smb/smb1-main.bro

@@ -309,7 +309,7 @@ event smb_pipe_request(c: connection, hdr: SMB1::Header, op_num: count)
 	if ( ! f?$uuid )
 		{
 		# TODO: figure out why this is happening.
-		event conn_weird("smb_pipe_request_missing_uuid", c, "");
+		Reporter::conn_weird("smb_pipe_request_missing_uuid", c, "");
 		return;
 		}
 	local arg = fmt("%s: %s",

scripts/base/protocols/ssl/main.bro

@@ -263,7 +263,7 @@ event ssl_extension_server_name(c: connection, is_orig: bool, names: string_vec)
 		{
 		c$ssl$server_name = names[0];
 		if ( |names| > 1 )
-			event conn_weird("SSL_many_server_names", c, cat(names));
+			Reporter::conn_weird("SSL_many_server_names", c, cat(names));
 		}
 	}
 

scripts/policy/misc/weird-stats.bro

@@ -0,0 +1,100 @@
+##! Log weird statistics.
+
+@load base/frameworks/sumstats
+@load base/frameworks/cluster
+
+module WeirdStats;
+
+export {
+	redef enum Log::ID += { LOG };
+
+	## How often stats are reported.
+	const weird_stat_interval = 15min &redef;
+
+	type Info: record {
+		## Timestamp for the measurement.
+		ts: time &log;
+		## Name of the weird.
+		name: string &log;
+		## Number of times weird was seen since the last stats interval.
+		num_seen: count &log;
+	};
+
+	global log_weird_stats: event(rec: Info);
+}
+
+global this_epoch_weirds: table[string] of double;
+global last_epoch_weirds: table[string] of double;
+
+function weird_epoch_results(ts: time, key: SumStats::Key, result: SumStats::Result)
+	{
+	this_epoch_weirds[key$str]=result["weirds.encountered"]$sum;
+	}
+
+function weird_epoch_finished(ts: time)
+	{
+	for ( n in this_epoch_weirds )
+		{
+		local last_count: double = 0.0;
+
+		if ( n in last_epoch_weirds )
+			last_count = last_epoch_weirds[n];
+
+		local num_seen: double = this_epoch_weirds[n] - last_count;
+
+		if ( num_seen > 0.0 )
+			Log::write(LOG, Info($ts = ts, $name = n,
+			                     $num_seen = double_to_count(num_seen)));
+		}
+
+	last_epoch_weirds = this_epoch_weirds;
+	this_epoch_weirds = table();
+	}
+
+event bro_init() &priority=5
+	{
+	Log::create_stream(WeirdStats::LOG,
+	                   [$columns = Info, $ev = log_weird_stats,
+	                    $path="weird_stats"]);
+	local r1 = SumStats::Reducer($stream = "weirds.encountered",
+	                             $apply = set(SumStats::SUM));
+	SumStats::create([$name = "weirds.statistics",
+	                  $epoch = weird_stat_interval, $reducers = set(r1),
+	                  $epoch_result = weird_epoch_results,
+	                  $epoch_finished = weird_epoch_finished]);
+	}
+
+module SumStats;
+
+function observe_weird_stats()
+	{
+	local rs = get_reporter_stats();
+
+	for ( n in rs$weirds_by_type )
+		SumStats::observe("weirds.encountered", SumStats::Key($str = n),
+		                  SumStats::Observation($dbl=rs$weirds_by_type[n]+0.0));
+	}
+
+@if ( Cluster::is_enabled() )
+
+# I'm not sure if this is a hack or not: the manager will generate this
+# event at the end of its epoch so workers can handle it just in time to
+# generate the necessary stats.  Alternative may be workers generating the
+# stats individually/proactively in their own finish_epoch, but that may be
+# less synchronized?
+event SumStats::cluster_ss_request(uid: string, ss_name: string, cleanup: bool) &priority=10
+	{
+	observe_weird_stats();
+	}
+
+@else
+
+event SumStats::finish_epoch(ss: SumStat) &priority=10
+	{
+	if ( ss$name != "weirds.statistics" )
+		return;
+
+	observe_weird_stats();
+	}
+
+@endif

scripts/test-all-policy.bro

@@ -54,6 +54,7 @@
 @load misc/profiling.bro
 @load misc/scan.bro
 @load misc/stats.bro
+@load misc/weird-stats.bro
 @load misc/trim-trace-file.bro
 @load protocols/conn/known-hosts.bro
 @load protocols/conn/known-services.bro

src/Conn.cc

@@ -1071,3 +1071,28 @@ void Connection::CheckFlowLabel(bool is_orig, uint32 flow_label)
 	else
 		saw_first_resp_packet = 1;
 	}
+
+bool Connection::PermitWeird(const char* name, uint64 threshold, uint64 rate,
+                             double duration)
+	{
+	auto& state = weird_state[name];
+	++state.count;
+
+	if ( state.count < threshold )
+		return true;
+
+	if ( state.count == threshold )
+		state.sampling_start_time = network_time;
+	else
+		{
+		if ( network_time > state.sampling_start_time + duration )
+			{
+			state.sampling_start_time = 0;
+			state.count = 1;
+			return true;
+			}
+		}
+
+	auto num_above_threshold = state.count - threshold;
+	return num_above_threshold % rate == 0;
+	}

src/Conn.h

@@ -5,6 +5,9 @@
 
 #include <sys/types.h>
 
+#include <unordered_map>
+#include <string>
+
 #include "Dict.h"
 #include "Val.h"
 #include "Timer.h"
@@ -275,6 +278,9 @@ public:
 	uint32 GetOrigFlowLabel() { return orig_flow_label; }
 	uint32 GetRespFlowLabel() { return resp_flow_label; }
 
+	bool PermitWeird(const char* name, uint64 threshold, uint64 rate,
+	                 double duration);
+
 protected:
 
 	Connection()	{ persistent = 0; }
@@ -339,6 +345,14 @@ protected:
 	analyzer::pia::PIA* primary_PIA;
 
 	Bro::UID uid;	// Globally unique connection ID.
+
+	struct WeirdState {
+		WeirdState() { count = 0; sampling_start_time = 0; }
+		uint64 count = 0;
+		double sampling_start_time = 0;
+	};
+
+	std::unordered_map<std::string, WeirdState> weird_state;
 };
 
 class ConnectionTimer : public Timer {

src/Func.cc

@@ -722,6 +722,7 @@ void init_builtin_funcs()
 	FileAnalysisStats = internal_type("FileAnalysisStats")->AsRecordType();
 	ThreadStats = internal_type("ThreadStats")->AsRecordType();
 	BrokerStats = internal_type("BrokerStats")->AsRecordType();
+	ReporterStats = internal_type("ReporterStats")->AsRecordType();
 
 	var_sizes = internal_type("var_sizes")->AsTableType();
 

src/Reporter.cc

@@ -10,6 +10,7 @@
 #include "NetVar.h"
 #include "Net.h"
 #include "Conn.h"
+#include "Timer.h"
 #include "plugin/Plugin.h"
 #include "plugin/Manager.h"
 
@@ -36,6 +37,11 @@ Reporter::Reporter()
 	warnings_to_stderr = true;
 	errors_to_stderr = true;
 
+	weird_count = 0;
+	weird_sampling_rate = 0;
+	weird_sampling_duration = 0;
+	weird_sampling_threshold = 0;
+
 	openlog("bro", 0, LOG_LOCAL5);
 	}
 
@@ -49,6 +55,24 @@ void Reporter::InitOptions()
 	info_to_stderr = internal_const_val("Reporter::info_to_stderr")->AsBool();
 	warnings_to_stderr = internal_const_val("Reporter::warnings_to_stderr")->AsBool();
 	errors_to_stderr = internal_const_val("Reporter::errors_to_stderr")->AsBool();
+	weird_sampling_rate = internal_const_val("Weird::sampling_rate")->AsCount();
+	weird_sampling_threshold = internal_const_val("Weird::sampling_threshold")->AsCount();
+	weird_sampling_duration = internal_const_val("Weird::sampling_duration")->AsInterval();
+	auto wl_val = internal_const_val("Weird::sampling_whitelist")->AsTableVal();
+	auto wl_table = wl_val->AsTable();
+
+	HashKey* k;
+	IterCookie* c = wl_table->InitForIteration();
+	TableEntryVal* v;
+
+	while ( (v = wl_table->NextEntry(k, c)) )
+		{
+		auto index = wl_val->RecoverIndex(k);
+		string key = index->Index(0)->AsString()->CheckString();
+		weird_sampling_whitelist.emplace(move(key));
+		Unref(index);
+		delete k;
+		}
 	}
 
 void Reporter::Info(const char* fmt, ...)
@@ -221,23 +245,121 @@ void Reporter::WeirdFlowHelper(const IPAddr& orig, const IPAddr& resp, const cha
 	delete vl;
 	}
 
+void Reporter::UpdateWeirdStats(const char* name)
+	{
+	++weird_count;
+	++weird_count_by_type[name];
+	}
+
+class NetWeirdTimer : public Timer {
+public:
+	NetWeirdTimer(double t, const char* name, double timeout)
+	: Timer(t + timeout, TIMER_NET_WEIRD_EXPIRE), weird_name(name)
+		{}
+
+	void Dispatch(double t, int is_expire) override
+		{ reporter->ResetNetWeird(weird_name); }
+
+	std::string weird_name;
+};
+
+class FlowWeirdTimer : public Timer {
+public:
+	using IPPair = std::pair<IPAddr, IPAddr>;
+
+	FlowWeirdTimer(double t, IPPair p, double timeout)
+	: Timer(t + timeout, TIMER_FLOW_WEIRD_EXPIRE), endpoints(p)
+		{}
+
+	void Dispatch(double t, int is_expire) override
+		{ reporter->ResetFlowWeird(endpoints.first, endpoints.second); }
+
+	IPPair endpoints;
+};
+
+void Reporter::ResetNetWeird(const std::string& name)
+	{
+	net_weird_state.erase(name);
+	}
+
+void Reporter::ResetFlowWeird(const IPAddr& orig, const IPAddr& resp)
+	{
+	flow_weird_state.erase(std::make_pair(orig, resp));
+	}
+
+bool Reporter::PermitNetWeird(const char* name)
+	{
+	auto& count = net_weird_state[name];
+	++count;
+
+	if ( count == 1 )
+		timer_mgr->Add(new NetWeirdTimer(network_time, name,
+		                                 weird_sampling_duration));
+
+	if ( count < weird_sampling_threshold )
+		return true;
+
+	auto num_above_threshold = count - weird_sampling_threshold;
+	return num_above_threshold % weird_sampling_rate == 0;
+	}
+
+bool Reporter::PermitFlowWeird(const char* name,
+                               const IPAddr& orig, const IPAddr& resp)
+	{
+	auto endpoints = std::make_pair(orig, resp);
+	auto& map = flow_weird_state[endpoints];
+
+	if ( map.empty() )
+		timer_mgr->Add(new FlowWeirdTimer(network_time, endpoints,
+		                                  weird_sampling_duration));
+
+	auto& count = map[name];
+	++count;
+
+	if ( count < weird_sampling_threshold )
+		return true;
+
+	auto num_above_threshold = count - weird_sampling_threshold;
+	return num_above_threshold % weird_sampling_rate == 0;
+	}
+
 void Reporter::Weird(const char* name)
 	{
+	UpdateWeirdStats(name);
+
+	if ( ! WeirdOnSamplingWhiteList(name) )
+		{
+		if ( ! PermitNetWeird(name) )
+			return;
+		}
+
 	WeirdHelper(net_weird, 0, 0, "%s", name);
 	}
 
 void Reporter::Weird(Connection* conn, const char* name, const char* addl)
 	{
-	WeirdHelper(conn_weird, conn->BuildConnVal(), addl, "%s", name);
+	UpdateWeirdStats(name);
+
+	if ( ! WeirdOnSamplingWhiteList(name) )
+		{
+		if ( ! conn->PermitWeird(name, weird_sampling_threshold,
+		                         weird_sampling_rate, weird_sampling_duration) )
+			return;
 		}
 
-void Reporter::Weird(Val* conn_val, const char* name, const char* addl)
+	WeirdHelper(conn_weird, conn->BuildConnVal(), addl, "%s", name);
-	{
-	WeirdHelper(conn_weird, conn_val, addl, "%s", name);
 	}
 
 void Reporter::Weird(const IPAddr& orig, const IPAddr& resp, const char* name)
 	{
+	UpdateWeirdStats(name);
+
+	if ( ! WeirdOnSamplingWhiteList(name) )
+		{
+		if ( ! PermitFlowWeird(name, orig, resp) )
+			 return;
+		}
+
 	WeirdFlowHelper(orig, resp, "%s", name);
 	}
 

src/Reporter.h

@@ -7,6 +7,10 @@
 
 #include <list>
 #include <utility>
+#include <string>
+#include <map>
+#include <unordered_set>
+#include <unordered_map>
 
 #include "util.h"
 #include "EventHandler.h"
@@ -36,6 +40,11 @@ protected:
 
 class Reporter {
 public:
+	using IPPair = std::pair<IPAddr, IPAddr>;
+	using WeirdCountMap = std::unordered_map<std::string, uint64>;
+	using WeirdFlowMap = std::map<IPPair, WeirdCountMap>;
+	using WeirdSet = std::unordered_set<std::string>;
+
 	Reporter();
 	~Reporter();
 
@@ -76,7 +85,6 @@ public:
 	// that may lead to incorrectly processing a connnection.
 	void Weird(const char* name);	// Raises net_weird().
 	void Weird(Connection* conn, const char* name, const char* addl = "");	// Raises conn_weird().
-	void Weird(Val* conn_val, const char* name, const char* addl = "");	// Raises conn_weird().
 	void Weird(const IPAddr& orig, const IPAddr& resp, const char* name);	// Raises flow_weird().
 
 	// Syslog a message. This methods does nothing if we're running
@@ -120,6 +128,30 @@ public:
 	// Signals that we're done processing an error handler event.
 	void EndErrorHandler()	{ --in_error_handler; }
 
+	/**
+	 * Reset/cleanup state tracking for a "net" weird.
+	 */
+	void ResetNetWeird(const std::string& name);
+
+	/**
+	 * Reset/cleanup state tracking for a "flow" weird.
+	 */
+	void ResetFlowWeird(const IPAddr& orig, const IPAddr& resp);
+
+	/**
+	 * Return the total number of weirds generated (counts weirds before
+	 * any rate-limiting occurs).
+	 */
+	uint64 GetWeirdCount() const
+		{ return weird_count; }
+
+	/**
+	 * Return number of weirds generated per weird type/name (counts weirds
+	 * before any rate-limiting occurs).
+	 */
+	const WeirdCountMap& GetWeirdsByType() const
+		{ return weird_count_by_type; }
+
 private:
 	void DoLog(const char* prefix, EventHandlerPtr event, FILE* out,
 		   Connection* conn, val_list* addl, bool location, bool time,
@@ -129,6 +161,11 @@ private:
 	// contain format specifiers
 	void WeirdHelper(EventHandlerPtr event, Val* conn_val, const char* addl, const char* fmt_name, ...) __attribute__((format(printf, 5, 6)));;
 	void WeirdFlowHelper(const IPAddr& orig, const IPAddr& resp, const char* fmt_name, ...) __attribute__((format(printf, 4, 5)));;
+	void UpdateWeirdStats(const char* name);
+	inline bool WeirdOnSamplingWhiteList(const char* name)
+		{ return weird_sampling_whitelist.find(name) != weird_sampling_whitelist.end(); }
+	bool PermitNetWeird(const char* name);
+	bool PermitFlowWeird(const char* name, const IPAddr& o, const IPAddr& r);
 
 	int errors;
 	bool via_events;
@@ -138,6 +175,18 @@ private:
 	bool errors_to_stderr;
 
 	std::list<std::pair<const Location*, const Location*> > locations;
+
+	uint64 weird_count;
+	WeirdCountMap weird_count_by_type;
+
+	WeirdCountMap net_weird_state;
+	WeirdFlowMap flow_weird_state;
+
+	WeirdSet weird_sampling_whitelist;
+	uint64 weird_sampling_threshold;
+	uint64 weird_sampling_rate;
+	double weird_sampling_duration;
+
 };
 
 extern Reporter* reporter;

src/Timer.cc

@@ -18,12 +18,14 @@ const char* TimerNames[] = {
 	"ConnectionStatusUpdateTimer",
 	"DNSExpireTimer",
 	"FileAnalysisInactivityTimer",
+	"FlowWeirdTimer",
 	"FragTimer",
 	"IncrementalSendTimer",
 	"IncrementalWriteTimer",
 	"InterconnTimer",
 	"IPTunnelInactivityTimer",
 	"NetbiosExpireTimer",
+	"NetWeirdTimer",
 	"NetworkTimer",
 	"NTPExpireTimer",
 	"ProfileTimer",

src/Timer.h

@@ -23,12 +23,14 @@ enum TimerType {
 	TIMER_CONN_STATUS_UPDATE,
 	TIMER_DNS_EXPIRE,
 	TIMER_FILE_ANALYSIS_INACTIVITY,
+	TIMER_FLOW_WEIRD_EXPIRE,
 	TIMER_FRAG,
 	TIMER_INCREMENTAL_SEND,
 	TIMER_INCREMENTAL_WRITE,
 	TIMER_INTERCONN,
 	TIMER_IP_TUNNEL_INACTIVITY,
 	TIMER_NB_EXPIRE,
+	TIMER_NET_WEIRD_EXPIRE,
 	TIMER_NETWORK,
 	TIMER_NTP_EXPIRE,
 	TIMER_PROFILE,

src/analyzer/protocol/finger/Finger.cc

@@ -41,6 +41,9 @@ void Finger_Analyzer::DeliverStream(int length, const u_char* data, bool is_orig
 	const char* line = (const char*) data;
 	const char* end_of_line = line + length;
 
+	if ( length == 0 )
+		return;
+
 	if ( is_orig )
 		{
 

src/analyzer/protocol/ftp/FTP.cc

@@ -69,6 +69,10 @@ void FTP_Analyzer::DeliverStream(int length, const u_char* data, bool orig)
 	const char* line = (const char*) data;
 	const char* end_of_line = line + length;
 
+	if ( length == 0 )
+		// Could emit "ftp empty request/reply" weird, but maybe not worth it.
+		return;
+
 	val_list* vl = new val_list;
 	vl->append(BuildConnVal());
 

src/analyzer/protocol/ident/Ident.cc

@@ -56,6 +56,9 @@ void Ident_Analyzer::DeliverStream(int length, const u_char* data, bool is_orig)
 	if ( TCP() )
 		s = is_orig ? TCP()->Orig() : TCP()->Resp();
 
+	if ( length == 0 )
+		return;
+
 	if ( is_orig )
 		{
 		if ( ! ident_request )

src/analyzer/protocol/pop3/POP3.cc

@@ -132,6 +132,9 @@ static string trim_whitespace(const char* in)
 
 void POP3_Analyzer::ProcessRequest(int length, const char* line)
 	{
+	if ( length == 0 )
+		return;
+
 	if ( waitingForAuthentication )
 		{
 		++authLines;

src/analyzer/protocol/smtp/SMTP.cc

@@ -269,6 +269,8 @@ void SMTP_Analyzer::ProcessLine(int length, const char* line, bool orig)
 			if ( smtp_request )
 				{
 				int data_len = end_of_line - line;
+
+				if ( cmd_len > 0 || data_len > 0 )
 					RequestEvent(cmd_len, cmd, data_len, line);
 				}
 
@@ -379,8 +381,18 @@ void SMTP_Analyzer::NewCmd(const int cmd_code)
 		if ( first_cmd < 0 )
 			first_cmd = cmd_code;
 		else
+			{
+			auto constexpr max_pending_cmd_q_size = 1000;
+
+			if ( pending_cmd_q.size() == max_pending_cmd_q_size )
+				{
+				Weird("smtp_excessive_pending_cmds");
+				pending_cmd_q.clear();
+				}
+
 			pending_cmd_q.push_back(cmd_code);
 			}
+		}
 	else
 		first_cmd = cmd_code;
 	}
@@ -805,12 +817,22 @@ void SMTP_Analyzer::UpdateState(const int cmd_code, const int reply_code, bool o
 #endif
 	}
 
+static bool istrequal(const char* s, const char* cmd, int s_len)
+	{
+	auto cmd_len = strlen(cmd);
+
+	if ( cmd_len != s_len )
+		return false;
+
+	return strncasecmp(s, cmd, s_len) == 0;
+	}
+
 void SMTP_Analyzer::ProcessExtension(int ext_len, const char* ext)
 	{
 	if ( ! ext )
 		return;
 
-	if ( ! strncasecmp(ext, "PIPELINING", ext_len) )
+	if ( istrequal(ext, "PIPELINING", ext_len) )
 		pipelining = 1;
 	}
 
@@ -820,11 +842,11 @@ int SMTP_Analyzer::ParseCmd(int cmd_len, const char* cmd)
 		return -1;
 
 	// special case because we cannot define our usual macros with "-"
-	if ( strncmp(cmd, "X-ANONYMOUSTLS", cmd_len) == 0 )
+	if ( istrequal(cmd, "X-ANONYMOUSTLS", cmd_len) )
 		return SMTP_CMD_X_ANONYMOUSTLS;
 
 	for ( int code = SMTP_CMD_EHLO; code < SMTP_CMD_LAST; ++code )
-		if ( ! strncasecmp(cmd, smtp_cmd_word[code - SMTP_CMD_EHLO], cmd_len) )
+		if ( istrequal(cmd, smtp_cmd_word[code - SMTP_CMD_EHLO], cmd_len) )
 			return code;
 
 	return -1;

src/reporter.bif

@@ -71,3 +71,44 @@ function Reporter::fatal%(msg: string%): bool
 	reporter->PopLocation();
 	return new Val(1, TYPE_BOOL);
 	%}
+
+## Generates a "net" weird.
+##
+## name: the name of the weird.
+##
+## Returns: Always true.
+function Reporter::net_weird%(name: string%): bool
+	%{
+	reporter->Weird(name->CheckString());
+	return new Val(1, TYPE_BOOL);
+	%}
+
+## Generates a "flow" weird.
+##
+## name: the name of the weird.
+##
+## orig: the originator host associated with the weird.
+##
+## resp: the responder host associated with the weird.
+##
+## Returns: Always true.
+function Reporter::flow_weird%(name: string, orig: addr, resp: addr%): bool
+	%{
+	reporter->Weird(orig->AsAddr(), resp->AsAddr(), name->CheckString());
+	return new Val(1, TYPE_BOOL);
+	%}
+
+## Generates a "conn" weird.
+##
+## name: the name of the weird.
+##
+## c: the connection associated with the weird.
+##
+## addl: additional information to accompany the weird.
+##
+## Returns: Always true.
+function Reporter::conn_weird%(name: string, c: connection, addl: string &default=""%): bool
+	%{
+	reporter->Weird(c, name->CheckString(), addl->CheckString());
+	return new Val(1, TYPE_BOOL);
+	%}

src/stats.bif

@@ -16,6 +16,7 @@ RecordType* ThreadStats;
 RecordType* TimerStats;
 RecordType* FileAnalysisStats;
 RecordType* BrokerStats;
+RecordType* ReporterStats;
 %%}
 
 ## Returns packet capture statistics. Statistics include the number of
@@ -35,6 +36,7 @@ RecordType* BrokerStats;
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_net_stats%(%): NetStats
 	%{
 	uint64 recv = 0;
@@ -83,6 +85,7 @@ function get_net_stats%(%): NetStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_conn_stats%(%): ConnStats
 	%{
 	RecordVal* r = new RecordVal(ConnStats);
@@ -133,6 +136,7 @@ function get_conn_stats%(%): ConnStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_proc_stats%(%): ProcStats
 	%{
 	struct rusage ru;
@@ -189,6 +193,7 @@ function get_proc_stats%(%): ProcStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_event_stats%(%): EventStats
 	%{
 	RecordVal* r = new RecordVal(EventStats);
@@ -215,6 +220,7 @@ function get_event_stats%(%): EventStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_reassembler_stats%(%): ReassemblerStats
 	%{
 	RecordVal* r = new RecordVal(ReassemblerStats);
@@ -243,6 +249,7 @@ function get_reassembler_stats%(%): ReassemblerStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_dns_stats%(%): DNSStats
 	%{
 	RecordVal* r = new RecordVal(DNSStats);
@@ -276,6 +283,7 @@ function get_dns_stats%(%): DNSStats
 ##              get_reassembler_stats
 ##              get_thread_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_timer_stats%(%): TimerStats
 	%{
 	RecordVal* r = new RecordVal(TimerStats);
@@ -303,6 +311,7 @@ function get_timer_stats%(%): TimerStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_file_analysis_stats%(%): FileAnalysisStats
 	%{
 	RecordVal* r = new RecordVal(FileAnalysisStats);
@@ -330,6 +339,7 @@ function get_file_analysis_stats%(%): FileAnalysisStats
 ##              get_reassembler_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_thread_stats%(%): ThreadStats
 	%{
 	RecordVal* r = new RecordVal(ThreadStats);
@@ -355,6 +365,7 @@ function get_thread_stats%(%): ThreadStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_gap_stats%(%): GapStats
 	%{
 	RecordVal* r = new RecordVal(GapStats);
@@ -386,6 +397,7 @@ function get_gap_stats%(%): GapStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_matcher_stats%(%): MatcherStats
 	%{
 	RecordVal* r = new RecordVal(MatcherStats);
@@ -423,6 +435,7 @@ function get_matcher_stats%(%): MatcherStats
 ##              get_thread_stats
 ##              get_timer_stats
 ##              get_broker_stats
+##              get_reporter_stats
 function get_broker_stats%(%): BrokerStats
 	%{
 	RecordVal* r = new RecordVal(BrokerStats);
@@ -441,3 +454,39 @@ function get_broker_stats%(%): BrokerStats
 
 	return r;
 	%}
+
+## Returns statistics about reporter messages and weirds.
+##
+## Returns: A record with reporter statistics.
+##
+## .. bro:see:: get_conn_stats
+##              get_dns_stats
+##              get_event_stats
+##              get_file_analysis_stats
+##              get_gap_stats
+##              get_matcher_stats
+##              get_net_stats
+##              get_proc_stats
+##              get_reassembler_stats
+##              get_thread_stats
+##              get_timer_stats
+##              get_broker_stats
+function get_reporter_stats%(%): ReporterStats
+	%{
+	RecordVal* r = new RecordVal(ReporterStats);
+	int n = 0;
+
+	TableVal* weirds_by_type = new TableVal(internal_type("table_string_of_count")->AsTableType());
+
+	for ( auto& kv : reporter->GetWeirdsByType() )
+		{
+		Val* weird = new StringVal(kv.first);
+		weirds_by_type->Assign(weird, new Val(kv.second, TYPE_COUNT));
+		Unref(weird);
+		}
+
+	r->Assign(n++, new Val(reporter->GetWeirdCount(), TYPE_COUNT));
+	r->Assign(n++, weirds_by_type);
+
+	return r;
+	%}

testing/btest/Baseline/core.reporter-weird-sampling/output

@@ -0,0 +1,351 @@
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird
+net_weird, my_net_weird
+flow_weird, my_flow_weird
+conn_weird, my_conn_weird
+net_weird, whitelisted_net_weird
+flow_weird, whitelisted_flow_weird
+conn_weird, whitelisted_conn_weird

testing/btest/Baseline/coverage.find-bro-logs/out

@@ -55,4 +55,5 @@ traceroute
 tunnel
 unified2
 weird
+weird_stats
 x509

testing/btest/Baseline/scripts.policy.misc.weird-stats-cluster/manager-1.weird_stats.log

@@ -0,0 +1,13 @@
+#separator \x09
+#set_separator	,
+#empty_field	(empty)
+#unset_field	-
+#path	weird_stats
+#open	2018-07-26-23-11-27
+#fields	ts	name	num_seen
+#types	time	string	count
+1532646687.827249	weird3	1
+1532646687.827249	weird2	1000
+1532646687.827249	weird1	2000
+1532646692.877464	weird1	2
+#close	2018-07-26-23-11-34

testing/btest/Baseline/scripts.policy.misc.weird-stats/bro.weird_stats.log

@@ -0,0 +1,12 @@
+#separator \x09
+#set_separator	,
+#empty_field	(empty)
+#unset_field	-
+#path	weird_stats
+#open	2018-07-27-00-20-35
+#fields	ts	name	num_seen
+#types	time	string	count
+1532650834.978616	my_weird	1000
+1532650840.011592	my_weird	2000
+1532650845.043367	my_weird	10
+#close	2018-07-27-00-20-47

testing/btest/core/reporter-weird-sampling.bro

@@ -0,0 +1,55 @@
+# @TEST-EXEC: bro -b -r $TRACES/http/bro.org.pcap %INPUT >output
+# @TEST-EXEC: btest-diff output
+
+redef Weird::sampling_duration = 5sec;
+redef Weird::sampling_threshold = 10;
+redef Weird::sampling_rate = 10;
+redef Weird::sampling_whitelist = set("whitelisted_net_weird",
+                                      "whitelisted_flow_weird",
+                                      "whitelisted_conn_weird");
+
+event conn_weird(name: string, c: connection, addl: string)
+	{
+	print "conn_weird", name;
+	}
+
+event flow_weird(name: string, src: addr, dst: addr)
+	{
+	print "flow_weird", name;
+	}
+
+event net_weird(name: string)
+	{
+	print "net_weird", name;
+	}
+
+event gen_weirds(c: connection)
+	{
+	local num = 30;
+
+	while ( num != 0 )
+		{
+		Reporter::net_weird("my_net_weird");
+		Reporter::flow_weird("my_flow_weird", c$id$orig_h, c$id$resp_h);
+		Reporter::conn_weird("my_conn_weird", c);
+
+		Reporter::net_weird("whitelisted_net_weird");
+		Reporter::flow_weird("whitelisted_flow_weird", c$id$orig_h, c$id$resp_h);
+		Reporter::conn_weird("whitelisted_conn_weird", c);
+		--num;
+		}
+	}
+
+global did_one_connection = F;
+
+event new_connection(c: connection)
+	{
+	if ( did_one_connection )
+		return;
+
+	did_one_connection = T;
+	event gen_weirds(c);             # should permit 10 + 2 of each "my" weird
+	schedule 2sec { gen_weirds(c) }; # should permit 3 of each "my" weird
+	schedule 7sec { gen_weirds(c) }; # should permit 10 + 2 of each "my" weird
+	# Total of 27 "my" weirds of each type and 90 of each "whitelisted" type
+	}

testing/btest/scripts/policy/misc/weird-stats-cluster.bro

@@ -0,0 +1,93 @@
+# @TEST-SERIALIZE: comm
+#
+# @TEST-EXEC: btest-bg-run manager-1 BROPATH=$BROPATH:.. CLUSTER_NODE=manager-1 bro %INPUT
+# @TEST-EXEC: btest-bg-run worker-1  BROPATH=$BROPATH:.. CLUSTER_NODE=worker-1 bro %INPUT
+# @TEST-EXEC: btest-bg-run worker-2  BROPATH=$BROPATH:.. CLUSTER_NODE=worker-2 bro %INPUT
+# @TEST-EXEC: btest-bg-wait 20
+
+# @TEST-EXEC: btest-diff manager-1/weird_stats.log
+
+@TEST-START-FILE cluster-layout.bro
+redef Cluster::nodes = {
+	["manager-1"] = [$node_type=Cluster::MANAGER, $ip=127.0.0.1, $p=37757/tcp],
+	["worker-1"]  = [$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
+
+@load misc/weird-stats
+
+redef Cluster::retry_interval = 1sec;
+redef Broker::default_listen_retry = 1sec;
+redef Broker::default_connect_retry = 1sec;
+
+redef Log::enable_local_logging = T;
+redef Log::default_rotation_interval = 0secs;
+redef WeirdStats::weird_stat_interval = 5secs;
+
+event terminate_me()
+	{
+	terminate();
+	}
+
+event Broker::peer_lost(endpoint: Broker::EndpointInfo, msg: string)
+	{
+	terminate();
+	}
+
+event ready_again()
+	{
+	Reporter::net_weird("weird1");
+
+	if ( Cluster::node == "worker-2" )
+		{
+		schedule 5secs { terminate_me() };
+		}
+	}
+
+event ready_for_data()
+	{
+	local n = 0;
+
+	if ( Cluster::node == "worker-1" )
+		{
+		while ( n < 1000 )
+			{
+			Reporter::net_weird("weird1");
+			++n;
+			}
+
+		Reporter::net_weird("weird3");
+		}
+	else if ( Cluster::node == "worker-2" )
+		{
+		while ( n < 1000 )
+			{
+			Reporter::net_weird("weird1");
+			Reporter::net_weird("weird2");
+			++n;
+			}
+		}
+
+	schedule 5secs { ready_again() };
+	}
+
+
+@if ( Cluster::local_node_type() == Cluster::MANAGER )
+
+event bro_init()
+	{
+	Broker::auto_publish(Cluster::worker_topic, ready_for_data);
+	}
+
+global peer_count = 0;
+
+event Broker::peer_added(endpoint: Broker::EndpointInfo, msg: string)
+	{
+	++peer_count;
+
+	if ( peer_count == 2 )
+		event ready_for_data();
+	}
+
+@endif

testing/btest/scripts/policy/misc/weird-stats.bro

@@ -0,0 +1,32 @@
+# @TEST-EXEC: btest-bg-run bro bro %INPUT
+# @TEST-EXEC: btest-bg-wait 20
+# @TEST-EXEC: btest-diff bro/weird_stats.log
+
+@load misc/weird-stats.bro
+
+redef exit_only_after_terminate = T;
+redef WeirdStats::weird_stat_interval = 5sec;
+
+event die()
+	{
+	terminate();
+	}
+
+event gen_weirds(n: count, done: bool &default = F)
+	{
+	while ( n != 0 )
+		{
+		Reporter::net_weird("my_weird");
+		--n;
+		}
+
+	if ( done )
+		schedule 5sec { die() };
+	}
+
+event bro_init()
+	{
+	event gen_weirds(1000);
+	schedule 7.5sec { gen_weirds(2000) } ;
+	schedule 12.5sec { gen_weirds(10, T) } ;
+	}

testing/scripts/diff-remove-timestamps

@@ -9,5 +9,5 @@ else
    sed="sed -E"
 fi
 
-$sed 's/(0\.000000)|([0-9]{10}\.[0-9]{2,8})/XXXXXXXXXX.XXXXXX/g' | \
+$sed 's/(0\.000000)|([0-9]{9,10}\.[0-9]{2,8})/XXXXXXXXXX.XXXXXX/g' | \
 $sed 's/^ *#(open|close).(19|20)..-..-..-..-..-..$/#\1 XXXX-XX-XX-XX-XX-XX/g'