* origin/topic/johanna/conn-threshold:
Wrap threshold stuff up - fix two small bugs and update baselines.
update GridFTP analyzer to use connection thresholding instead of polling
Add high level api for thresholding that holds lists of thresholds and raises an event for each threshold exactly once.
Allow setting packet and byte thresholds for connections.
BIT-1377 #merged
one can now add an option "offset" to the config map. Positive offsets
are interpreted to be from the beginning of the file, negative from the
end of the file (-1 is end of file).
Only works for raw reader in streaming or manual mode. Does not work
with executables.
Addresses BIT-985
This extends the ConnSize analyzer to be able to raise events when each
direction of a connection crosses a certain amount of bytes or packets.
Thresholds are set using
set_conn_bytes_threshold(c$id, [num-bytes], [direction]);
and
set_conn_packets_threshold(c$id, [num-packets], [direction]);
respectively.
They raise the event
event conn_bytes_threshold_crossed(c: connection, threshold: count, is_orig: bool)
and
event conn_packets_threshold_crossed(c: connection, threshold: count, is_orig: bool)
respectively.
Current thresholds can be examined using
get_conn_bytes_threshold and get_conn_packets_threshold
Currently only one threshold can be set per connection.
This also fixes a bug where child packet analyzers of the TCP analyzer
where not found using FindChild.
Under remote communication overload conditions, the child->parent
chunked IO may start rejecting chunks if over the hard cap. Some
messages are made of two chunks, accepting the first part, but rejecting
the second can put the parent in a bad state and the next two chunks it
reads are likely to cause the error.
This patch just removes the rejecting functionality completely and so
now relies solely on shutting down remote peer connections to help
alleviate temporary overload conditions. The
"chunked_io_buffer_soft_cap" script variable can now tune when this
shutting down starts happening and the default setting is now double
what it used to be. For constant overload conditions, communication.log
should keep stating "queue to parent filling up; shutting down heaviest
connection".
An alternative to completely removing the hard cap rejection code could
be ensuring that messages that involve a pair of chunks can never have
the second chunk be rejected when attempting to write it.
Addresses BIT-1376
Note: loading external-ca-list.bro in the external tests increases
execution times by 1-2%; if I remove that @load, things get back to
normal so doesn't seem to indicate a problem.
* origin/topic/johanna/ca-list:
Update mozilla CA list.
BIT-1375 #merged
With this patch the model is:
- "print" cleans the data so that non-printable characters get
escaped. This is not necessarily reversible.
- to print in a reversible way, one can go through
escape_string(); this escapes backslashes as well to make the
decoding non-ambigious.
- Logging always escapes similar to escape_string(), making it
reversible.
Compared to master, we also change the escaping as follows:
- We now only escape with "\xXX", no more "^X" or "\0". Exception:
backslashes.
- We escape backlashes as "\\".
- There's no "alternative" output style anymore, i.e., fmt() '%A'
qualifier is gone.
Baselines in testing/btest are updated, external tests not yet.
Addresses BIT-1333.
Turns out that in error situations, the final finish message might not
reach the writer anymore, as communication between the threads will be
shut down. Instead of aborting, we now just clean up in that case and
proceed. This isn't changing any other behaviour. The original error
check was in place mostly for helping debug the data flow between the
threads anyways.
Addresses BIT-1331.
This also patches a few tests to contain certificates that were removed.
Furthermore, we include the old CA file with the external tests and load
it automatically. Those traces are kind of old now, more and more of the
CAs in them are no longer valid and it does not really make sense to
update them on each change...
There were two problems actually: the iomanager wasn't properly
deleting sourcesl; and in some situations, the remote serialize wasn't
registered with it to begin with.
Addresses BIT-1306 and probably also BIT-1356.
Removed "file_mime_type" and "file_mime_types" event, replacing them
with a new event called "file_metadata_inferred". It has a record
argument of type "inferred_file_metadata", which contains the mime type
information that the earlier events used to supply. The idea here is
that future extensions to the record with new metadata will be less
likely to break user code than the alternatives (adding new events or
new event parameters).
Addresses BIT-1368.
When files had gaps prior to the bof_buffer completely filling, the
file gap handling code was never sniffing and passing along as much
data as possible so file type identification wasn't working correctly.
The HTTP analyzer was propogating Gaps to the files framework even
in the case of a packet drop occurring immediately after the headers
are completed in an HTTP response when the response content length
was declared to be zero (no file started, so no loss).
Includes passing test.
