util::safe_write() calls abort() in case of EAGAIN errors. This is
easily observed when starting clusters with 32 workers or more.
Add a custom write_message() function handling EAGAIN by retrying
after a small sleep. It's not clear a more complicated poll() would be
much better: The pipe might be ready for writing, but then our message
might not actually fit in, resulting in another EAGAIN error. And even
poll() would introduce blocking/sleeping code.
Take some precautions against the stem and the supervisor dead-locking
when both pipes are full by draining the other end on EAGAIN errors.
Closes#3043
The ssl_history field may grow unbounded (e.g., ssl_alert event). Prevent this
by capping using a configurable limit (default 100) and raise a weird once reached.
Previously, seq was computed as the result of |pending_commands|+1. This
opened the possibility to override queued commands, as well as logging
the same pending ftp reply multiple times.
For example, when commands 1, 2, 3 are pending, command 1 may be dequeued,
but the incoming command then receives seq 3 and overrides the already
pending command 3. The second scenario happens when ftp_reply() selected
command 3 as pending for logging, but is then followed by many ftp_request()
events. This resulted in command 3's response being logged for every
following ftp_request() over and over again.
Avoid both scenarios by tracking the command sequence as an absolute counter.
Unsure what it's used for today and also results in the situation that on
some platforms we generate a reporter.log in bare mode, while on others
where spicy is disabled, we do not.
If we want base/frameworks/version loaded by default, should put it into
init-bare.zeek and possibly remove the loading of the reporter framework
from it - Reporter::error() would still work and be visible on stderr,
just not create a reporter.log.
The first pcap only contained packets from the originator, not the responder.
What stands out here is that the Linux kernel doesn't seem to use a symmetric
flow hash for the tunneled connection, resulting in a total of four tunnel
connections for the two inner connections. Sigh.
* origin/topic/awelzel/2326-import-quic:
ci/btest: Remove spicy-quic helper, disable Spicy on CentOS 7
btest/core/ppp: Run test in bare mode
btest/quic: Update other tests
testing/quic: Fixups and simplification after Zeek integration
quic: Integrate as default analyzer
quic: Include Copyright lines to the analyzer's source code contributed by Fox-IT
quic: Squashed follow-ups: quic.log, tests, various fixes, performance
quic: Initial implementation
Without this, Zeekygen won't generate documentation about exported
enum types as it can not resolve the identifier. Also, only register a
type as item with the Spicy plugin if there's no _module_info currently
active.
* origin/topic/bbannier/issue-3234:
Introduce dedicated `LDAP::Info`
Remove redundant storing of protocol in LDAP logs
Use LDAP `RemovalHook` instead of implementing `connection_state_remove`
Tidy up LDAP code by using local references
Pluralize container names in LDAP types
Move LDAP script constants to their own file
Name `LDAP::Message` and `LDAP::Search` `*Info`
Make ports for LDAP analyzers fully configurable
Require have-spicy for tests which log spicy-ldap information
Fix LDAP analyzer setup for when Spicy analyzers are disabled
Bump zeek-testing-private
Integrate spicy-ldap test suite
Move spicy-ldap into Zeek protocol analyzer tree
Explicitly use all of spicy-ldap's modules
Explicitly list `asn1.spicy` as spicy-ldap source
Remove uses of `zeek` module in spicy-ldap
Fix typos in spicy-ldap
Remove project configuration files in spicy-ldap
Integrate spicy-ldap into build
Import zeek/spicy-ldap@57b5eff988
This moves the ports the LDAP analyzers should be triggered on from the
EVT file to the Zeek module. This gives users full control over which
ports the analyzers are registered for while previously they could only
register them for additional ports (there is no Zeek script equivalent
of `Manager::UnregisterAnalyzerForPort`).
The analyzers could still be triggered via DPD, but this is intentional.
To fully disable analyzers users can use e.g.,
```zeek
event zeek_init()
{
Analyzer::disable_analyzer(Analyzer::ANALYZER_LDAP_TCP);
}
```
On Linux with a default ext4 or tmpfs filesystem, the default buffer size for
reading a pcap is chosen as 4k (strace/gdb validated). When reading large pcaps
containing raw data transfers, the syscall overhead for read becomes visible
in profiles. Support configurability of the buffer size and default to 128kb.
When processing a ~830M PCAP (16 UDP connections, each transferring ~50MB) in
bare mode, this change improves runtime from 1.39 sec to 1.29 sec. Increasing
the buffer further didn't provide a noticeable boost.
Motivation is basically the same as in 88bb527026.
For plugin.hooks, one example is that adding a new option in the default script
changes the baseline due registration of change handlers. Also, the connection
record is printed in various places, resulting in churn when the default
scripts change.
Justin pointed out that the misc/dump-events test shows added fields to
the connection record. Add a new test that prints the connection record
recursively in bare and default mode to cover that use-case
specifically.
The dump-events baseline changes are pure noise and have spurred confusion
for internal and external contributors. For example, adding new
analyzers have perturbed orderings of sets holding analyzer tags.
Running in non-bare mode, the baselines change almost whenever any of the
record types attached to connections change in the default scripts. This
causes continuous and seemingly little useful updates to the baselines.
This change switches the test to run in bare mode and explicitly loads
just base/protocols/conn and base/protocols/smtp. The primary intention
of the test should be testing the functionality of the misc/dump-events
script, not the raised events of all loaded default scripts (for that the
used PCAP is too narrow).
Protocol specific scripts that do want to leverage misc/dump-events for
baseline creation of their or their analyzer's events can add additional
specific tests with suitable PCAP files.
A bit larger follow-up to what Tim pointed out: Function prototype descriptions
previously used semicolons to separate parameters.
Switch to use commas when a RecordType is used as function parameter.
Use existing "func_args" naming for consistency.
