Problems could happen if the queue worker needed to be cancelled
and this cancellation happened inside queue-code (including
wtp, wti). We have now solved this by disabling cancellation while
in this code and only enabling it when working inside the user consumer.
This exactly matches the use case for which cancellation may be needed.
these occured in very unusual scenarios where we had a DA-queue running
in parallel and very lengthy actions. Then, in some situations, the
shutdown could hang. The code needs some addition lab time, but
is believed to be much better than any previous version.
... but this brings a lot of problems with it. The issue is that
we still have a sequential store and we do not know how we could
delete the one entry right in the middle of processing. I keep this
branch if we intend to move on with it - but for now I look into a
different solution...
support for enhancing probability of memory addressing failure by
using non-NULL default value for malloced memory (optional, only if
requested by configure option). This helps to track down some
otherwise undetected issues within the testbench and is expected
to be very useful in the future.
We are "abusing" the parser tests to test the TCP/UDP reception
handling. Thus we do not run the other tests through to an IPv4
only config. It acutally is not that parser that matters but rather
the hope to find some init issues when running with only a single
listening socket (in IPv6, we usually both have an IPv4 AND an IPv6
listening socket, what - as practice has shown - may hide listener
setup errors).
however, this had no negative effect, as the message processing state
was not evaluated when a batch was deleted, and that was the only case
where the state could be wrong.
simplified and thus speeded up the queue engine, also fixed some
potential race conditions (in very unusual shutdown conditions)
along the way. The threading model has seriously changes, so there may
be some regressions.
NOTE: the code passed basic tests, but there is still more work
and testing to be done. This commit should be treated with care.
- bugfix: solved potential (temporary) stall of messages when the queue was
almost empty and few new data added (caused testbench to sometimes hang!)
- fixed some race condition in testbench
- added more elaborate diagnostics to parts of the testbench
- solved a potential race inside the queue engine
made shutdown more reliable by makeing sure that the main queue DA worker
is only cancelled if this is actually unavoidable. Also moved down the
deletion of rsyslogd's pid file to immediately before termination, so
that absence of the file is a proper indication that rsyslogd has
finished (in the past, e.g. the testbench accidently ran two intances
as the pid file was deleted too early). Also some improvments to the
testbench, namely to handle aborts more intelligently (but still not
perfect).
This resulted in build errors if no Java was present on the build system,
even though none of the selected option actually required Java.
(I forgot to backport a similar fix to newer releases).
This was a result of an internal processing error if maximum field
sizes had been specified in the property replacer.
Also did some testbench improvements, including omstdout.
