e154c2cefe
I'm trying to debug a problem where the last thing in the logs is "Starting instance...". Unfortunately this occurs twice in the same source file, so it's hard to figure out a priori which log statement it came from. Audit and info messages don't disambiguate their function/line number, all you get is the logger name like: 2014-04-06 19:51:01.871 8288 AUDIT nova.compute.manager [req-...] [instance: ...] Starting instance... I think it doesn't hurt to to make sure all logs at this level are unique, even if upstream were to log the function (see [1]). Most duplicate LOG.info and LOG.audit level messages are in separate files so we can disambiguate by the name, but there are a few that overlap. For reference, the less-then-perfect but useful check I used was: $ grep -r 'LOG.audit' * | sed 's/:\s*/ /' | sort -k1 | uniq -f1 -c [1] https://review.openstack.org/#/c/85638/ Change-Id: If1ff573ffed1d12fcadfbdfd3365aae68a5fb977 |
||
---|---|---|
.. | ||
__init__.py | ||
xvp_proxy.py |