On 19/07/16 16:44, Nathan Scott wrote:
...
----- Original Message -----
...
1108 is a mystery ... we get 2 primary pmloggers started ...
I finally nailed this one ... see ugly commit from this morning for
pmlogger.
361 has gone a bit under the radar ...
> ...
Fixed now.
Yep, passing everywhere now, thanks.
381 is possibly due to pmlogger being more resilient to pmcd &| pmda restarts
now ... but I'd have expected it to see the same failure signature everywhere?
This one remains unresolved.
That 581 failure we've talked about before too I think - seems to be sensitive
to number of open fds in pmcd,...
I think this is 578 (not 581) .. the +/-1 slop I added reduced the
failure rate, but I'm still seeing some failures ... I'd have to
increase it to -1/+4 slop to get 'em all passing, and I'm not sure if
that is the right thing to do, but will probably make that change given
what the test is trying to establish (and in particular that we have no
evidence of fd leaks in this part of the code).
823 I'm certain is also a _notrun candidate - some versions of SASL seem buggy
and a newly created user becomes oddly invisible. It may be worth collecting
"pmconfig -L sasl_version" from the failing machines and looking for a pattern
that could be squashed by _notrun? Certainly passes reliably for me on recent
SASL library versions anyway.
No luck on this one ... random checking suggests the same version on
passing and failing hosts, e.g. passes on vm03, fails on vm02, but ...
kenj@bozo:~$ ssh vm02 pmconfig -L sasl_version
sasl_version=2.1.26
kenj@bozo:~$ ssh vm03 pmconfig -L sasl_version
sasl_version=2.1.26
|