BIND 10 #2157: Create an interface to pass statistics counters in Auth module
BIND 10 Development
do-not-reply at isc.org
Tue Feb 5 06:22:57 UTC 2013
#2157: Create an interface to pass statistics counters in Auth module
-------------------------------------+-------------------------------------
Reporter: y-aharen | Owner:
Type: enhancement | jinmei
Priority: medium | Status:
Component: b10-auth | reviewing
Keywords: | Milestone:
Sensitive: 0 | Sprint-20130205
Sub-Project: DNS | Resolution:
Estimated Difficulty: 8 | CVSS Scoring:
Total Hours: 36.66 | Defect Severity: N/A
| Feature Depending on Ticket:
| Add Hours to Ticket: 0
| Internal?: 0
-------------------------------------+-------------------------------------
Comment (by jinmei):
Replying to [comment:62 y-aharen]:
> > > I applied your suggestion for b10-auth.8. Thank you.
> > >
> > > While testing master with the branch in some environment, I found
lettuce test
> > > fails occasionally. I added a workaround for it (git 1c470ce). If
it's OK,
> >
> > Please be more specific: what was wrong and how it solved the issue.
> > And, in any case, when Jeremy wakes up we should probably create a
> > merge branch and run it on buildbots.
> It was insufficient check for log message output.
> In lettuce test, it checks the counters are expected. Counters are read
from Stats module.
> To avoid reading counters before they've got updated, it checks log
message output; whether
> Auth module receives 'getstats' command and responded to it. It had
checked only for
> Auth module receives 'getstats' command and it failed occasionally in
some environments.
> I added a check for response.
I'd be concerned about the keywords used there. Both "CC_REPLY" and
'"v4":' are pretty short and could match other things (especially the
case for the former); it's also not clear what we are waiting for by
just seeing them in the test feature. Further, don't they necessarily
ensure the statistics are updated at b10-stats, just indicating some
related messages arrived?
But, I guess they were chosen because there was no better alternative
(I've not looked at the code); if so, and if it worked in practice, I
wouldn't block the merge just for that. But I'd suggest the
following:
- add comment about what these expect at least at one of the places
where they were used
{{{
And wait for new bind10 stderr message CC_REPLY
And wait for new bind10 stderr message "v4":
}}}
- create a ticket to introduce a better keyword (probably adding a log
message to b10-stats) and update the test. It should be an easy
task, so probably do it soonish.
--
Ticket URL: <http://bind10.isc.org/ticket/2157#comment:63>
BIND 10 Development <http://bind10.isc.org>
BIND 10 Development
More information about the bind10-tickets
mailing list