lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Dawid Weiss <dawid.we...@cs.put.poznan.pl>
Subject Re: [JENKINS] Lucene-Solr-4.x-Linux-Java7-64 - Build # 288 - Failure!
Date Sun, 01 Jul 2012 16:29:53 GMT
> :) After my shoulder surgery I have time to follow the Jenkins runs with one
> hand only :)

Hope it's not anything serious. I had shoulder tendon reattached. It
felt like hell for three weeks until it kind of healed. Very, very
bad.

> Can we use test timing logs to automatically define large timeouts? Like 10
> times normal runtime?

Not really because they're computed on one machine and the scaling
factor is unknown (only their relative values matter for scheduling).
Also, all this takes place at <junit4> level which is one level
"above" the runner (for example maven builds wouldn't have it).

> If a test hangs, I would use System.exit(failure) and print stack traces
> before (and switch all threads to daemon mode if not already).

This isn't so simple -- a call to System.exit would stop the forked vm
without passing all the cleanup events back to the aggregator
(<junit4>)... and it's not a nice solution anyway. I think a regular
thread dump within the test scope will suffice (followed by graceful
ignored completion for all remaining tests). I don't see a reason why
we should terminate immediately if a timeout occurs -- it's an event
much like anything else, only problematic in that resuming normal
operation is potentially impossible.

Dawid

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message