idp-Integration-tests

[Tom Zeller]

> I should add that in my testing just stopping the maven test left a jetty
> process hanging around which of course meant that the next time around we
> didn't start properly. So if the tests hang (and it rather looks as if it
> has) we will need to go in and kill all the jetty processing that have been
> thrown into existence on the Jenkins box.

Yup.

I find those couple of tests useful, but we may change our minds. I've tried to minimize time spent on them, with some success. Really I would like them to run the installer as part of setup.

(mobile)



> 
>> -----Original Message-----
>> From: dev-bounces at shibboleth.net [mailto:dev-bounces at shibboleth.net] On
>> Behalf Of Rod Widdowson
>> Sent: 11 December 2014 15:06
>> To: Shib Dev
>> Subject: idp-Integration-tests
>> 
>> This is probably for Tom.
>> 
>> Now that Jetty logs to disk rather than to stdout, the trick we used for
>> doing a popen("jetty") and waiting for the "I'm started" messages, doesn't
>> work to good.
>> 
>> I've have checked in a bit of a "patch" (pronounced "hack") to get around
>> that: we continually wait for the log to exist and then for the "I'm
>> started" line.  We then have to delete the log when we kill the process.
>> This appears to work locally, but I'd be glad of some review of the code.
>> 
>> Longer term there is probably a better way to wait for jetty to be
> started:
>> maybe we can ping the status page?
>> 
>> R
>> 
>> --
>> To unsubscribe from this list send an email to
> dev-unsubscribe at shibboleth.net
> 
> -- 
> To unsubscribe from this list send an email to dev-unsubscribe at shibboleth.net