Dan Price wrote:
On Fri 07 Aug 2009 at 05:24PM, Shawn Walker wrote:
Dan Price wrote:
3061 depotcontroller exposes current process file descriptors to subprocess
6035 depotcontroller doesn't check exit status causing misleading timeout
errors
9626 start_expected_fail sleeptime could be more generous
Some cleanup of the depot controller. I did test under umem
since I'm changing the spawn code.
http://cr.opensolaris.org/~dp/depotcontroller-fixes/
I didn't review the pspawn parts, but the depotcontroller changes look
right to me.
Cool thanks. I wasn't sure if I had interpreted 6035 properly, and
can't readily test it in the context of hudson...
The easiest way to test would be to add a test to cli/t_pkg_depotd.py
that attempted to start the depot with bogus options. If it fails with
a DepotStateException, then things are probably good.
To be more specific, I've also encountered this problem outside of
hudson, where I had introduced an error into depot.py (such as
forgetting to import a module) and every test that depended on the depot
process would just hang for a minute or so before failing and moving to
the next, despite the fact that the depot had exited immediately with a
return code or syntax error.
Cheers,
--
Shawn Walker
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
