[torqueusers] deadlock in torque p6

[Garrick Staples]

On Thu, Feb 03, 2005 at 07:56:55PM -0800, Garrick Staples alleged:
> On Fri, Feb 04, 2005 at 02:29:04PM +1100, Chris Samuel alleged:
> > On Fri, 4 Feb 2005 12:03 am, Marcin Mogielnicki wrote:
> > 
> > > The solution would be to update the state of the node every time mom is
> > > started. It can be done in a very simple way. The patch is given below.
> > 
> > Hi Marcin,
> > 
> > Is this the release version of p6 that we're talking about ?
> > 
> > Have you looked at any of the 1.2 betas to see if those work?  I believed this 
> > has been addressed in these..
> 
> A small patch seems to have been dropped in 1.1.0p6.  I remember seeing it in
> p5.
> 
> Further testing reveals that my original patch didn't address all scenerios
> anyways.  There's a couple that I missed before, and Marcin's is still another
> one.

And even more testing reveals the original patch was the wrong direction
anyways.

Please see the patch at http://clusterresources.com/bugzilla/show_bug.cgi?id=41
and please comment if it works or breaks something.  It should apply to
1.2.0p6.

-- 
Garrick Staples, Linux/HPCC Administrator
University of Southern California
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://www.supercluster.org/pipermail/torqueusers/attachments/20050203/3926b2cc/attachment.bin