On Wed, Jun 30, 2010 at 06:51:08PM -0600, sanjay nadkarni wrote: > On 05/ 5/10 11:21 AM, [email protected] wrote: > >On Tue, May 04, 2010 at 11:41:44PM -0700, Danek Duvall wrote: > >>[ Replying to pkg-discuss, since that's where we discuss these things. ] > >> > >>sanjay nadkarni wrote: > >> > >>>My desktop is running build 136. I tried to upgrade it to build 138 > >>>and pkg image-update hung. > >>>I have taken a pstack of this and it is available at: > >>>http://paste.east.sun.com/1821 > >>>My system has 8GB memory. > >>2503: /usr/bin/python2.6 /usr/bin/pkg image-update --be-name osol-138 -v > >> feee1476 __clock_gettime (4, 8045d80, 0, fcd9f31c) + 36 > >> fcd9f331 curlx_tvnow (8045dc8, f66b098, 51b6519f) + 21 > >> fcdc7ef1 curl_multi_perform (85998c0) + 71 > >> fce0850f do_multi_perform (88662d4, 0, 8045e78, fed48d92) + 83 > >> fed490c0 call_function (8045eec, 0, 30, 0) + 33c > >> fed46221 PyEval_EvalFrameEx (88e1d34, 0, 875ed74, 0) + 3029 > >> > >>That stack doesn't look like it's hung, just like it's waiting for the > >>network (but johansen might know better). > >So far, I agree with Danek's analysis. It could help if we got a pstack > >that uses the python stack helper that was added in b137, too. > > > >-j > Unfortunately, I could not get back to this earlier. I have not > been able to move to any release beyond this. I even went back to > an earlier build to try and move up and that failed too. Should I > simply take a coredump of the system when this is occurring and > provide that ?
That's not necessary. The bug has already been root-caused, and will be fixed in b144, IIRC. https://defect.opensolaris.org/bz/show_bug.cgi?id=16203 -j _______________________________________________ pkg-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
