A Topaz session on a staging server just suffered the following error:
potential infinite loop at 07/16/2014 07:49:17 SAST FindFreeFrame: found free page and exited loop at 07/16/2014 07:50:21 SAST (64 seconds) FindFreeFrame: potential infinite loop at 07/16/2014 07:50:31 SAST FindFreeFrame: found free page and exited loop at 07/16/2014 07:52:36 SAST (125 seconds) FindFreeFrame: potential infinite loop at 07/16/2014 07:52:46 SAST FindFreeFrame: found free page and exited loop at 07/16/2014 08:09:40 SAST (1014 seconds) [Info]: Logging out at 07/16/2014 08:09:45 SAST ----------------------------------------------------- GemStone: Error Fatal Your GemStone session has been forcibly terminated, , stone shutting down Error Category: 231169 [GemStone] Number: 4059 Arg Count: 1 Context : 20 exception : 20 Arg 1: 20 The symbol gem, and all our other logged in gems also logged about "FindFreeFrame: potential infinite loop" at the same time. Any advice on hunting this down? _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
Pieter,
This message indicates that you've filled up your SPC. The likely situation is that you've got one or more gems in the process of doing "large commits" that are hogging most of the pages in the SPC. Gems are allowed to preempt most pages from the cache, but pages that are involved in a commit cannot be pre-empted and _could_ lead this situation....
If you are not running statmon, you should. Statmonitor has a number of stats that make it possible to view the health of the SPC and we can pretty much tell what's wrong by looking at the statmon output.
Dale On Wed, Jul 16, 2014 at 12:32 AM, <[hidden email]> wrote: A Topaz session on a staging server just suffered the following error: _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
Hi,
I’m running into this issue during a restore of a backup in GS 3.1.0.5 It says: FindFreeFrame: potential infinite loop at 02/24/15 17:02:19 CET And I see the pgsvrmain process hitting 100% and I eventually kill the entire process because it’s taking too long. The stone log files are not giving me any info. After killing the restore process, The SPC is a 500MB one and the backup is of a database of 18GB. I’m going to try restoring on a larger machine later tonight and see what the issue is but I wonder what the constraints on the SPC are during a restore from backup? cheers Johan
_______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
Johan,
I've forwarded this to engineering ... did the large machine help you get past the problem? If I'm not mistaken, these days, you will get this message if you fill the SPC with ditry pages for the current transaction (i.e., the isze of your transaction in dirty pages is larger than the SPC).... statmons during the restore would help to confirm this ... Dale On 2/24/15 8:23 AM, Johan Brichau via
Glass wrote:
Hi, _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
Thanks. Yes, a small increment of the SPC to 750MB did the trick. I suspect there’s a very large collection (or maybe even a string) in that database. But it’s pending investigation (did not have a chance to look at that yet). Johan _______________________________________________ Glass mailing list [hidden email] http://lists.gemtalksystems.com/mailman/listinfo/glass |
Free forum by Nabble | Edit this page |