Message boards : BOINC client : Core client periodically crashes - I have the std err file!
Message board moderation
Author | Message |
---|---|
Send message Joined: 17 Mar 06 Posts: 5 |
Hi folks, Periodically (every few days) I notice that BOINC is not doing any work. I see the message that the manager cannot connect to the client. I check my task manager and the client is not running! I poked around a bit and I found the stderrdae.txt file with the following lines: (lines removed...) 2006-03-13 11:05:12 [rosetta@home] Unrecoverable error for result FA_RLXac_hom012_1acf__359_106_0 ( - exit code -164 (0xffffff5c)) 2006-03-13 15:54:22 [rosetta@home] Unrecoverable error for result FA_RLXb3_hom030_1b3aA_359_141_0 ( - exit code -164 (0xffffff5c)) ***UNHANDLED EXCEPTION**** Reason: Access Violation (0xc0000005) at address 0x0032F114 read attempt to address 0x00000008 1: 03/13/06 17:29:26 1: SymGetLineFromAddr(): GetLastError = 126 2006-03-14 08:45:02 [---] Couldn't resolve hostname [predictor.scripps.edu] 2006-03-14 08:45:02 [---] Couldn't resolve hostname [boinc.bakerlab.org] (lines removed...) 2006-03-14 18:49:48 [boincsimap] Scheduler request to http://boinc.bio.wzw.tum.de/boincsimap_cgi/cgi failed with a return value of 500 2006-03-14 18:49:48 [boincsimap] No schedulers responded ***UNHANDLED EXCEPTION**** Reason: Access Violation (0xc0000005) at address 0x0032F114 read attempt to address 0x00000008 1: 03/14/06 19:23:59 1: SymGetLineFromAddr(): GetLastError = 126 2006-03-15 09:17:26 [---] Failed to stop applications; aborting CPU benchmarks 2006-03-15 11:00:57 [boincsimap] Project is down 2006-03-15 17:58:19 [boincsimap] Scheduler request to http://boinc.bio.wzw.tum.de/boincsimap_cgi/cgi failed with a return value of 500 (lines removed...) 2006-03-16 14:53:12 [boincsimap] No work from project 2006-03-16 15:09:14 [boincsimap] No work from project ***UNHANDLED EXCEPTION**** Reason: Access Violation (0xc0000005) at address 0x0032F114 read attempt to address 0x00000008 1: 03/16/06 21:55:45 1: SymGetLineFromAddr(): GetLastError = 126 (EOF) Interesting that it's the same addresses in memory that the error keeps occurring at! Any ideas? Is this a bug? What can I do so I don't have to keep restarting BOINC? Ulysses |
Send message Joined: 17 Mar 06 Posts: 5 |
Ooops, I should have said I'm running 5.2.13 BOINC on my Windows XP (service pack 2) box. |
Send message Joined: 8 Sep 05 Posts: 168 |
Ok here are somethings to do, Rosetta likes to be left in memory, check your firewall for all the needed exceptions, Rom walton says Boinc needs the following: BOINC Manager needs to be able to open outbound connections to either 1043 or 31416. BOINC Screensaver needs to be able to open outbound connections to either 1043 or 31416. BOINC needs to open an inbound connection to either 1043 or 31416, and an outbound connection to port 80 for downloads and uploads of workunits, results, and science packages. My understanding was that 443 was already in there for the "account manager" stuff, would send all passwords via HTTPS. I don't know if that means "only" to AMS, or if it means to "attach project" for V5 servers as well. Run the following to test for stability and memory problems...' Prime95 Torture test Memtest86+ BOINC Wiki |
Send message Joined: 17 Mar 06 Posts: 5 |
OK, I set stay in memory true for all my connected projects. I don't have the windows firewall running, as I'm behind the corp firewall. Over the weekend I still get the same issue. I'm attached to 4 projects (simap, rosetta, Predictor@home, World Community Grid) are there any known issues with any of them? Any other ideas? |
Send message Joined: 17 Mar 06 Posts: 5 |
I noticed that the world community grid is using rosetta version 4.22 and the rosetta project is using 4.82. Could this older version be causing the issue? I sent the world community grid an email letting them know about the version issue. |
Send message Joined: 8 Sep 05 Posts: 168 |
I noticed that the world community grid is using rosetta version 4.22 and the rosetta project is using 4.82. Could this older version be causing the issue? Not the same Rosetta, one will not run the other.... BOINC Wiki |
Send message Joined: 30 Aug 05 Posts: 17 |
I have seen the same failure periodically. ***UNHANDLED EXCEPTION**** Reason: Access Violation (0xc0000005) at address 0x0032F114 read attempt to address 0x00000008 1: 02/03/06 13:12:34 1: SymGetLineFromAddr(): GetLastError = 126 |
Send message Joined: 24 Nov 05 Posts: 129 |
I have seen the same failure periodically. Jim, That particular error has proved to be an extremely pesky one to track down and eliminate, and is still being worked on. Hopefully, it will be fixed in the upcoming 5.4.x version, which has been in a particularly long beta "incubation" stage. Michael R. "The arc of history is long, but it bends toward Justice" |
Send message Joined: 29 Aug 05 Posts: 15585 |
Hum... If at least it is for the Einstein application. Multiple other apps have given the same errors. Without knowing which project this person was crunching for, it's useless to figure out. It's still the science application that breaks out in the error. Not BOINC itself. |
Send message Joined: 17 Mar 06 Posts: 5 |
Here is an update... I was attached to 4 projects when I was getting this issue. 1) Simap 2) Rosetta 3) Predictor@home 4) World Community Grid I suspended World Community Grid and left the remainnig 3 running. I've had no more client crashes. Sorry World Community Grid I'll have to leave you suspended! |
Copyright © 2025 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.