Computation errors

Message boards : Number crunching : Computation errors

To post messages, you must log in.

AuthorMessage
Nick Hill

Send message
Joined: 20 Nov 05
Posts: 7
Credit: 71,277
RAC: 0
Message 5033 - Posted: 3 Dec 2005, 14:35:28 UTC

03/12/2005 1:03:34 pm|rosetta@home|Unrecoverable error for result 1dcj__abrelax_rand_len10_jit02_omega_sim_filters_57294_0 ( - exit code -1073741819 (0xc0000005))

I've been using Boinc for a while & run 6 projects on 2 computers. Rosetta is a new project for me & I've noticed that I get regular computation errors - 1 or 2 per day. I've only ever seen 1 computation error before on any other project & that was a while ago. I assume that this is an issue with Rosetta & as I'm wasting resources on errors, I will reduce the resource share for Rosetta until it is fixed. Can anyone tell me if this issue is being investigated?

Boinc 5.2.12 running on 2 HT threads, WUs set to be removed from memory
Win XP Pro SP2
PIV 3.2gHz HT, 1gb dual ch RAM
ID: 5033 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Rebirther
Avatar

Send message
Joined: 17 Sep 05
Posts: 116
Credit: 41,315
RAC: 0
Message 5034 - Posted: 3 Dec 2005, 14:43:53 UTC - in response to Message 5033.  

03/12/2005 1:03:34 pm|rosetta@home|Unrecoverable error for result 1dcj__abrelax_rand_len10_jit02_omega_sim_filters_57294_0 ( - exit code -1073741819 (0xc0000005))

I've been using Boinc for a while & run 6 projects on 2 computers. Rosetta is a new project for me & I've noticed that I get regular computation errors - 1 or 2 per day. I've only ever seen 1 computation error before on any other project & that was a while ago. I assume that this is an issue with Rosetta & as I'm wasting resources on errors, I will reduce the resource share for Rosetta until it is fixed. Can anyone tell me if this issue is being investigated?

Boinc 5.2.12 running on 2 HT threads, WUs set to be removed from memory
Win XP Pro SP2
PIV 3.2gHz HT, 1gb dual ch RAM


The problem is known but what you can only do is to set the preferences "Leave in memory". You have enough memory for using that ;)

ID: 5034 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Nick Hill

Send message
Joined: 20 Nov 05
Posts: 7
Credit: 71,277
RAC: 0
Message 5035 - Posted: 3 Dec 2005, 15:34:20 UTC - in response to Message 5034.  

The problem is known but what you can only do is to set the preferences "Leave in memory". You have enough memory for using that ;)


Thanks I'll try that - I just had a 2nd error today. I'm a little concerned about the leaving in memory option, woud this affect other projects? I also run Climate & their WUs are pretty big.
ID: 5035 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Rebirther
Avatar

Send message
Joined: 17 Sep 05
Posts: 116
Credit: 41,315
RAC: 0
Message 5038 - Posted: 3 Dec 2005, 15:39:51 UTC - in response to Message 5035.  
Last modified: 3 Dec 2005, 15:40:21 UTC

The problem is known but what you can only do is to set the preferences "Leave in memory". You have enough memory for using that ;)


Thanks I'll try that - I just had a 2nd error today. I'm a little concerned about the leaving in memory option, woud this affect other projects? I also run Climate & their WUs are pretty big.


Yes, the preferences you are setting having effect to all projects!

ID: 5038 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Tern
Avatar

Send message
Joined: 25 Oct 05
Posts: 575
Credit: 4,646,626
RAC: 2,030
Message 5060 - Posted: 3 Dec 2005, 22:22:56 UTC - in response to Message 5038.  
Last modified: 3 Dec 2005, 22:24:31 UTC

Yes, the preferences you are setting having effect to all projects!


For many of the preferences, including "leave in memory", you can create "venues" - home, work, school - and assign different computers to different ones. However, all projects running on a _single_ computer, will use the same set of preferences.

I would reassure you that if you have adequate RAM to even run Rosetta (512MB recommended) then CPDN ("Climate") on the same system is NOT going to be a problem. I'm running Rosetta, CPDN, Einstein, Predictor, and SETI, on a 512MB PC, with "leave in memory" set to "Yes". I haven't sat and watched every project through an entire result, but whenever I've looked, Rosetta has required more RAM than CPDN did. And again, when they aren't running, they take only a few KB of "real" memory - the rest is switched out to disk as virtual memory.

ID: 5060 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile dgnuff
Avatar

Send message
Joined: 1 Nov 05
Posts: 350
Credit: 24,773,605
RAC: 0
Message 5061 - Posted: 3 Dec 2005, 22:28:15 UTC - in response to Message 5035.  

The problem is known but what you can only do is to set the preferences "Leave in memory". You have enough memory for using that ;)


Thanks I'll try that - I just had a 2nd error today. I'm a little concerned about the leaving in memory option, woud this affect other projects? I also run Climate & their WUs are pretty big.


Unless I'm mistaken, the stats for the two boxes you have show that they both have 1 Gb of physical memory. If that's true, you should be OK leaving them in memory. Win XP will move much of the "resident but suspended" data out to the swap file, thus freeing up a good proportion of the physical "chip" memory.

As a followup to Bill Michael's post, if he can pull this off in 512Kb, you'll be OK with the 1 Gb you've got.
ID: 5061 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : Computation errors



©2024 University of Washington
https://www.bakerlab.org