Advanced search

Forums : Technical Support : 2.11 seems much better
Message board moderation

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
Profile aardvark1917

Send message
Joined: 13 Jan 08
Posts: 2
Credit: 36,410
RAC: 0
Message 5179 - Posted: 12 Mar 2008, 21:13:48 UTC
Last modified: 12 Mar 2008, 21:15:08 UTC

Only running Cosmology, and the same WU\'s are running forever under 2.11:

3/12/2008 3:52:02 PM|Cosmology@Home|If this happens repeatedly you may need to reset the project.
3/12/2008 3:52:02 PM|Cosmology@Home|Restarting task wu_030608_091046_1_3 using camb version 211
3/12/2008 3:52:54 PM|Cosmology@Home|Task wu_030608_111315_1_0 exited with zero status but no \'finished\' file

This is been happening all day; it takes about 3.5 hours generally for my system to process a WU.
ID: 5179 · Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 11 Dec 07
Posts: 420
Credit: 270,580
RAC: 0
Message 5181 - Posted: 12 Mar 2008, 21:32:21 UTC - in response to Message 5178.  
Last modified: 12 Mar 2008, 21:33:12 UTC



Did you suspend/resume at any point throughout the computation?


Yes, I did.


I now have one example of this in my messages, but it appears that there was no major consequence to it. I\'m putting the Cosmology parts in red, but pay attention to the blue too:

3/11/2008 9:31:57 PM|Cosmology@Home|Starting wu_022708_080900_0_2
3/11/2008 9:31:57 PM|Cosmology@Home|Starting task wu_022708_080900_0_2 using camb version 211

3/11/2008 10:27:38 PM|lhcathome|Starting wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0
3/11/2008 10:27:38 PM|lhcathome|Starting task wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0 using sixtrack version 467
3/11/2008 10:27:39 PM|lhcathome|Computation for task wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0 finished
3/11/2008 10:27:39 PM|lhcathome|Starting wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2
3/11/2008 10:27:39 PM|lhcathome|Starting task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 using sixtrack version 467

3/11/2008 10:27:41 PM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0_0
3/11/2008 10:27:42 PM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0_0
3/11/2008 10:27:42 PM|lhcathome|[file_xfer] Throughput 686 bytes/sec
3/11/2008 10:29:15 PM|lhcathome|Starting wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3
3/11/2008 10:29:15 PM|lhcathome|Starting task wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3 using sixtrack version 467
3/12/2008 3:40:20 AM|lhcathome|Computation for task wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3 finished
3/12/2008 3:40:20 AM|lhcathome|Resuming task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 using sixtrack version 467
3/12/2008 3:40:22 AM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3_0
3/12/2008 3:40:25 AM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3_0
3/12/2008 3:40:25 AM|lhcathome|[file_xfer] Throughput 28264 bytes/sec
3/12/2008 4:43:08 AM|Cosmology@Home|Task wu_022708_080900_0_2 exited with zero status but no \'finished\' file
3/12/2008 4:43:08 AM|Cosmology@Home|If this happens repeatedly you may need to reset the project.

3/12/2008 4:43:08 AM|lhcathome|Task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 exited with zero status but no \'finished\' file
3/12/2008 4:43:08 AM|lhcathome|If this happens repeatedly you may need to reset the project.
3/12/2008 4:43:08 AM|lhcathome|Restarting task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 using sixtrack version 467

3/12/2008 8:40:22 AM|lhcathome|Computation for task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 finished
3/12/2008 8:40:22 AM|lhcathome|Starting wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0
3/12/2008 8:40:22 AM|lhcathome|Starting task wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0 using sixtrack version 467
3/12/2008 8:40:24 AM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2_0
3/12/2008 8:40:27 AM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2_0
3/12/2008 8:40:27 AM|lhcathome|[file_xfer] Throughput 29610 bytes/sec
3/12/2008 1:41:55 PM|lhcathome|Computation for task wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0 finished
3/12/2008 1:41:55 PM|Cosmology@Home|Restarting task wu_022708_080900_0_2 using camb version 211
3/12/2008 1:41:57 PM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0_0
3/12/2008 1:42:00 PM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0_0
3/12/2008 1:42:00 PM|lhcathome|[file_xfer] Throughput 29630 bytes/sec
3/12/2008 3:01:16 PM|Cosmology@Home|Computation for task wu_022708_080900_0_2 finished


Based on what the log says, that Cosmology task was running for 3341 seconds before the \"LHC break\", then 4761 seconds after, which is 8102 seconds. The result says that it ran for 7649 seconds, making the overhead on the entire workunit to be 453 seconds, or 7 minutes and 33 seconds. Part of this time is likely to be other tasks running causing timeslices to be taken away from the computation, but part of it could be having to go back to a prior checkpoint.

Anyway, the real fun thing is that the LHC task was reported the same way at the exact same time, which means that if it is a problem with the science app, LHC is affected by the same thing, as both projects had a task in memory when the message was put into the log file (LHC was running, Cosmology was suspended).

Dunno what this all means and if it is relevant to the other problems... Just throwing it out there to the wind...


ID: 5181 · Report as offensive     Reply Quote
Profile Ananas

Send message
Joined: 19 Jan 08
Posts: 180
Credit: 2,500,290
RAC: 0
Message 5184 - Posted: 13 Mar 2008, 1:48:02 UTC
Last modified: 13 Mar 2008, 1:54:20 UTC

First : I don\'t have any issues with 2.11 on my boxes.

I wonder if the results that complain about \"too many exits\" have the old BOINC heartbeat problems - but if that applies, where did the heartbeat messages go?

Have you seen any of those in the stderr files in the Cosmology slot directory?

Imo. there is a problem in BOINC (Windows) with the way they use the sleep function, the argument to sleep() is often a random value, which can be 0. If an application is ment to sleep for a short time, 0 isn\'t a good idea as this specific value gives control to Windows usually in a critical situation on BOINC side.

Loosing control for more than 30 seconds on the BOINC core client side will make the BOINC API on application side miss the heartbeat.
ID: 5184 · Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 11 Dec 07
Posts: 420
Credit: 270,580
RAC: 0
Message 5185 - Posted: 13 Mar 2008, 2:25:19 UTC - in response to Message 5181.  
Last modified: 13 Mar 2008, 2:31:11 UTC

Top-posting this so people won\'t have to dig as far for my \"a-ha!\" moment... LOL.

I know why this happened to me. I was in the process of making a decision on what tasks to run, so I was browsing the LHC and Cosmology account pages for my computer and was selecting tasks that were either already at quorum (LHC) or needed my result to form quorum (Cosmology). I had one instance where a LHC task was running, but when I resumed another LHC task that was seemingly \"below\" it in my local queue, that task suddenly started running and the other LHC task got paused. What happened is the cpu scheduler went into \"EDF\" mode without explicitly mentioning it in the log file, as the task that fired up was assigned and downloaded to me before the one that got paused, thus its\' due date was before the currently running task, so the scheduler decided that it needed to run, and it did.

However, mine completed normally after that noise...so I doubt that it is related to the renaming of the output file. Might start checking into virus scanners in use, what OS versions, and whatnot...

Edit: As for aardvark1917\'s task, it looks like that same workunit has now bombed on 2 other systems, eventually getting a \"too many exits\".



Did you suspend/resume at any point throughout the computation?


Yes, I did.


I now have one example of this in my messages, but it appears that there was no major consequence to it. I\'m putting the Cosmology parts in red, but pay attention to the blue too:

3/11/2008 9:31:57 PM|Cosmology@Home|Starting wu_022708_080900_0_2
3/11/2008 9:31:57 PM|Cosmology@Home|Starting task wu_022708_080900_0_2 using camb version 211

3/11/2008 10:27:38 PM|lhcathome|Starting wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0
3/11/2008 10:27:38 PM|lhcathome|Starting task wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0 using sixtrack version 467
3/11/2008 10:27:39 PM|lhcathome|Computation for task wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0 finished
3/11/2008 10:27:39 PM|lhcathome|Starting wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2
3/11/2008 10:27:39 PM|lhcathome|Starting task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 using sixtrack version 467

3/11/2008 10:27:41 PM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0_0
3/11/2008 10:27:42 PM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__50__64.31_59.32__12_14__6__25_1_sixvf_boinc342852_0_0
3/11/2008 10:27:42 PM|lhcathome|[file_xfer] Throughput 686 bytes/sec
3/11/2008 10:29:15 PM|lhcathome|Starting wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3
3/11/2008 10:29:15 PM|lhcathome|Starting task wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3 using sixtrack version 467
3/12/2008 3:40:20 AM|lhcathome|Computation for task wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3 finished
3/12/2008 3:40:20 AM|lhcathome|Resuming task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 using sixtrack version 467
3/12/2008 3:40:22 AM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3_0
3/12/2008 3:40:25 AM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__33__64.31_59.32__10_12__6__5_1_sixvf_boinc340808_3_0
3/12/2008 3:40:25 AM|lhcathome|[file_xfer] Throughput 28264 bytes/sec
3/12/2008 4:43:08 AM|Cosmology@Home|Task wu_022708_080900_0_2 exited with zero status but no \'finished\' file
3/12/2008 4:43:08 AM|Cosmology@Home|If this happens repeatedly you may need to reset the project.

3/12/2008 4:43:08 AM|lhcathome|Task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 exited with zero status but no \'finished\' file
3/12/2008 4:43:08 AM|lhcathome|If this happens repeatedly you may need to reset the project.
3/12/2008 4:43:08 AM|lhcathome|Restarting task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 using sixtrack version 467

3/12/2008 8:40:22 AM|lhcathome|Computation for task wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2 finished
3/12/2008 8:40:22 AM|lhcathome|Starting wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0
3/12/2008 8:40:22 AM|lhcathome|Starting task wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0 using sixtrack version 467
3/12/2008 8:40:24 AM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2_0
3/12/2008 8:40:27 AM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__53__64.31_59.32__14_16__6__5_1_sixvf_boinc343222_2_0
3/12/2008 8:40:27 AM|lhcathome|[file_xfer] Throughput 29610 bytes/sec
3/12/2008 1:41:55 PM|lhcathome|Computation for task wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0 finished
3/12/2008 1:41:55 PM|Cosmology@Home|Restarting task wu_022708_080900_0_2 using camb version 211
3/12/2008 1:41:57 PM|lhcathome|[file_xfer] Started upload of file wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0_0
3/12/2008 1:42:00 PM|lhcathome|[file_xfer] Finished upload of file wm72nwp_m72all__49__64.31_59.32__8_10__6__85_1_sixvf_boinc342711_0_0
3/12/2008 1:42:00 PM|lhcathome|[file_xfer] Throughput 29630 bytes/sec
3/12/2008 3:01:16 PM|Cosmology@Home|Computation for task wu_022708_080900_0_2 finished


Based on what the log says, that Cosmology task was running for 3341 seconds before the \"LHC break\", then 4761 seconds after, which is 8102 seconds. The result says that it ran for 7649 seconds, making the overhead on the entire workunit to be 453 seconds, or 7 minutes and 33 seconds. Part of this time is likely to be other tasks running causing timeslices to be taken away from the computation, but part of it could be having to go back to a prior checkpoint.

Anyway, the real fun thing is that the LHC task was reported the same way at the exact same time, which means that if it is a problem with the science app, LHC is affected by the same thing, as both projects had a task in memory when the message was put into the log file (LHC was running, Cosmology was suspended).

Dunno what this all means and if it is relevant to the other problems... Just throwing it out there to the wind...



ID: 5185 · Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 11 Dec 07
Posts: 420
Credit: 270,580
RAC: 0
Message 5186 - Posted: 13 Mar 2008, 2:42:24 UTC - in response to Message 5184.  


I wonder if the results that complain about \"too many exits\" have the old BOINC heartbeat problems - but if that applies, where did the heartbeat messages go?


To the tinman? ;-)

I dunno. I think Scott should perhaps build a new version based on the science approach of 2.05 and the fix for the progress indicator and then test 2.05 with a workunit that failed with \"too many exits\" on versions 2.09 and 2.11, then unit test the new build with the same workunit. If the science change that was mentioned was in the app and not in the INI file, then this testing will give direction as to what addition caused the failure...or if the failure existed in version 2.05 as well...
ID: 5186 · Report as offensive     Reply Quote
Profile DoctorNow
Volunteer tester
Avatar

Send message
Joined: 22 May 07
Posts: 24
Credit: 203,321
RAC: 0
Message 5188 - Posted: 13 Mar 2008, 11:19:23 UTC
Last modified: 13 Mar 2008, 11:20:38 UTC

Just had the first WU of 2.11 which did stuck in a loop: 3624007
It was at 3:35 h and 95%, but did jump back to 50%, crunched to 3:37 and 95% back, then the time code did jump back to 3:35 h and so on.
Don\'t know the exact cause, I guess it happened after the last task switch...
Life is Science, and Science rules. To the universe and beyond
Member of BOINC@Heidelberg
My BOINC-Stats
ID: 5188 · Report as offensive     Reply Quote
Yeti
Avatar

Send message
Joined: 26 Jul 07
Posts: 21
Credit: 3,448,022
RAC: 0
Message 5191 - Posted: 13 Mar 2008, 17:59:46 UTC

@all with problems restarting 2.11er WUs:

Would be interesting to know, wheather you have set \"Leave applications in memory while suspended\" to no ?

If so, switch it to Yes and give it a try

Yeti




Supporting BOINC, a great concept !
ID: 5191 · Report as offensive     Reply Quote
Profile kevint

Send message
Joined: 30 Aug 07
Posts: 46
Credit: 6,502,980
RAC: 0
Message 5192 - Posted: 13 Mar 2008, 18:15:13 UTC


always on leave in memory - since the olden days of Rosetta issues.


ID: 5192 · Report as offensive     Reply Quote
Profile DoctorNow
Volunteer tester
Avatar

Send message
Joined: 22 May 07
Posts: 24
Credit: 203,321
RAC: 0
Message 5193 - Posted: 13 Mar 2008, 18:37:23 UTC - in response to Message 5191.  
Last modified: 13 Mar 2008, 18:39:14 UTC

Would be interesting to know, wheather you have set \"Leave applications in memory while suspended\" to no ?

It is ON in my case. I have enough RAM for it. :-)
So I guess my loop WU was more a day-fly. ;-)
Life is Science, and Science rules. To the universe and beyond
Member of BOINC@Heidelberg
My BOINC-Stats
ID: 5193 · Report as offensive     Reply Quote
muumi

Send message
Joined: 23 Aug 07
Posts: 9
Credit: 2,014,260
RAC: 0
Message 5194 - Posted: 13 Mar 2008, 19:19:48 UTC - in response to Message 5193.  
Last modified: 13 Mar 2008, 19:21:30 UTC

Well now...

I have had several of all these error messages already mentioned.
Now I have a new one, WU \"waiting for memory\".
At the same time the comp is happily running 4 WU.
And allways leave in memory...

wu_030608_211037_8_0
3:49 minutes and 14.651%

http://www.cosmologyathome.org/show_host_detail.php?hostid=15826
2 Gb memory, 143 Gb free HD-space, 2 Gb pagefile.

I have babysitted it and have seen 4 WU\'s start after
the \"waiting for memory\" appeared.

The log does not say anything except that the WU started...

My understanding of this = 0

Maybe I\'ll have a beer or several and wait for tomorrow...
;-)
ID: 5194 · Report as offensive     Reply Quote
Profile Ananas

Send message
Joined: 19 Jan 08
Posts: 180
Credit: 2,500,290
RAC: 0
Message 5195 - Posted: 13 Mar 2008, 19:27:40 UTC - in response to Message 5194.  
Last modified: 13 Mar 2008, 19:30:17 UTC

... \"waiting for memory\"....


The default \"when in use\" value for the memory BOINC is allowed to use is quite low.

Try to increase \"Use at most ... of memory when computer is in use\" and maybe the \"not in use\" one as well (Global settings, Disk and memory usage section) then contact the server (update project) once.
ID: 5195 · Report as offensive     Reply Quote
muumi

Send message
Joined: 23 Aug 07
Posts: 9
Credit: 2,014,260
RAC: 0
Message 5197 - Posted: 13 Mar 2008, 20:25:23 UTC - in response to Message 5195.  

... \"waiting for memory\"....


The default \"when in use\" value for the memory BOINC is allowed to use is quite low.

Try to increase \"Use at most ... of memory when computer is in use\" and maybe the \"not in use\" one as well (Global settings, Disk and memory usage section) then contact the server (update project) once.

Never before any problems with these original settings:
...
Disk and memory usage
Use at most 100 GB disk space
Leave at least
(Values smaller than 0.001 are ignored) 1 GB disk space free
Use at most 50% of total disk space
Write to disk at most every 60 seconds
Use at most 75% of page file (swap space)
Use at most
Enforced by version 5.8+ 50% of memory when computer is in use
Use at most
Enforced by version 5.8+ 90% of memory when computer is not in use
---
As written before, it has done 4 WU\'s, one for each core, all the time, this one WU just decided that it has not enough memory.
The WU in question has changed to \"waiting\" now.
Total credit for the machine is (at this moment) 1,234,200

I\'m still waiting...
(Wasn\'t it Diana Ross who sang it...)
ID: 5197 · Report as offensive     Reply Quote
Profile Ananas

Send message
Joined: 19 Jan 08
Posts: 180
Credit: 2,500,290
RAC: 0
Message 5199 - Posted: 13 Mar 2008, 21:13:19 UTC

Let\'s estimate that you have 1.8GB free RAM - if there are other projects sleeping in the background, it\'s probably less than that, but let\'s assume it is 1.8GB

50% of that (allowed while computer is in use) are 900k

4 tasks are 225k for each task

You box has a lot of RAM but if you allow BOINC to use only half of it ...
ID: 5199 · Report as offensive     Reply Quote
muumi

Send message
Joined: 23 Aug 07
Posts: 9
Credit: 2,014,260
RAC: 0
Message 5200 - Posted: 13 Mar 2008, 22:25:02 UTC - in response to Message 5199.  

The computer is dedicated to BOINC.
Nothing else is running.
It is behind a hardware firewall so no software firewall, nor active AV.
AV is run once a week manually.

\"First connected at 2007-10-11 / 5:21:58\" - from Boincstats

The only project at the moment is C@H.

As I Said before, it is happily running 4 C@H WU\'s while this one was waiting for memory.

Right now the WU in question has started again, about 33 % done.
30 min CPU time, about 6,5 hours of wall time...



Let\'s estimate that you have 1.8GB free RAM - if there are other projects sleeping in the background, it\'s probably less than that, but let\'s assume it is 1.8GB

50% of that (allowed while computer is in use) are 900k

4 tasks are 225k for each task

You box has a lot of RAM but if you allow BOINC to use only half of it ...

ID: 5200 · Report as offensive     Reply Quote
Profile aardvark1917

Send message
Joined: 13 Jan 08
Posts: 2
Credit: 36,410
RAC: 0
Message 5201 - Posted: 13 Mar 2008, 22:41:30 UTC

I simply aborted the WU and no problems now.

BTW, running on a laptop. I frequently put BOINC on pause, then close the lid. Out of hibernation, BOINC starts right back up. Never had any problems with that.
ID: 5201 · Report as offensive     Reply Quote
Brian Silvers

Send message
Joined: 11 Dec 07
Posts: 420
Credit: 270,580
RAC: 0
Message 5204 - Posted: 13 Mar 2008, 23:56:57 UTC - in response to Message 5200.  


As I Said before, it is happily running 4 C@H WU\'s while this one was waiting for memory.


Different results have different memory requirements. It could be that this particular task needs a little more than normal. If you look at the memory usage in Task Manager, is this one using more memory or is there one task that is using much more than another?
ID: 5204 · Report as offensive     Reply Quote
Profile Ananas

Send message
Joined: 19 Jan 08
Posts: 180
Credit: 2,500,290
RAC: 0
Message 5215 - Posted: 14 Mar 2008, 7:20:35 UTC - in response to Message 5200.  
Last modified: 14 Mar 2008, 7:25:15 UTC

The computer is dedicated to BOINC.
Nothing else is running.
...


One more reason to increase the setting for

Use at most 50% of memory when computer is in use

If you do not need RAM for yourself, why not let BOINC use it?

Set it to 90% and the setting when computer is not in use to 95% and your \"waiting for memory\" message will disappear :-)


p.s.: Some Cosmo WUs do need a bit more, I already have seen 198MB used by camb 2.11, whereas others seem to be happy with 160MB
ID: 5215 · Report as offensive     Reply Quote
Previous · 1 · 2

Forums : Technical Support : 2.11 seems much better