climateprediction.net home page
Pirates interfering with CPDN ?

Pirates interfering with CPDN ?

Message boards : Number crunching : Pirates interfering with CPDN ?
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 28592 - Posted: 11 May 2007, 7:41:00 UTC
Last modified: 11 May 2007, 7:42:35 UTC

This happened today - is it just a coincidence or did the Pirates WU have something to do with it?


11/05/2007 08:20:22|Pirates@Home|Computation for task wu_1178857808_157_1 finished
11/05/2007 08:20:24|climateprediction.net|Restarting task hadcm3inct_ckyb_1920_160_05887912_0 using hadcm3i version 540
11/05/2007 08:20:25|Pirates@Home|[file_xfer] Started upload of file wu_1178857808_157_1_0
11/05/2007 08:20:28|Pirates@Home|[file_xfer] Finished upload of file wu_1178857808_157_1_0
11/05/2007 08:20:28|Pirates@Home|[file_xfer] Throughput 3552 bytes/sec
11/05/2007 08:20:44|Pirates@Home|Sending scheduler request: To fetch work
11/05/2007 08:20:44|Pirates@Home|Requesting 8518 seconds of new work, and reporting 1 completed tasks
11/05/2007 08:20:51|climateprediction.net|Deferring communication for 1 min 0 sec
11/05/2007 08:20:51|climateprediction.net|Reason: Unrecoverable error for result hadcm3inct_ckyb_1920_160_05887912_0 (The device does not recognize the command. (0x16) - exit code 22 (0x16))
11/05/2007 08:20:51|climateprediction.net|Computation for task hadcm3inct_ckyb_1920_160_05887912_0 finished
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_1.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_2.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_3.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_4.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_5.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_6.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:51|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_7.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_8.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_9.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_10.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_11.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_12.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_13.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_14.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_15.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent
11/05/2007 08:20:52|climateprediction.net|Output file hadcm3inct_ckyb_1920_160_05887912_0_16.zip for task hadcm3inct_ckyb_1920_160_05887912_0 absent


Heres the logfile from CPDN -
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=6508719

Very frustrating - the WU was just getting started...........

ID: 28592 · Report as offensive     Reply Quote
Jord
Avatar

Send message
Joined: 5 Aug 04
Posts: 250
Credit: 93,274
RAC: 0
Message 28595 - Posted: 11 May 2007, 8:32:33 UTC

You do not have enough memory to run CPDN. A minimum of 512MB is needed.
Jord.
ID: 28595 · Report as offensive     Reply Quote
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 28596 - Posted: 11 May 2007, 8:55:12 UTC

That\'s strange. I\'ve run CPDN before on this PC and managed to finish a sulphur model previously.....

Perhaps the requirements for this project have changed.
I\'ll run other projects where 512MB of RAM is sufficient.

ID: 28596 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 28597 - Posted: 11 May 2007, 9:27:40 UTC

John

Your computer info says you have 479.29 MB of ram. This tends to indicate that you don\'t have a separate display card, so your computer is using some of the system ram for display purposes.
The latest version of the TCMs has been optimised to write to the hard disk less often, but the trade off is that it uses more ram then the earlier TCMs, and a fair bit more than the sulphur/slab models, which were totally different. (No ocean calcs.)

Adding a card, or increasing ram to 1GB will help. Both would help a lot.


Backups: Here
ID: 28597 · Report as offensive     Reply Quote
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 28598 - Posted: 11 May 2007, 9:53:36 UTC
Last modified: 11 May 2007, 9:55:02 UTC

Thanks for the explanation, Les - much appreciated!
Correct - I don\'t have a display card.
Do you need to change the system requirements section in the \'Getting Started\' section? Other prospective crunchers may be in the same situation as myself.....

ID: 28598 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 28599 - Posted: 11 May 2007, 11:01:04 UTC

Hi John

Your 479Mb RAM is a bit short of the 512Mb one should have per model, but you\'ll probably still manage as long as you

*disable the screensaver if you haven\'t already
*when you look at the globe thro the View graphics button, don\'t maximise the globe window
*suspend the model thro the Activity menu before you need to do anything intensive like games
*only run ONE model even tho the computer is dual-core. I think you should set your preferences to \'Use at most one core\' so you don\'t get workunits from other projects running alongside it. That\'s because it\'s ideally 512Mb PER MODEL.

I think your model may have crashed because you ran an AV scan without first exiting from boinc. There\'s a \'scan_lockfile\' message.

It would be worth checking on the project READMEs which have a lot of very useful advice on keeping models safe. Get to them thro my sig. I\'d recommend the top tips in the README called Running the model.

And in the README about avoiding crashes, item #5 by Mike (which advises for example on avoiding 161 error crashes, which you\'ve had in the past). Plus item #1 by Les which is an easy backup method. That way, if your model does crash, you just restore the backup and continue working on it.

Otherwise, there\'s a whole README about different backup methods.


Cpdn news
ID: 28599 · Report as offensive     Reply Quote
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 28600 - Posted: 11 May 2007, 11:17:19 UTC

Thanks everyone for the help here.
Two things mentioned by mo.v that probably caused this morning\'s crash -
1) Yes - my machine does an automatic AV scan every morning at 0900 GMT.
2) My PC is a Pentium 4 H/T and it was running ABC alongside the CPDN model with no problems. It must have been the addition of the AV scan that tipped it \'over the edge\'.
3) (For fun....) Is there really a screensaver that you can see on BOINC projects?
ID: 28600 · Report as offensive     Reply Quote
Profile mo.v
Volunteer moderator
Avatar

Send message
Joined: 29 Sep 04
Posts: 2363
Credit: 14,611,758
RAC: 0
Message 28605 - Posted: 11 May 2007, 14:28:47 UTC

You need to exclude the boinc directory from the AV scan options.
Cpdn news
ID: 28605 · Report as offensive     Reply Quote
Profile Pooh Bear 27
Avatar

Send message
Joined: 5 Feb 05
Posts: 465
Credit: 1,914,189
RAC: 0
Message 28606 - Posted: 11 May 2007, 16:01:08 UTC - in response to Message 28605.  

You need to exclude the boinc directory from the AV scan options.

I can state that some AVs I have tested need to exclude BOINC directory and all sub-directories. I have tested only Norton, AVG and McAfee (home version and enterprise). Norton and AVG need to not scan BOINC. It seems McAfee is OK to scan (at least I have not had an issue) on XP and Vista.

Your mileage may vary, but it is what I have been doing for a while now. My McAfee boxes sit nice without incident. I excluded BOINC on Norton and AVG, and they also run without incident.


ID: 28606 · Report as offensive     Reply Quote
Jord
Avatar

Send message
Joined: 5 Aug 04
Posts: 250
Credit: 93,274
RAC: 0
Message 28609 - Posted: 11 May 2007, 16:45:44 UTC - in response to Message 28606.  

I excluded BOINC on Norton and AVG, and they also run without incident.

The trouble with AVG Free is that you can\'t exclude any directories in the automatic test. Yes, you can set up a personalized scan, but you can only run that when running it by hand. Not automatically. The User Test is only available in AVG Professional. Not something most of us run.
Jord.
ID: 28609 · Report as offensive     Reply Quote
m.mitch
Avatar

Send message
Joined: 10 Jan 06
Posts: 55
Credit: 1,256,164
RAC: 8,648
Message 28620 - Posted: 12 May 2007, 1:58:24 UTC

I think John, who\'s a wealthy Insurance Executive 8O and part time spy ;), should buy another 512M stick of RAM.



Click here to join the #1 Aussie Alliance on Climate Prediction
ID: 28620 · Report as offensive     Reply Quote
Profile Strathpeffer
Avatar

Send message
Joined: 9 Jan 07
Posts: 497
Credit: 342,899
RAC: 0
Message 28621 - Posted: 12 May 2007, 3:13:13 UTC
Last modified: 12 May 2007, 3:15:34 UTC

I expect that\'s right, but:

I\'ve completed 3 BBC models on computers with only 512Mb of RAM, part of it shared - but I wouldn\'t attempt to run another BOINC project at the same time and I suspend the models while doing anything else that demands a lot of the computer.

I\'ve also run (but not completed) BBC and CPDN models on an old computer with only 256Mb of RAM. The furthest I ever got with a BBC model was 1943 (and then I transferred the backup to a newer computer), but the only CPDN model I\'ve attempted so far got to 1976 without ever crashing - and has now been transferred to a newer computer only because the newer one completed its BBC model.

An automatic AVG Free scan runs on all of these computers at 8 am daily and has never yet caused a model to crash - although sometimes the old computer shuts itself off during the scan.

Maybe I\'ve just been lucky - or careful - and I make regular backups. ;-)
Visit the Scotland team
ID: 28621 · Report as offensive     Reply Quote
Profile John Hunt
Avatar

Send message
Joined: 5 Mar 05
Posts: 64
Credit: 790,577
RAC: 0
Message 28622 - Posted: 12 May 2007, 5:57:37 UTC - in response to Message 28620.  

I think John, who\'s a wealthy Insurance Executive 8O and part time spy ;), should buy another 512M stick of RAM.


Definitely not wealthy! I\'ll have to put in a bit of overtime to fund a RAM upgrade.
This is the 3rd time I\'ve had the problem of \'not enough memory\' -
1st - I couldn\'t do SAP (when it was still going)
2nd - Some of the WCG WUs need 1G to run.
3rd - now I can\'t run CPDN alongside other projects as I have done in the past.

Looking on the positive side though - this old machine of mine hasn\'t done too badly. I\'ll do the upgrade as a reward for when it passes the quarter million credits......




ID: 28622 · Report as offensive     Reply Quote
old_user95764

Send message
Joined: 30 Aug 05
Posts: 2
Credit: 61,753
RAC: 0
Message 28623 - Posted: 12 May 2007, 7:55:57 UTC

I haven\'t been able to recieve any wu\'s since 4-21 is there any reason. I\'ve reinstalled to make sure I got updated version. I know seti has and is still having probs. Will soon be up and running but I don\'t know what\'s up w/ CP any help would be helpful

Thanks
ID: 28623 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 28628 - Posted: 12 May 2007, 8:45:26 UTC


I see that you got one soon after posting, so it must have been something to do with how you had BOINC set up previously.
Perhaps you had No new tasks selected in the Projects tab.


Backups: Here
ID: 28628 · Report as offensive     Reply Quote
Profile MikeMarsUK
Volunteer moderator
Avatar

Send message
Joined: 13 Jan 06
Posts: 1498
Credit: 15,613,038
RAC: 0
Message 28629 - Posted: 12 May 2007, 8:45:56 UTC
Last modified: 12 May 2007, 8:46:29 UTC

Looks like you sorted out your problem (a work unit dated 20 minutes after your post).

Was it the \'no more work\' button by any chance?


-- Edit: SNAP :-)
I'm a volunteer and my views are my own.
News and Announcements and FAQ
ID: 28629 · Report as offensive     Reply Quote
Profile old_user35834
Avatar

Send message
Joined: 12 Jan 05
Posts: 12
Credit: 40,824
RAC: 0
Message 28633 - Posted: 12 May 2007, 10:40:57 UTC
Last modified: 12 May 2007, 10:44:16 UTC

I got the same error as John for this work unit: http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=6506855

The machine is running XP. It\'s got 2GB of RAM, and plenty of disk space. The model runs alongside either MalariaControl or Leiden.

I have no antivirus software installed, so the crash definitely was not the result of a virus scan in progress.

One thing worth mentioning about the work unit. It was only running for a couple of days. After about 40 hours CPU time, BOINC suddenly showed only 24 hours CPU time. The BOINC message log did not show any crash or stop/restart. But in task manager I saw that the process was only running for 3 hours, where it had ran non-stop since the start. Luckily (I thought) the current timestep had not dropped back, so I thought no work was lost. You can see this clearly in the work unit\'s trickle info.

BOINC.BE: For Belgians who love the smell of glowing red cpu's in the morning
Tutta55's Lair
ID: 28633 · Report as offensive     Reply Quote
Les Bayliss
Volunteer moderator

Send message
Joined: 5 Sep 04
Posts: 7629
Credit: 24,240,330
RAC: 0
Message 28634 - Posted: 12 May 2007, 11:13:46 UTC


I\'ve sent a note to the programmers.

(John\'s was a different problem, even though he had the same error code.)

ID: 28634 · Report as offensive     Reply Quote
old_user95764

Send message
Joined: 30 Aug 05
Posts: 2
Credit: 61,753
RAC: 0
Message 28647 - Posted: 12 May 2007, 19:41:56 UTC - in response to Message 28623.  

I haven\'t been able to recieve any wu\'s since 4-21 is there any reason. I\'ve reinstalled to make sure I got updated version. I know seti has and is still having probs. Will soon be up and running but I don\'t know what\'s up w/ CP any help would be helpful

Thanks


That\'s just the way it goes I guess. Thanks
ID: 28647 · Report as offensive     Reply Quote

Message boards : Number crunching : Pirates interfering with CPDN ?

©2024 climateprediction.net