Message boards : Number crunching : Not getting tasks. Probably need help
Message board moderation

To post messages, you must log in.

AuthorMessage
Profile Kathryn Tombaugh Weber

Send message
Joined: 12 Sep 11
Posts: 38
Credit: 218,154
RAC: 0
Message 26323 - Posted: 6 Apr 2014, 23:51:10 UTC



Not sure if you can see this image or not. I'm kind of new at it. Basically I'm not getting any tasks even though there are hundreds of thousands, and I was trying to show you my event log. Probably need help here please. [/url]
ID: 26323 · Report as offensive     Reply Quote
Profile Conan
Avatar

Send message
Joined: 6 Jul 06
Posts: 107
Credit: 511,942
RAC: 0
Message 26324 - Posted: 7 Apr 2014, 1:39:09 UTC
Last modified: 7 Apr 2014, 1:42:24 UTC

According to you computer it shows that you have 2 running jobs since the 3/4/14 and 4 pending.
2 of the pending jobs date back to 2013, I am not sure what their problem is.

Is this the only project you do on this computer?
If not then time is shared across all projects.

When you complete at least one of your current jobs you will get another.

If you don't actually have the 2 jobs on your computer that are listed at the project, then a reset of the project via the BOINC Manager might be in order.

Conan
ID: 26324 · Report as offensive     Reply Quote
Profile Tom95134

Send message
Joined: 4 May 07
Posts: 250
Credit: 826,541
RAC: 0
Message 26325 - Posted: 7 Apr 2014, 5:00:22 UTC

If you are running multiple Projects then I suggest that you Suspend the other Projects and see if LHC gets a number of Tasks as a test.

Sometimes other Projects can be a queue hog and LCH ends up on the tail of the train.

I run SETI, T4T, and Einstein. T4T runs itself since it uses the Oracle VM and will only use the number or cores that you have VM running. I allow only 1. Einstein will grab as many cores as it can within the priority system on BOINC. SETI will do the same but since I allow SETI to run Tasks on my GPU I try to make sure it doesn't have a chance to get CPU based Tasks by keeping the CPU queue filled with either LCH or Einstein Tasks. (Using Suspend and Allow Task to modulate how much CPU is available to LHC.)

Yeah, I know this is "cheating" but when LHC has something to run I would like to run as much as possible and not have all the CPU cycles eaten by Einstein or SETI.
ID: 26325 · Report as offensive     Reply Quote
Profile Ananas

Send message
Joined: 17 Jul 05
Posts: 102
Credit: 542,016
RAC: 0
Message 26326 - Posted: 7 Apr 2014, 8:25:40 UTC - in response to Message 26323.  
Last modified: 7 Apr 2014, 8:26:17 UTC

You're running a >10 days CPDN task AND you had some of those underrated LHC tasks lately which messed up the task duration correction factor of your host (don't worry, that will fix itself, but it needs quite a few "normal" results), maybe that's why your BOINC client "thinks" that the work cache contents are sufficient to meet your preferences.

You can try to modify the "... and up to an additional" cache settings in your global preferences, sometimes that helps.

Don't set this value too high, that sometimes has the opposite effect (when the server side scheduler "thinks" that your cache is overloaded).
ID: 26326 · Report as offensive     Reply Quote
Profile Kathryn Tombaugh Weber

Send message
Joined: 12 Sep 11
Posts: 38
Credit: 218,154
RAC: 0
Message 26327 - Posted: 7 Apr 2014, 10:54:20 UTC

I'm running a climateprediction model that takes months. maybe that's it. Maybe I'll just have to let that run and then get back to normal.
ID: 26327 · Report as offensive     Reply Quote
alvin
Avatar

Send message
Joined: 12 Mar 12
Posts: 128
Credit: 20,013,377
RAC: 0
Message 26329 - Posted: 8 Apr 2014, 2:27:26 UTC - in response to Message 26327.  

you could set "switch between tasks" option to more frequency
in this case it will do LHC often and finish it in day timeframe, not year
ID: 26329 · Report as offensive     Reply Quote
Profile MAGIC Quantum Mechanic
Avatar

Send message
Joined: 24 Oct 04
Posts: 853
Credit: 37,649,936
RAC: 25,190
Message 26330 - Posted: 8 Apr 2014, 2:56:05 UTC
Last modified: 8 Apr 2014, 2:58:19 UTC

Well Kathryn your main problem is your computer is just one single core.

http://lhcathomeclassic.cern.ch/sixtrack/show_host_detail.php?hostid=9932712

So you can only do one task at a time and if you are going to run that "climate" task you won't get any tasks anywhere else.

Not to mention that is an old,slow processor from 2008.

If you want any LHC tasks you will have to suspend any other tasks you have running.

And your 2.2 will take about 40 hours to run one of the tasks we have here right now.

SixTrack v446.03 (pni)

It takes my modern Intel Core i7-2670QM CPU @ 2.20GHz about that long but I have 8 cores so it can run 7 LHC tasks and a T4T along with a pair of Einstein cuda's at the same time (laptop I am on right now)

I don't even run these LHC tasks on my quad and 3-core pc's since they all run T4T and Einstein cuda BRP X2 and it runs better and faster leaving a free core.
Volunteer Mad Scientist For Life
ID: 26330 · Report as offensive     Reply Quote
Profile Kathryn Tombaugh Weber

Send message
Joined: 12 Sep 11
Posts: 38
Credit: 218,154
RAC: 0
Message 26331 - Posted: 8 Apr 2014, 3:35:06 UTC

I have now aborted the other two tasks, much to my regret, and reinstalled the latest version of BOINC with that header thing. forced update several times and now have zero tasks.

Should I now commit suicide because I'm such a lousy BOINCer?
ID: 26331 · Report as offensive     Reply Quote
Profile Conan
Avatar

Send message
Joined: 6 Jul 06
Posts: 107
Credit: 511,942
RAC: 0
Message 26332 - Posted: 8 Apr 2014, 4:39:25 UTC
Last modified: 8 Apr 2014, 4:44:55 UTC

The system still shows you having those 2 tasks from the 3/4/14 with a deadline of 11/4/14.
So if you have aborted them the system does not know about it.
Perhaps resetting the LHC project might help in this case, as that should get rid of old files and download some new ones.

It is something to try anyway.

Suicide is not an option. Smiling is far better as it makes others wonder what you have been up too.

Conan
ID: 26332 · Report as offensive     Reply Quote
alvin
Avatar

Send message
Joined: 12 Mar 12
Posts: 128
Credit: 20,013,377
RAC: 0
Message 26334 - Posted: 8 Apr 2014, 4:58:33 UTC - in response to Message 26331.  
Last modified: 8 Apr 2014, 5:13:49 UTC

I have now aborted the other two tasks, much to my regret, and reinstalled the latest version of BOINC with that header thing. forced update several times and now have zero tasks.

Should I now commit suicide because I'm such a lousy BOINCer?

the best for you was to put "no new tasks" for projects but leave old tasks to finish cause you've done a lot before.
anyway just manage your tasks this way to get result you want: suspend, no new tasks, priority and switch between tasks frequesncy will make sense
ID: 26334 · Report as offensive     Reply Quote
Profile Kathryn Tombaugh Weber

Send message
Joined: 12 Sep 11
Posts: 38
Credit: 218,154
RAC: 0
Message 26337 - Posted: 8 Apr 2014, 17:56:22 UTC

ah, I see. Resetting LHC worked. Always I panic too soon.
ID: 26337 · Report as offensive     Reply Quote

Message boards : Number crunching : Not getting tasks. Probably need help


©2019 CERN