Is there a credit problem?

Rodrigo Lobos Aleuy
Rodrigo Lobos Aleuy
Joined: 28 Sep 11
Posts: 1
Credit: 25887138
RAC: 0
Topic 220257

Computer crunching but haven-t received credit for several weeks

Is there a problem?

@rodrigolobos

Richie
Richie
Joined: 7 Mar 14
Posts: 656
Credit: 1702989778
RAC: 0

Your computer has been

Your computer has been granted credits for GPU tasks all the time, also today. The oldest CPU tasks that are waiting for validation have been reported 7 days ago. It is normal that tasks may need to wait that long until another computer will return a tasks and validation can happen. You just need to wait and see.

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109406144599
RAC: 35389906

Rodrigo Lobos Aleuy

Rodrigo Lobos Aleuy wrote:
Computer crunching but haven-t received credit for several weeks

Maybe you should check your list of validated tasks on the website before you make such a wildly incorrect statement.  That list has 100 tasks currently, all validated and credited within the last week or so.

Rodrigo Lobos Aleuy wrote:
Is there a problem?

Yes, there is a problem but it's certainly not the one you are wrongly complaining about.  If you look at the top of the page I linked to, you will see (currently) 560 error tasks listed - 28 pages of them, more than half of those in the last week.  I only looked at a couple of pages (starting around 14th December) and I got the impression that many were either aborted or failed because they weren't started by the deadline.

The GW GPU tasks have a 7 day deadline.  I'm guessing that you have way too large a work cache size for that deadline.  The list currently shows a total of 666 tasks (of all types) in progress.  I guess many of those are headed for the trash can as well.  Please reduce your total work cache size to a maximum of 0.1 days until such time as the tasks you have on hand falls to a more reasonable number.  Any time you find your tasks on hand exceeding say 50-100, you must have too large a cache of tasks.

Part of the reason for your problem is that you are trying to support all 4 current CPU and GPU searches.  They have wildly different relationships between estimated and actual crunch times and they are drastically interfering with each other.  The only way that BOINC will be able to cope with all 4 searches is by permanently setting a very small cache size until such time as the project is able to issue work with much better time estimates than is currently available.

Cheers,
Gary.

lazlo
lazlo
Joined: 20 Nov 19
Posts: 9
Credit: 2958620
RAC: 0

Gary,   There are also a

Gary,

 

There are also a lot of tasks that show up as "Aborted" but actually had errors while computing like this one:

 

https://einsteinathome.org/task/893981190

 

Other tasks are listed as having an "Error while computing" but I do not use Windows so I am sure what to make of the output:

 

https://einsteinathome.org/task/904845811 and https://einsteinathome.org/task/904856179

 

 

 

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5842
Credit: 109406144599
RAC: 35389906

lazlo wrote:There are also a

lazlo wrote:
There are also a lot of tasks that show up as "Aborted" but actually had errors while computing like this one ...

If you look at the output for that task ID link, there is some interesting information.

           Created:            31 Oct 2019 7:34:41 UTC
           Sent:                5 Nov 2019 3:18:41 UTC
           Report deadline:    19 Nov 2019 3:18:41 UTC
           Received:           21 Nov 2019 17:25:06 UTC
           Server state:       Over
           Outcome:            Computation error
           Client state:       Aborted by user
           Exit status:        203 (0x000000CB) EXIT_ABORTED_VIA_GUI


Some points can be gleaned from this.  The work is almost 7 weeks old and doesn't relate to current behaviour.  Lots of things have changed since then which is why I mentioned only looking at current stuff (the last week or so).   The work was sent on Nov 5 with a deadline of Nov 19 and the result was returned on Nov 21.  In other words, the task was crunching after the deadline so the OP probably decided to abort it simply because of that fact.  The last line of the excerpt seems to confirm that the task was aborted whilst crunching.

The take home message is the same.  The OP needs to drastically reduce the work cache size so that tasks are not under this sort of deadline pressure.

 The other two tasks are much more recent - Dec 12 - and are CPU tasks not under deadline pressure.  Both show multiple stops/restarts so without knowing what the OP may have been doing with the machine at the time, it's impossible to know the true cause of the failures.  I suspect that if the number of tasks on hand is reduced so there is no longer any deadline pressure, any other issues may also tend to disappear.

Cheers,
Gary.

lazlo
lazlo
Joined: 20 Nov 19
Posts: 9
Credit: 2958620
RAC: 0

Thank you for clarifying what

Thank you for clarifying what I was seeing.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.