wasted time

Mark Reiss
Mark Reiss
Joined: 22 Jan 05
Posts: 16
Credit: 1996687
RAC: 0
Topic 197798

this task:
PB0056_00821_88_0 202720988 6 Nov 2014 17:07:22 UTC 15 Nov 2014 0:31:11 UTC Error while computing 554,121.40 4,593.92 9.42 --- Binary Radio Pulsar Search (Perseus Arm Survey) v1.39 (BRP5-opencl-ati)

took 6.413 days & completed with an error - what a waste - I won't run another 1 of these on this computer!

Claggy
Claggy
Joined: 29 Dec 06
Posts: 560
Credit: 2694028
RAC: 0

wasted time

Quote:

this task:
PB0056_00821_88_0 202720988 6 Nov 2014 17:07:22 UTC 15 Nov 2014 0:31:11 UTC Error while computing 554,121.40 4,593.92 9.42 --- Binary Radio Pulsar Search (Perseus Arm Survey) v1.39 (BRP5-opencl-ati)

took 6.413 days & completed with an error - what a waste - I won't run another 1 of these on this computer!


Boinc aborted the task with 'Maximum elapsed time exceeded' as it took 10 times longer than it should, Have you freed a CPU core to stop the GPU apps being slowed down?

Quote:

7.2.42

Maximum elapsed time exceeded

That host is running a particularly old CAL driver of 1.4.1417, this equates to Cat 11.6 and should have OpenCL runtime of 650.9 (assuming they haven't become mismatched), this was an SDK 2.4 driver/runtime:

ATI Driver Version Cheat Sheet

It is quite possible that your host has a too old driver/OpenCL runtime for the projects apps, They may have been built with SDK's as recent as SDK 2.7 or 2.8

Claggy

Logforme
Logforme
Joined: 13 Aug 10
Posts: 332
Credit: 1714373961
RAC: 0

Very rarely I get BRP5 tasks

Very rarely I get BRP5 tasks that runs "forever". I abort them when I check the boinc manager and see tasks taking waaaay too long. Like this one:
http://einsteinathome.org/workunit/202984243
Only happens on my ATI card, never seen it on a Nvidia card.

Gavin
Gavin
Joined: 21 Sep 10
Posts: 191
Credit: 40644234415
RAC: 140758

I also see this from time to

I also see this from time to time, there is no need to abort the stuck/overly long running task as a simple suspend/resume of the task will get it going again from the last checkpoint and in my case they will always then run to completion and validate.

I concur that this only happens with (my) AMD cards and not NVidia, don't know why but if you can keep a regular eye on your hosts you can spot the problematic tasks without wasting too much time and as I have said a simple suspend/resume sorts it out.

Gav.

ExtraTerrestrial Apes
ExtraTerrestria...
Joined: 10 Nov 04
Posts: 770
Credit: 540147475
RAC: 134944

Mark, that host is very slow.

Mark, that host is very slow. But it shouldn't be this slow, so I agree with Claggy: try to update your driver. That "E-300 APU" is just a system-on-chip with the codename "Brazos", so you should be able to get drivers from AMD just fine, even if your laptop/small box manufacturer doesn't provide any update. Which would not be unusual, normally they'd rather sell you a new system than make your current one work properly.

MrS

Scanning for our furry friends since Jan 2002

Comment viewing options

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