Markus,
This WU is definitely not running as it should.
In case uplinger or another tech wants to check anything I suggest that you
1. make sure you have enough work in queue until the next time you can monitor this device; if necessary, increase your cache size as appropriate and wait for WUs to be downloaded before going to the next step
2. suspend this type A WU until you get new instructions or you need to refill the cache (there is no work fetch while there are any suspended tasks).
3. if you need to refill the cache before you get feedback from the techs resume the suspended task while WUs are downloaded, then suspend it again.

If you don't like this suggestion you can obviously decide to kill this WU, because it is most probable that it will never reach a happy end.

Sorry for the trouble.
Jean.

Markus,

I am sorry for my slow response. From my initial looks at what you have commented on, it appears that the work unit is not converging as it should. We encountered one of these in BETA but are not able to prevent from the very few that squeak through. We are going to set the status of the work unit as unloaded and you should get server abort for that work unit.

When I say very few, this is the 2nd one we have seen other than the beta work unit referred to above.

Thanks Jean for bringing this to my attention.

Thanks,
-Uplinger

rembertw,

If you post the stderr from the results status page you may see why it has errored out. I believe what you'll see is a either a memory access error or "ENERGY TOLERANCE" error. Both of which are the same and is a positive negative for the researchers.

-Uplinger