Forum > Discussion Forum

No tasks at beta?

(1/2) > >>

Raistmer:
29/11/2011 14:27:29   SETI@home Beta Test   update requested by user
29/11/2011 14:27:32   SETI@home Beta Test   Sending scheduler request: Requested by user.
29/11/2011 14:27:32   SETI@home Beta Test   Requesting new tasks for GPU
29/11/2011 14:27:34   SETI@home Beta Test   Scheduler request completed: got 0 new tasks
29/11/2011 14:27:34   SETI@home Beta Test   Message from server: No tasks sent

Is it normal ?

Urs Echternacht:
Di 29 Nov 2011 10:07:40 CET   SETI@home Beta Test   Scheduler request completed: got 1 new tasks


Not over here.

Raistmer:

--- Quote from: Raistmer on 29 Nov 2011, 05:28:54 am ---
29/11/2011 14:27:34   SETI@home Beta Test   Message from server: No tasks sent


--- End quote ---

This  message  suspicious. Why no tasks sen if client requesed ?...

Urs Echternacht:
Because queue was empty at time of request. Also there have been connection drop outs throughout the last half hour or so. Maybe later there are wus for your hw-type available again.

Richard Haselgrove:

--- Quote from: Raistmer on 29 Nov 2011, 05:28:54 am ---29/11/2011 14:27:32    SETI@home Beta Test  Requesting new tasks for GPU
29/11/2011 14:27:34    SETI@home Beta Test  Message from server: No tasks sent

--- End quote ---


--- Quote from: Raistmer on 29 Nov 2011, 05:33:07 am ---This  message  suspicious. Why no tasks sen if client requesed ?...

--- End quote ---

'No tasks sent' at Beta means exactly the same as it does at the main project - at the moment of your request (but not necessarily before or after), there were no tasks in the 100-job feeder cache that were suitable to fulfil your request. You were asking for GPU tasks only - without debug logging, such as [sched_op], I have no way of knowing whether that was an ATI request or a NVIDIA request, MB or AP.

If NV, then it may be relevent that *every single task* in my Beta/CPU queue at the moment is tagged .vlar: since they come from the 12jl11aa tape, there may be specific scientific reasons why Arecibo was operating in 'static observe' mode that day - was that your suspicion?

Otherwise, if it's merely a renewal of your request that VLAR tasks be issued to NV cards at Beta for testing purposes, then I think you'll have to send that message to Berkeley - I suggest you phrase it as a request, with reasoning, rather than anything that could be mis-interpreted as a demand.

Navigation

[0] Message Index

[#] Next page