problem getting new work

Desperado
Desperado
Joined: 23 Feb 05
Posts: 4
Credit: 10574
RAC: 0
Topic 192890

haven't been using Boinc much for the last month or so...now when I try to get new work from Einstein by clicking on update button, I get the following msg:

6/22/2007 6:12:17 AM|Einstein@Home|Sending scheduler request: Requested by user

6/22/2007 6:12:17 AM|Einstein@Home|(not requesting new work or reporting completed tasks)

6/22/2007 6:12:22 AM|Einstein@Home|Scheduler RPC succeeded [server version 509]

can somebody explain what's going on? Thanks

Michael Karlinsky
Michael Karlinsky
Joined: 22 Jan 05
Posts: 888
Credit: 23502182
RAC: 0

problem getting new work

Quote:

haven't been using Boinc much for the last month or so...now when I try to get new work from Einstein by clicking on update button, I get the following msg:

6/22/2007 6:12:17 AM|Einstein@Home|Sending scheduler request: Requested by user

6/22/2007 6:12:17 AM|Einstein@Home|(not requesting new work or reporting completed tasks)

6/22/2007 6:12:22 AM|Einstein@Home|Scheduler RPC succeeded [server version 509]

can somebody explain what's going on? Thanks

Wow, this seems to be the question of the week! Please see thread_1, thread_2 or thread_3

Michael

Desperado
Desperado
Joined: 23 Feb 05
Posts: 4
Credit: 10574
RAC: 0

ok...I am running Rosetta

Message 68465 in response to message 68464

ok...I am running Rosetta projects as well. When that last SR change (think I said that right) was made, I started handling things manually between Einstein and Rosetta. I never had problems before getting new work while a Rosetta project was running but is that what the problem could be?

Desperado
Desperado
Joined: 23 Feb 05
Posts: 4
Credit: 10574
RAC: 0

I suspended my rosetta task

Message 68466 in response to message 68465

I suspended my rosetta task temporarily and then einstein dled and is working fine.

Thanks,

Alinator
Alinator
Joined: 8 May 05
Posts: 927
Credit: 9352143
RAC: 0

Keep in mind that

Keep in mind that micro-managing BOINC in this fashion is usually not a good idea.

The real source of the apparent issue is currently EAH is a very tight deadline project compared to what it was before. This is due to the workunits being 'tougher' from a runtime POV, the experimental new app being less optimized and therefore slower than S5R1/I was, and the deadline for the work was left at two weeks.

All of this in combination with other host side factors, such as the number of hours you run your machine for per day, other restrictions you put on BOINC for running work when you are using the machine, and the resource shares you have set for the other projects you run can lead BOINC to determine it has to run the EAH result more or less exclusively when it has one in order to make sure it can complete and report it on time. When this happens, BOINC responds by not DL'ing work from EAH for awhile after it completes the result to pay back the other projects for the time it had to 'borrow' to run the EAH result to maintain your resource share.

The problem with forcing it to DL a result when it ordinarily wouldn't is BOINC will have to borrow more time from the other projects to complete that one. So what happens then is once that result finishes, there's just that much more time to be paid back to the other projects and it will take that much longer for BOINC to get around to coming back to EAH on it's own.

HTH,

Alinator

Comment viewing options

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