It seems to be because of the transfer from BRP3 to BRP4. BRP 4 tasks are not yet shipped permanently, only once or twice per day. Perhaps the work generators are not quick enough. I'm sure the developers will take a look at tomorrow.
It seems to be because of the transfer from BRP3 to BRP4. BRP 4 tasks are not yet shipped permanently, only once or twice per day. Perhaps the work generators are not quick enough. I'm sure the developers will take a look at tomorrow.
I think it would be helpful, if of the 20 or so work generators some more were not disabled. If you look at the server status page, you see this:
BRP3 work generator daemon #0 einstein-wug Disabled
BRP3 work generator task #0 einstein-wug Disabled
BRP4 work generator #0 einstein-wug Running
BRP4 work generator #1 einstein-wug Running
BRP4 work generator #2 einstein-wug Disabled
BRP4 work generator #3 einstein-wug Disabled
BRP4 work generator #4 einstein-wug Disabled
BRP4 work generator #5 einstein-wug Disabled
BRP4 work generator #6 einstein-wug Disabled
BRP4 work generator #7 einstein-wug Disabled
BRP4 work generator #8 einstein-wug Disabled
BRP4 work generator #9 einstein-wug Disabled
BRP4 work generator #10 einstein-wug Disabled
BRP4 work generator #11 einstein-wug Disabled
BRP4 work generator #12 einstein-wug Disabled
BRP4 work generator #13 einstein-wug Disabled
BRP4 work generator #14 einstein-wug Disabled
BRP4 work generator #15 einstein-wug Disabled
BRP4 work generator #16 einstein-wug Disabled
BRP4 work generator #17 einstein-wug Disabled
BRP4 work generator #18 einstein-wug Disabled
BRP4 work generator #19 einstein-wug Disabled
It seems to be because of the transfer from BRP3 to BRP4. BRP 4 tasks are not yet shipped permanently, only once or twice per day. Perhaps the work generators are not quick enough. I'm sure the developers will take a look at tomorrow.
I think it would be helpful, if of the 20 or so work generators some more were not disabled. If you look at the server status page, you see this:
Eleven of them were enabled this morning, shortly after Bernd wrote:
Quote:
Workunit generation appears to be much slower (5x) for BRP4 compared to BRP3.
My guess is that the multiple-daemon list was an attempted weekend workround for the slow code (it didn't work, from what I could see of the rate of change of 'Tasks total'), and that they'll try to find the bug in the code and fix it properly when they're back in the lab sometime this week.
I too can't seem to get any BRP4 GPU work since they finished BRP3. What's up? The Tasks to Send for both BRP3 and 4 are always at 0 when I check the server status. This is what my log always shows when I contact E@H.
Quote:
7/26/2011 8:15:01 PM | Einstein@Home | Sending scheduler request: To fetch work.
7/26/2011 8:15:01 PM | Einstein@Home | Requesting new tasks for NVIDIA GPU
7/26/2011 8:15:02 PM | Einstein@Home | Scheduler request completed: got 0 new tasks
7/26/2011 8:15:02 PM | Einstein@Home | To get more Einstein@Home work, finish current work, stop BOINC, remove app_info.xml file, and restart.
7/26/2011 8:15:02 PM | Einstein@Home | No work sent
7/26/2011 8:15:02 PM | Einstein@Home | No work is available for Binary Radio Pulsar Search
7/26/2011 8:15:02 PM | Einstein@Home | No work is available for Gravitational Wave S6 GC search
7/26/2011 8:15:02 PM | Einstein@Home | Your app_info.xml file doesn't have a version of Gravitational Wave S6 GC search .
What I can't understand is why it always states the business about Gravitational Wave S6 GC search. I also continue to get a message in the BM about not having a suitable N-body application. I shouldn't. I only have the project set up to provide BRP GPU work. No CPU work at all. So the only application I should have running is the 1.07 BRP OpenCL app. I have updated my app_info to reflect the new BRP4 name change and the app version change to 1.00.
I too can't seem to get any BRP4 GPU work since they finished BRP3. What's up? The Tasks to Send for both BRP3 and 4 are always at 0 when I check the server status. This is what my log always shows when I contact E@H.
check your preferences page here - you need to opt in for it.
I had checked my preferences when I setup for BRP4. At that time there was only one option for BRP work. Now I see that they have doubled up the same BRP work. I assume the first statement is the original for BRP3 work and the second interation is for BRP4 work even though there is no identification as to which statement belongs to which BRP survey. Also I notice that I have a greyed out checkmark for the Gravitational Wave S6 GC work. That probably explains the error messages. I never opted for this work. I can't change the value by editing, it remains a greyed out checkmark in the box. I can't uncheck it. Can an administrator correct this?
Also I notice that I have a greyed out checkmark for the Gravitational Wave S6 GC work. That probably explains the error messages. I never opted for this work. I can't change the value by editing, it remains a greyed out checkmark in the box. I can't uncheck it. Can an administrator correct this?
Unfortunately for anyone who's only interested in GPU crunching, the the S6GC being non-disablable is by design. The money that's funding the whole E@H program comes from that research effort, with the other search efforts just riders. I'm not sure if they're explicitly not allowed to enable that checkbox, or if they're only unwilling to do so because it would just look bad with the people paying the bills. Either way the end result is the same for the rest of us.
What you can do is to use an app_info.xml file to tell the server that your computer is only able to do BRP4-cuda tasks. The caveat is that you'll need to manually update the file each time a new app comes out, and your message log will be cluttered with messages from the server saying you could get more work (S6GC, etc) if you removed the file.
There is a simple setting on the prefs screen to just do GPU work on E@H. Yes, if you enable the CPU, you will get S6, but you don't have to do anything with your app_info to only do GPU units.
![[B^S] Dimitrij](https://einsteinathome.org/sites/default/files/imagefield_default_images/default-avatar.png?1347855344 "[B^S] Dimitrij"){kind=avatar}
no more cuda work?
)
It seems to be because of the transfer from BRP3 to BRP4. BRP 4 tasks are not yet shipped permanently, only once or twice per day. Perhaps the work generators are not quick enough. I'm sure the developers will take a look at tomorrow.
thanks, that helps.
)
thanks, that helps.
RE: It seems to be because
)
I think it would be helpful, if of the 20 or so work generators some more were not disabled. If you look at the server status page, you see this:
Grüße vom Sänger
RE: RE: It seems to be
)
Eleven of them were enabled this morning, shortly after Bernd wrote:
My guess is that the multiple-daemon list was an attempted weekend workround for the slow code (it didn't work, from what I could see of the rate of change of 'Tasks total'), and that they'll try to find the bug in the code and fix it properly when they're back in the lab sometime this week.
i got some more work, thanks
)
i got some more work, thanks for your quick answers.
I too can't seem to get any
)
I too can't seem to get any BRP4 GPU work since they finished BRP3. What's up? The Tasks to Send for both BRP3 and 4 are always at 0 when I check the server status. This is what my log always shows when I contact E@H.
What I can't understand is why it always states the business about Gravitational Wave S6 GC search. I also continue to get a message in the BM about not having a suitable N-body application. I shouldn't. I only have the project set up to provide BRP GPU work. No CPU work at all. So the only application I should have running is the 1.07 BRP OpenCL app. I have updated my app_info to reflect the new BRP4 name change and the app version change to 1.00.
RE: I too can't seem to get
)
check your preferences page here - you need to opt in for it.
I had checked my preferences
)
I had checked my preferences when I setup for BRP4. At that time there was only one option for BRP work. Now I see that they have doubled up the same BRP work. I assume the first statement is the original for BRP3 work and the second interation is for BRP4 work even though there is no identification as to which statement belongs to which BRP survey. Also I notice that I have a greyed out checkmark for the Gravitational Wave S6 GC work. That probably explains the error messages. I never opted for this work. I can't change the value by editing, it remains a greyed out checkmark in the box. I can't uncheck it. Can an administrator correct this?
Cheers, Keith
RE: Also I notice that I
)
Unfortunately for anyone who's only interested in GPU crunching, the the S6GC being non-disablable is by design. The money that's funding the whole E@H program comes from that research effort, with the other search efforts just riders. I'm not sure if they're explicitly not allowed to enable that checkbox, or if they're only unwilling to do so because it would just look bad with the people paying the bills. Either way the end result is the same for the rest of us.
What you can do is to use an app_info.xml file to tell the server that your computer is only able to do BRP4-cuda tasks. The caveat is that you'll need to manually update the file each time a new app comes out, and your message log will be cluttered with messages from the server saying you could get more work (S6GC, etc) if you removed the file.
There is a simple setting on
)
There is a simple setting on the prefs screen to just do GPU work on E@H. Yes, if you enable the CPU, you will get S6, but you don't have to do anything with your app_info to only do GPU units.