Yes, if the computer shows up to BOINC as to be the same as before, there should be no problem. Just restore the BOINC data directory and make sure that the BOINC installer points to that place afterwards.
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
I've just done a fresh windows install on a new drive. Can I migrate my project settings/in progress work units over to the new image?
You may need/want to reinstall after you move the dir over. BOINC now fiddles with permissions so a re-install is usually required to set these correctly.
I would strongly suggest till you proof the new install that you disconnect the network so that if bad things happen the tasks don't get reported until you can get it sorted out. So, disconnect the network, copy the folder in place, install BOINC (the version you were using, no need to complicate matters), start BOINC and make sure it runs correctly for a bit ... then enable the network (plug teh cable back in).
If bad things happen you can just copy the folder back over and start again ...
Yes, if the computer shows up to BOINC as to be the same as before, there should be no problem. Just restore the BOINC data directory and make sure that the BOINC installer points to that place afterwards.
Gruß,
Gundolf
by this do you mean that it's not showing as two unique computers under the my account webpage?
I accomplished the merge successfully (including merging 2 in progress GPUgrid tasks that I had running in the new version along with placeholder tasks on the CPU).
I've got one issue that might just be coincidental timing. After the merge EAH is complaining about my app_info file (for the 3.09 ABP executable) and refusing to issue any new work. Can I safely nuke the file without running out my cache?
I think my previous post can be disregarded. It appears that it's only asking for GPUwork and bouncing because I don't have an app info for that. Until the EAH gets significantly faster I'm going to keep something else on my GPU (currently GPUgrid).
10/5/2009 12:04:04 AM Einstein@Home update requested by user
10/5/2009 12:04:05 AM Einstein@Home Sending scheduler request: Requested by user.
10/5/2009 12:04:05 AM Einstein@Home Reporting 2 completed tasks, requesting new tasks for GPU
10/5/2009 12:04:10 AM Einstein@Home Scheduler request completed: got 0 new tasks
10/5/2009 12:04:10 AM Einstein@Home Message from server: To get more Einstein@Home work, finish current work, stop BOINC, remove app_info.xml file, and restart.
10/5/2009 12:04:10 AM Einstein@Home Message from server: No work sent
I think my previous post can be disregarded. It appears that it's only asking for GPUwork and bouncing because I don't have an app info for that. Until the EAH gets significantly faster I'm going to keep something else on my GPU (currently GPUgrid).
10/5/2009 12:04:04 AM Einstein@Home update requested by user
10/5/2009 12:04:05 AM Einstein@Home Sending scheduler request: Requested by user.
10/5/2009 12:04:05 AM Einstein@Home Reporting 2 completed tasks, requesting new tasks for GPU
10/5/2009 12:04:10 AM Einstein@Home Scheduler request completed: got 0 new tasks
10/5/2009 12:04:10 AM Einstein@Home Message from server: To get more Einstein@Home work, finish current work, stop BOINC, remove app_info.xml file, and restart.
10/5/2009 12:04:10 AM Einstein@Home Message from server: No work sent
You are still seeing the artifact where the BOINC client asks for the wrong classes of work until it gives up or the server sets the back off high enough to make it stop for some longer period of time.
You are still seeing the artifact where the BOINC client asks for the wrong classes of work until it gives up or the server sets the back off high enough to make it stop for some longer period of time.
*sigh* Does this mean I have to switch client versions to get more CPU work?
You are still seeing the artifact where the BOINC client asks for the wrong classes of work until it gives up or the server sets the back off high enough to make it stop for some longer period of time.
*sigh* Does this mean I have to switch client versions to get more CPU work?
I'm using 6.6.38
It should also be asking for CPU work. The problem is that I cannot recall which versions do what other than we have been sliding into a ditch for some time (since 6.6.1) and there are all these weird interactions.
If the web site is still set to issue CPU work the client should be asking for it. The bad news is that many of the clients will also ask for GPU work when they should not ... or ask for CPU work when they should not (like at GPU Grid where there is no CPU work) ... eventually it is supposed to ask for the right class and move on ... the problem is that since 6.5.0 the way that the BOINC client balances work from project to project has also been allowed to "drift" so that it does not always ask as we might expect.
Thus my complaints about the way Resource Share is handled.
My only suggestion is to suspend or NNT all other projects, increase the cache size temporarily and make sure that there is work here and see if you can get some.
boinc migration
)
Yes, if the computer shows up to BOINC as to be the same as before, there should be no problem. Just restore the BOINC data directory and make sure that the BOINC installer points to that place afterwards.
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
RE: I've just done a fresh
)
You may need/want to reinstall after you move the dir over. BOINC now fiddles with permissions so a re-install is usually required to set these correctly.
I would strongly suggest till you proof the new install that you disconnect the network so that if bad things happen the tasks don't get reported until you can get it sorted out. So, disconnect the network, copy the folder in place, install BOINC (the version you were using, no need to complicate matters), start BOINC and make sure it runs correctly for a bit ... then enable the network (plug teh cable back in).
If bad things happen you can just copy the folder back over and start again ...
RE: Yes, if the computer
)
by this do you mean that it's not showing as two unique computers under the my account webpage?
RE: by this do you mean
)
yep.
But if you'll see that, the downloaded tasks will have been aborted already. So, you should follow Paul D. Buck's procedure.
Gruß,
Gundolf
Computer sind nicht alles im Leben. (Kleiner Scherz)
I accomplished the merge
)
I accomplished the merge successfully (including merging 2 in progress GPUgrid tasks that I had running in the new version along with placeholder tasks on the CPU).
I've got one issue that might just be coincidental timing. After the merge EAH is complaining about my app_info file (for the 3.09 ABP executable) and refusing to issue any new work. Can I safely nuke the file without running out my cache?
I think my previous post can
)
I think my previous post can be disregarded. It appears that it's only asking for GPUwork and bouncing because I don't have an app info for that. Until the EAH gets significantly faster I'm going to keep something else on my GPU (currently GPUgrid).
10/5/2009 12:04:04 AM Einstein@Home update requested by user
10/5/2009 12:04:05 AM Einstein@Home Sending scheduler request: Requested by user.
10/5/2009 12:04:05 AM Einstein@Home Reporting 2 completed tasks, requesting new tasks for GPU
10/5/2009 12:04:10 AM Einstein@Home Scheduler request completed: got 0 new tasks
10/5/2009 12:04:10 AM Einstein@Home Message from server: To get more Einstein@Home work, finish current work, stop BOINC, remove app_info.xml file, and restart.
10/5/2009 12:04:10 AM Einstein@Home Message from server: No work sent
RE: I think my previous
)
You are still seeing the artifact where the BOINC client asks for the wrong classes of work until it gives up or the server sets the back off high enough to make it stop for some longer period of time.
RE: You are still seeing
)
*sigh* Does this mean I have to switch client versions to get more CPU work?
I'm using 6.6.38
RE: RE: You are still
)
It should also be asking for CPU work. The problem is that I cannot recall which versions do what other than we have been sliding into a ditch for some time (since 6.6.1) and there are all these weird interactions.
If the web site is still set to issue CPU work the client should be asking for it. The bad news is that many of the clients will also ask for GPU work when they should not ... or ask for CPU work when they should not (like at GPU Grid where there is no CPU work) ... eventually it is supposed to ask for the right class and move on ... the problem is that since 6.5.0 the way that the BOINC client balances work from project to project has also been allowed to "drift" so that it does not always ask as we might expect.
Thus my complaints about the way Resource Share is handled.
My only suggestion is to suspend or NNT all other projects, increase the cache size temporarily and make sure that there is work here and see if you can get some.
Patience is key during the
)
Patience is key during the cache reallocation process to!
Overclock with the MSI G31M3-L and Intel E8600 3.33Ghz
Intel D865GLC Socket 478 Motherboard Review
Overclock your ASUS 1005HA netbook and crunch more