We essentially stopped the "O1 Engineering run", i.e aren't generating new workunits anymore.
Instead we will continue the previously suspended "O2AS" run as a GW run. The current GPU App will not give much benefit in that setup, so "O2AS" will (for now) continue to be CPU-only.
If all goes as planned we will start with the "O1OD1 injection run" on GPUs.
Which applications do we check-off under our project preferences to get these work units?
We essentially stopped the "O1 Engineering run", i.e aren't generating new workunits anymore.
Instead we will continue the previously suspended "O2AS" run as a GW run. The current GPU App will not give much benefit in that setup, so "O2AS" will (for now) continue to be CPU-only.
If all goes as planned we will start with the "O1OD1 injection run" on GPUs.
Which applications do we check-off under our project preferences to get these work units?
Thanks.
I have Gamma-ray pulsar binary search #1 (GPU) selected and I am getting GPU work on my Nvidia 1070 TI. For many tasks are running around the 12 minute mark. I hope this is helpful
We essentially stopped the "O1 Engineering run", i.e aren't generating new workunits anymore.
Instead we will continue the previously suspended "O2AS" run as a GW run. The current GPU App will not give much benefit in that setup, so "O2AS" will (for now) continue to be CPU-only.
If all goes as planned we will start with the "O1OD1 injection run" on GPUs.
Which applications do we check-off under our project preferences to get these work units?
Thanks.
Theoretically, O2AS should be "Continuous Gravitational Wave search O2 All-Sky"; but looking at the numbers on the server status page it doesn't appear that any more work is available. I've had fermi CPU and O1 Engineering off for 2+ days and haven't gotten any. Either only a little was left at the time of Bernd's notice, the work generators failed, or something happened and they intentionally stopped it but forgot to post an update. Since O1 engineering does appear to be getting more work generated I assume that whatever happened with O2 the project staff are aware and turned the engineering run back on to keep a supply of work available.
There's nothing that looks like O1OD1 injection run listed among the apps on the settings page, so I assume it's not running yet.
Which applications do we check-off under our project preferences to get these work units?
The O2AS (All-Sky) for CPU tasks. 12 hours ago the WU generator showed as "Running", as it does now. There were 4 tasks ready to send then, there are 133 now. Something is happening so you should be able to get that work now. No idea why generating tasks is taking so long. In the last 12 hours, the 'in progress' tasks has increased by more than 1,000 (from memory) so somebody is getting that work. To state the obvious, they seem to be getting distributed as fast as they are being created, more or less
The injection run (for GPUs) was mentioned in passing with the proviso "if all goes as planned". There was no time frame given. To me this means quite some work to do and probably not all that soon. While waiting, just look for a new check box entry for it in your preferences and for it to be listed on the server status page and the apps page. When all the hurdles for it are overcome, I would expect to see an announcement.
DanNeely wrote:
Since O1 engineering does appear to be getting more work generated I assume that whatever happened with O2 the project staff are aware and turned the engineering run back on to keep a supply of work available.
The O1OD1E run has not (at the moment) been "turned back on". The WU generator is "Disabled" and the remaining stockpile of tasks is slowly being distributed. There was an unusually large number 'pre-generated' at the time of the "no more new WUs" announcement. My guess is that they deliberately left a large stock of work in case of delays in getting the 'new plan' up to speed. If things are 'ready to roll' soon, they may well withdraw any remaining WUs that haven't as yet been distributed. I don't imagine they would want to overly delay the resumption of O2AS because of leftover O1OD1E tasks. I presume analysis of Observation run #2 data is preferable to continuing with the O1 data if the former is of better 'quality'.
Remember that the O2AS run was suddenly suspended last year over some sort of concern about the data. I seem to recall something about 'de-jittering' being mentioned. My guess is that the data has either been 'corrected' or shown not to be a problem. I imagine that's probably why there was this sudden change in plan.
We just need to be patient. "It will be ready when it's ready" is preferable to "Try this and see if it works" :-).
I presume the Windows software for the O2AS is 1.01?
That's what was being used last year when the search was suspended. As the search seems to be slowly progressing now, and there is no new version listed, it looks like the reason for suspending the search was nothing to do with the app itself. Anyone with tasks would know for sure if the app had changed. Sometimes changes happen before there is time to announce them :-).
Which applications do we check-off under our project preferences to get these work units?
The O2AS (All-Sky) for CPU tasks. 12 hours ago the WU generator showed as "Running", as it does now. There were 4 tasks ready to send then, there are 133 now. Something is happening so you should be able to get that work now. No idea why generating tasks is taking so long. In the last 12 hours, the 'in progress' tasks has increased by more than 1,000 (from memory) so somebody is getting that work. To state the obvious, they seem to be getting distributed as fast as they are being created, more or less
Somebody might be; but despite being locked out of E@H's other CPU projects none of my boxes have gotten any even as their CPU caches wound down toward zero until I turned the engineering run back on to keep them from running out.
Maybe they're running them internally on Atlas as a trial to make sure everything is in order for a restart. They might even be testing a new version of the app as well. I've been cleaning up some of the O1OD1E stuff so don't have O2AS selected.
I guess we'll get an announcement when they're generally available.
Given the lack of new gravity waves, which is my main interest in this project. I will have to make a decision. Do I allow my Einstein host fall back on it's back up project or restart looking for pulsars and watch my RAC grow?
If this is the biggest problem I have my life is pretty good.
I've been running O2AS over the weekend and I had no trouble getting enough work. My preferences are CPU only, O2AS only, no betas.
So far all tasks are double-checked by Atlas. I'm sure this will change soon and the search will probably speed up then.
Bernd Machenschalk wrote:We
)
Which applications do we check-off under our project preferences to get these work units?
Thanks.
rbpeake wrote:Bernd
)
I have Gamma-ray pulsar binary search #1 (GPU) selected and I am getting GPU work on my Nvidia 1070 TI. For many tasks are running around the 12 minute mark. I hope this is helpful
rbpeake wrote:Bernd
)
Theoretically, O2AS should be "Continuous Gravitational Wave search O2 All-Sky"; but looking at the numbers on the server status page it doesn't appear that any more work is available. I've had fermi CPU and O1 Engineering off for 2+ days and haven't gotten any. Either only a little was left at the time of Bernd's notice, the work generators failed, or something happened and they intentionally stopped it but forgot to post an update. Since O1 engineering does appear to be getting more work generated I assume that whatever happened with O2 the project staff are aware and turned the engineering run back on to keep a supply of work available.
There's nothing that looks like O1OD1 injection run listed among the apps on the settings page, so I assume it's not running yet.
rbpeake wrote:Which
)
The O2AS (All-Sky) for CPU tasks. 12 hours ago the WU generator showed as "Running", as it does now. There were 4 tasks ready to send then, there are 133 now. Something is happening so you should be able to get that work now. No idea why generating tasks is taking so long. In the last 12 hours, the 'in progress' tasks has increased by more than 1,000 (from memory) so somebody is getting that work. To state the obvious, they seem to be getting distributed as fast as they are being created, more or less
The injection run (for GPUs) was mentioned in passing with the proviso "if all goes as planned". There was no time frame given. To me this means quite some work to do and probably not all that soon. While waiting, just look for a new check box entry for it in your preferences and for it to be listed on the server status page and the apps page. When all the hurdles for it are overcome, I would expect to see an announcement.
The O1OD1E run has not (at the moment) been "turned back on". The WU generator is "Disabled" and the remaining stockpile of tasks is slowly being distributed. There was an unusually large number 'pre-generated' at the time of the "no more new WUs" announcement. My guess is that they deliberately left a large stock of work in case of delays in getting the 'new plan' up to speed. If things are 'ready to roll' soon, they may well withdraw any remaining WUs that haven't as yet been distributed. I don't imagine they would want to overly delay the resumption of O2AS because of leftover O1OD1E tasks. I presume analysis of Observation run #2 data is preferable to continuing with the O1 data if the former is of better 'quality'.
Remember that the O2AS run was suddenly suspended last year over some sort of concern about the data. I seem to recall something about 'de-jittering' being mentioned. My guess is that the data has either been 'corrected' or shown not to be a problem. I imagine that's probably why there was this sudden change in plan.
We just need to be patient. "It will be ready when it's ready" is preferable to "Try this and see if it works" :-).
Cheers,
Gary.
Thank you all for the great
)
Thank you all for the great information!
I presume the Windows software for the O2AS is 1.01?
Thanks!
rbpeake wrote:I presume the
)
That's what was being used last year when the search was suspended. As the search seems to be slowly progressing now, and there is no new version listed, it looks like the reason for suspending the search was nothing to do with the app itself. Anyone with tasks would know for sure if the app had changed. Sometimes changes happen before there is time to announce them :-).
Cheers,
Gary.
Gary Roberts wrote:rbpeake
)
Somebody might be; but despite being locked out of E@H's other CPU projects none of my boxes have gotten any even as their CPU caches wound down toward zero until I turned the engineering run back on to keep them from running out.
Maybe they're running them
)
Maybe they're running them internally on Atlas as a trial to make sure everything is in order for a restart. They might even be testing a new version of the app as well. I've been cleaning up some of the O1OD1E stuff so don't have O2AS selected.
I guess we'll get an announcement when they're generally available.
Cheers,
Gary.
Given the lack of new gravity
)
Given the lack of new gravity waves, which is my main interest in this project. I will have to make a decision. Do I allow my Einstein host fall back on it's back up project or restart looking for pulsars and watch my RAC grow?
If this is the biggest problem I have my life is pretty good.
I've been running O2AS over
)
I've been running O2AS over the weekend and I had no trouble getting enough work. My preferences are CPU only, O2AS only, no betas.
So far all tasks are double-checked by Atlas. I'm sure this will change soon and the search will probably speed up then.