Gravitational Wave search O2 All-Sky search ("O2AS20-500")

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4266
Credit: 244924143
RAC: 16679
Topic 214401

Following the "tuning" run, we have just launched the actual analysis run for the full frequency range. This will be E@H's first analysis of data from the second Observation Run of the LIGO detectors, and search the whole sky, covering the frequencies 20-500Hz.

We'll leave the workunit generation run somewhat throttled over the weekend and ramp it up next week (unless we see signs of serious trouble).

Also validation will be suspended until we got back the first results and are back ourselves at our desks to monitor the validator.

BM

Logforme
Logforme
Joined: 13 Aug 10
Posts: 332
Credit: 1714373961
RAC: 0

Can we get a 14 day deadline

Can we get a 14 day deadline please? Saves me a lot of aborting

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4266
Credit: 244924143
RAC: 16679

This will also be changed

This will also be changed after the weekend.

BM

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3562358667
RAC: 1580

almost 1/3rd of the tasks

almost 1/3rd of the tasks reported so far are failing.  This doesn't look promising...

Zalster
Zalster
Joined: 26 Nov 13
Posts: 3117
Credit: 4050672230
RAC: 0

Actually, if you look at the

Actually, if you look at the results. None have validated, invalidated or been diagnosis as inconclusive.   1022 are pending, 408 have failed.  So we will need to wait and see what happens 

rbpeake
rbpeake
Joined: 18 Jan 05
Posts: 266
Credit: 967950387
RAC: 1205585

It looks like the percentage

It looks like the percentage of successes is increasing.   I know personally I have not had any failures .  It will be interesting to see if that trend continues.

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3562358667
RAC: 1580

Down to around 10% now; which

Down to around 10% now; which suggests that the fails are happening around app start.  A bad batch of WUs, or one app/system combo not working perhaps.  It'll be interesting to see what happens when the devs start looking Monday morning.

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4266
Credit: 244924143
RAC: 16679

Errors are always what's

Errors are always what's returned first, I don't care much about these and ignore them at least for the duration it takes to successfully complete & report a task (here: ~12h), unless some particular error occurs on a large variety of hosts, which may indicate a real problem.

Currently we are down to a ('computing") error rate below 2%.

BM

Zalster
Zalster
Joined: 26 Nov 13
Posts: 3117
Credit: 4050672230
RAC: 0

Thanks Bernd for the update. 

Thanks Bernd for the update.   Looks like just a hair over 9 hours per work unit on my computer. Would add a second to help out but having internet problems at the moment.

Conan
Conan
Joined: 19 Jun 05
Posts: 172
Credit: 7099171
RAC: 2415

No errors at all for me so

No errors at all for me so far, running well. Tuning run was good also.

 

Conan

Jim1348
Jim1348
Joined: 19 Jan 06
Posts: 463
Credit: 257957147
RAC: 0

Mine are running well and

Mine are running well and completing, but with a hitch.  After starting up and running for 8 minutes, the estimated time remaining jumps up to a very high value, typically 7 to 9 days and briefly even more.  Then, after running for maybe half an hour or so, the time estimate returns to a reasonable value.  (They typically run for about 11 to 12 1/2 hours on this i7-4790 running Ubuntu 16.04.)

However, this brief jump causes most of them to run "high priority", and creates havoc with the scheduling of other projects that are running (or trying to run) at the time.  This has been going on for a couple of days, and I hoped it would correct itself, but it has not yet.

 

Comment viewing options

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