Panic Mode On (106) Server Problems?

Message boards : Number crunching : Panic Mode On (106) Server Problems?
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 15 · 16 · 17 · 18 · 19 · 20 · 21 . . . 29 · Next

AuthorMessage
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874393 - Posted: 22 Jun 2017, 6:49:00 UTC - in response to Message 1874229.  

Yes, it seems the 6.xx.xx clients are immune to the issue. I assume, Wiggo, you do your SoG tuning in the app_info since you don't have access to an app_config with your client?
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874393 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13703
Credit: 208,696,464
RAC: 304
Australia
Message 1874394 - Posted: 22 Jun 2017, 6:50:57 UTC - in response to Message 1874393.  

Yes, it seems the 6.xx.xx clients are immune to the issue. I assume, Wiggo, you do your SoG tuning in the app_info since you don't have access to an app_config with your client?

You can use a configuration file that's referenced in the app_info.xml file, to save the risk of trashing all your work.
Grant
Darwin NT
ID: 1874394 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874396 - Posted: 22 Jun 2017, 7:04:21 UTC - in response to Message 1874394.  

Yes, that is what I meant. Use the standard mb_cmdline_win_x86_SSE3_OpenCL_NV_SoG.txt callout in the app_info for tuning parameters. Wiggo has said all along he is not affected. Curious if anyone has gone backwards from a 7.x.xx client to a 6.x.xx client and seen whether that fixes the problem for them? I've never attempted it because I ran into roadblocks getting a download of an earlier client to not be corrupt. A few people have sent me older clients but I have not been brave enough to try.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874396 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13703
Credit: 208,696,464
RAC: 304
Australia
Message 1874398 - Posted: 22 Jun 2017, 7:12:11 UTC - in response to Message 1874396.  
Last modified: 22 Jun 2017, 7:18:24 UTC

Yes, that is what I meant. Use the standard mb_cmdline_win_x86_SSE3_OpenCL_NV_SoG.txt callout in the app_info for tuning parameters. Wiggo has said all along he is not affected. Curious if anyone has gone backwards from a 7.x.xx client to a 6.x.xx client and seen whether that fixes the problem for them? I've never attempted it because I ran into roadblocks getting a download of an earlier client to not be corrupt. A few people have sent me older clients but I have not been brave enough to try.

I remember that there was one major version upgrade that wasn't easily reversible; going back to the older version required empting the cache (or trashing all that work) and i'm pretty sure it was the v6 to v7 move that's not easily undone.
And I do like being able to use app_config.xml
Grant
Darwin NT
ID: 1874398 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1874400 - Posted: 22 Jun 2017, 7:25:32 UTC - in response to Message 1874390.  

And we're back to random allocation/non allocation of work.

EDIT- at least this time around the application flipping is working.
And GBT is scarce again, only GBT work i'm getting are resends. And that appears to be the case for Arecibo as well; at least for the last 2 requests that got work.


. . This is such a random issue ... I am getting heaps of GBT work, and Arecibo VLARs.

Stephen

<shrug>
ID: 1874400 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13703
Credit: 208,696,464
RAC: 304
Australia
Message 1874408 - Posted: 22 Jun 2017, 9:27:03 UTC - in response to Message 1874390.  

And we're back to random allocation/non allocation of work.

EDIT- at least this time around the application flipping is working.

Well, it worked for a while at least.
:-(
Grant
Darwin NT
ID: 1874408 · Report as offensive
Profile tazzduke
Volunteer tester

Send message
Joined: 15 Sep 07
Posts: 190
Credit: 28,269,068
RAC: 5
Australia
Message 1874429 - Posted: 22 Jun 2017, 12:52:47 UTC

Well its my turn to join the party, this randomness is quirky.

3 machines, 1 x win 7 and 1 x Linux being fed, the other linux box is doing a hunger strike.

Oh well, the universe is bound to do a correction and we will be back on track lol.

Cheerio
ID: 1874429 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1874431 - Posted: 22 Jun 2017, 14:22:02 UTC - in response to Message 1874400.  

And we're back to random allocation/non allocation of work.

EDIT- at least this time around the application flipping is working.
And GBT is scarce again, only GBT work i'm getting are resends. And that appears to be the case for Arecibo as well; at least for the last 2 requests that got work.


. . This is such a random issue ... I am getting heaps of GBT work, and Arecibo VLARs.

Stephen

<shrug>


[update] Now I am getting almost consistent work on one machine while on the other 3 rigs just CPU Arecibo VLARs. This is painful and tedious. I am shutting down those 3 rigs for the night. I will check again in the morning. Someone really needs to work out just what this problem is and start a plan to fix it.
ID: 1874431 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874451 - Posted: 22 Jun 2017, 16:20:37 UTC - in response to Message 1874398.  


And I do like being able to use app_config.xml

That has been my primary reason why I have not attempted the task. I like app_config too. Good to hear it might not be possible without trashing your work if you don't clear them first.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874451 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874452 - Posted: 22 Jun 2017, 16:29:58 UTC

As my usual scenario, the Windows7 machines are having troubles getting work. The Windows10 computer is being constantly fed. Flipping preferences isn't working.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874452 · Report as offensive
Profile tazzduke
Volunteer tester

Send message
Joined: 15 Sep 07
Posts: 190
Credit: 28,269,068
RAC: 5
Australia
Message 1874498 - Posted: 22 Jun 2017, 20:51:44 UTC

Well I spoke to soon last night, no sooner I had just posted, the 3rd machine filled up.

A little bit of craziness is about.

So do we move on and try to figure out the sum of the universe lol.

Happy Crunching.
ID: 1874498 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874538 - Posted: 22 Jun 2017, 22:29:42 UTC - in response to Message 1874498.  

Yes, right when I was going to post that the Win10 machine now was having issues, it filled itself back up. Seems to be par for the course. The schedulers are still taking long lunch breaks.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874538 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874567 - Posted: 23 Jun 2017, 2:13:13 UTC

Sigh, just checked and all machines are down by 75 tasks at least. Wonder if this coincided with AP splitting again. Trying my usual tricks.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874567 · Report as offensive
Stephen "Heretic" Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Sep 12
Posts: 5557
Credit: 192,787,363
RAC: 628
Australia
Message 1874569 - Posted: 23 Jun 2017, 2:36:35 UTC - in response to Message 1874567.  
Last modified: 23 Jun 2017, 2:38:57 UTC

Sigh, just checked and all machines are down by 75 tasks at least. Wonder if this coincided with AP splitting again. Trying my usual tricks.


. . Hi Keith,

. . Frankly I am finding this erratic behaviour by the servers way toooo tiresome. It gets worse and worse and no mention of even an investigation into why. It really does undermine all enthuisiasm for the project. :(

Stephen

:(
ID: 1874569 · Report as offensive
Profile Keith Myers Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 29 Apr 01
Posts: 13161
Credit: 1,160,866,277
RAC: 1,873
United States
Message 1874571 - Posted: 23 Jun 2017, 2:51:10 UTC - in response to Message 1874569.  

I agree whole-heartedly. This is getting VERY tiresome. I am beginning to see the appeal that Zalster has for mostly, exclusively running Einstein. Just don't have the daily aggravations for that project. I might find myself flipping computer resources in that projects favor shortly. Without any indication that the scientists are aware of, or even have acknowledged the problem here, I may just take that route. My patience is wearing thin.
Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)
ID: 1874571 · Report as offensive
Profile betreger Project Donor
Avatar

Send message
Joined: 29 Jun 99
Posts: 11349
Credit: 29,581,041
RAC: 66
United States
Message 1874573 - Posted: 23 Jun 2017, 3:24:33 UTC - in response to Message 1874571.  
Last modified: 23 Jun 2017, 3:26:23 UTC

I am beginning to see the appeal that Zalster has for mostly, exclusively running Einstein.

Yes, but it usually lacks the excitement we have here with the amateure electrical hour. I say just let'm run and get whatever your resource shares are set to.
ID: 1874573 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1874579 - Posted: 23 Jun 2017, 5:11:56 UTC

It looks similar to Tuesday morning. The only tasks I can get to Download are Arecibo VLARs. A couple of APs are being sent to the two machines requesting them, otherwise three machines are not able to download any GPU tasks. The 4th machine is running Beta, no problem getting tasks from Beta.

Would someone please repeat what you did Tuesday morning to Open the Flood gates?
ID: 1874579 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13703
Credit: 208,696,464
RAC: 304
Australia
Message 1874581 - Posted: 23 Jun 2017, 5:20:27 UTC - in response to Message 1874567.  

Sigh, just checked and all machines are down by 75 tasks at least. Wonder if this coincided with AP splitting again. Trying my usual tricks.

In my case it's down to 75 tasks shared between the 2 GPUs.

There is a lot of Arecibo VLAR work still about, but if the splitters are all putting out work, the ready-to-send mix should be approx. 50/50, so on (at the absolute worst) every 4th request I should be able to get heaps of GPU work, instead of none. And even if there is no GPU work available in the feeder on a particular request, I should at least be able to some for the CPU, but that isn't happening either.
Even if I deselect Use Nvidia GPU, I still often get any CPU work.
Just the odd spurt here & there.


Oh for crying out loud.
Naturally as soon as I go to post this message, I get a batch of work. Still need another 150 or so for the GPUs alone just to reach the server side limits.
Grant
Darwin NT
ID: 1874581 · Report as offensive
TBar
Volunteer tester

Send message
Joined: 22 May 99
Posts: 5204
Credit: 840,779,836
RAC: 2,768
United States
Message 1874582 - Posted: 23 Jun 2017, 5:28:02 UTC - in response to Message 1874581.  

Yep, but that was a wimpy Kick; Fri Jun 23 01:25:32 2017 | SETI@home | Scheduler request completed: got 79 new tasks
You need to try it again...a little harder.
ID: 1874582 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13703
Credit: 208,696,464
RAC: 304
Australia
Message 1874585 - Posted: 23 Jun 2017, 5:35:21 UTC - in response to Message 1874582.  

Yep, but that was a wimpy Kick; Fri Jun 23 01:25:32 2017 | SETI@home | Scheduler request completed: got 79 new tasks

The most I've ever got from a single request is 54.
Generally after the outages when things are working i'll get batches of 52.
Back to the usual "No work available" response.

23/06/2017 14:59:12 | SETI@home | Reporting 1 completed tasks
23/06/2017 14:59:12 | SETI@home | Requesting new tasks for CPU and NVIDIA GPU
23/06/2017 14:59:15 | SETI@home | Scheduler request completed: got 0 new tasks
23/06/2017 14:59:15 | SETI@home | No tasks sent
23/06/2017 14:59:15 | SETI@home | No tasks are available for AstroPulse v7
23/06/2017 14:59:15 | SETI@home | Tasks for AMD/ATI GPU are available, but your preferences are set to not accept them
23/06/2017 14:59:15 | SETI@home | Tasks for Intel GPU are available, but your preferences are set to not accept them
23/06/2017 14:59:15 | SETI@home | This computer has reached a limit on tasks in progress


*bangs head on desk*
Once again, just as i'm about to post, new work. This time only 15.
Grant
Darwin NT
ID: 1874585 · Report as offensive
Previous · 1 . . . 15 · 16 · 17 · 18 · 19 · 20 · 21 . . . 29 · Next

Message boards : Number crunching : Panic Mode On (106) Server Problems?


 
©2024 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.