Panic Mode On (80) Server Problems?


log in

Advanced search

Message boards : Number crunching : Panic Mode On (80) Server Problems?

1 · 2 · 3 · 4 . . . 25 · Next
Author Message
Profile arkayn
Volunteer tester
Avatar
Send message
Joined: 14 May 99
Posts: 3545
Credit: 46,345,496
RAC: 30,943
United States
Message 1321335 - Posted: 29 Dec 2012, 16:12:44 UTC

Almost nothing is working currently, most of us are on backup projects.
____________

zoom314
Avatar
Send message
Joined: 30 Nov 03
Posts: 44610
Credit: 35,465,937
RAC: 8,777
Message 1321343 - Posted: 29 Dec 2012, 16:24:33 UTC - in response to Message 1321335.

Aye the plumbing is all plugged up, ach someone stuffed too many Klingons in there...
____________

Profile Donald L. Johnson
Avatar
Send message
Joined: 5 Aug 02
Posts: 5709
Credit: 566,288
RAC: 561
United States
Message 1321391 - Posted: 29 Dec 2012, 17:26:48 UTC

I'm seeing the same thing you guys are - Scheduler time-outs, them come back 5 minutes later and get through, and pick up the ghosts from the earlier contact.

First tries at downloads time out, second or third time they come down like greased lightning. Must depend on which download server I hit.
____________
Donald
Infernal Optimist / Submariner, retired

N9JFE
Volunteer tester
Avatar
Send message
Joined: 4 Oct 99
Posts: 9421
Credit: 12,021,112
RAC: 15,544
United States
Message 1321430 - Posted: 29 Dec 2012, 17:50:54 UTC
Last modified: 29 Dec 2012, 17:52:21 UTC

A couple days ago, I set the host with the new GT630 to NNT for Einstein because I wanted it to actually get some Seti done so I could see how long it was taking 2 at a time. So I check on it today and find it doing nothing because all the Seti downloads are stuck. Grrrrr.

Seeing that even babysitting it wasn't going to get me Seti work as fast as it could be finished, I turned Einstein loose again.

Now I want to know why Einsteins that are due on 1/12/13 and estimated to run in 5:32 are running HP ahead of the one Seti I managed to get that is due on 1/11/13 and estimated to run in 21:32.

Meanwhile, my i7 suddenly had a surge of successful downloads. It got everything and asked for more, and was resent 20 ghosts. Then another 20 ghosts. Then I discovered a typo in its hosts file and fixed it. Now it's moving downloads in little bits at a time (no networking pun intended). Maybe I should take out that line in hosts. [edit] Or change it from georgem to vader.
____________
David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.


Profile Brother Frank
Send message
Joined: 10 Dec 11
Posts: 26
Credit: 15,142,410
RAC: 0
United States
Message 1321491 - Posted: 29 Dec 2012, 18:13:56 UTC - in response to Message 1321335.

My upload timeouts began about Wednesday and got worse day by day. Noticed that my RAC rose and rose from Dec 5th until Dec 27th time point. I am getting work unit limit reached even on my little i3 notebook without dedicated graphics. Nothing is working here anymore. Am switching to alternate project GPU Grid and Rosetta @ Home today. Have tried proxy servers and no new work setting used during that very bad set of scheduler/server problems from November to early December time period. This last two + days feels as bad as the last meltdown.
____________
Frank Elliott,Member of Carepages.com,a chronic illness support site. Was FrankLivingFully there.Free user name & pw needed. My Google+ Profile is:
https://profiles.google.com/u/0/10871372137584 Science,SF,Space,Astronomy,Medicine,Psyc Topics.

Profile ivan
Volunteer tester
Avatar
Send message
Joined: 5 Mar 01
Posts: 556
Credit: 120,827,719
RAC: 93,049
United Kingdom
Message 1321498 - Posted: 29 Dec 2012, 18:16:28 UTC - in response to Message 1321391.

I'm seeing the same thing you guys are - Scheduler time-outs, them come back 5 minutes later and get through, and pick up the ghosts from the earlier contact.

First tries at downloads time out, second or third time they come down like greased lightning. Must depend on which download server I hit.

Yes, I synchronised my home machine with NTP (I presume the Labs are too; my lab servers are) and observed that after my PC sent a request for more work both the time-stamp on my "Last contact" and the "Sent" time for (a) recent file(s) updated to 3-5 seconds after my request time. Sometimes the scheduler request would time out, other times I got notification in 5-20 seconds that there was more work, often in the form of ghosts if the file entry existed before my request. Seems the database is updating quickly, but somewhere along the chain the return of information about the new file(s) is disappearing.
____________

TBar
Volunteer tester
Send message
Joined: 22 May 99
Posts: 1080
Credit: 31,117,987
RAC: 85,203
United States
Message 1321552 - Posted: 29 Dec 2012, 18:58:57 UTC
Last modified: 29 Dec 2012, 19:12:37 UTC

It's still sometimes taking me around 5-10 minutes to complete a simple file upload. Once that is accomplished, the rest seems to work fine. If you are working on 20-40 minute long tasks, as myself, it's not that much of a problem. First step is to Fix the Upload Server.

Antjest
Volunteer tester
Send message
Joined: 27 Oct 99
Posts: 27
Credit: 15,182,821
RAC: 1,273
Slovenia
Message 1321613 - Posted: 29 Dec 2012, 20:40:51 UTC

All the problems started when 4th and 5th AP splitter went online. That was the only visible change after the maintenance outage.
Wheather this is server overload problem or network congestion.

They should stop those two and my bet is everything will return to normal as it was for a few days before.

Grant (SSSF)
Send message
Joined: 19 Aug 99
Posts: 5567
Credit: 51,650,456
RAC: 44,635
Australia
Message 1321681 - Posted: 29 Dec 2012, 21:54:03 UTC - in response to Message 1321613.
Last modified: 29 Dec 2012, 22:05:50 UTC

The splitters are working better than they have for months, the assimilators are finally keeping up.
And it appears the Scheduler is no longer giving errors, it's just a shame it's not possible to upload and almost all Scheduler requests result in a timeout.


EDIT- and once again for some unfathonable reason, setting No New Tasks increases the chance of not getting a Scheduler time out. It still happens, just not as regularly as when trying to get work.
____________
Grant
Darwin NT.

WezH
Volunteer tester
Send message
Joined: 19 Aug 99
Posts: 78
Credit: 3,096,714
RAC: 19,663
Finland
Message 1321717 - Posted: 29 Dec 2012, 22:14:01 UTC

Panic mode? Again?

I know to the funding issue, I have made my donation, but....

Good luck for next year, I'm officially off-line again. Power bill vs S@H....

Set all my crunchers to NNT, only left Beta Big units to crunch to research.

See You All next Year and HOPE for Best!
____________

Rolf
Send message
Joined: 16 Jun 09
Posts: 114
Credit: 7,800,622
RAC: 1
Switzerland
Message 1321733 - Posted: 29 Dec 2012, 22:20:50 UTC - in response to Message 1321613.
Last modified: 29 Dec 2012, 22:22:40 UTC

All the problems started when 4th and 5th AP splitter went online. That was the only visible change after the maintenance outage.

This is more or less the same I wanted to say in my post:
http://setiathome.berkeley.edu/forum_thread.php?id=70070&postid=1321332
Easy to test, isn't too complicated! I think it's worth testing it!
Edit: And it will not increase anybody's power bill.

zoom314
Avatar
Send message
Joined: 30 Nov 03
Posts: 44610
Credit: 35,465,937
RAC: 8,777
Message 1321749 - Posted: 29 Dec 2012, 22:29:21 UTC - in response to Message 1321733.

All the problems started when 4th and 5th AP splitter went online. That was the only visible change after the maintenance outage.

This is more or less the same I wanted to say in my post:
http://setiathome.berkeley.edu/forum_thread.php?id=70070&postid=1321332
Easy to test, isn't too complicated! I think it's worth testing it!
Edit: And it will not increase anybody's power bill.

So that's what torpedoed the system... sigh.
____________

bill
Send message
Joined: 16 Jun 99
Posts: 848
Credit: 20,704,450
RAC: 17,479
United States
Message 1321833 - Posted: 29 Dec 2012, 23:56:29 UTC

FWIW Uploads are uploading with an occasional Retry
to help it along , downloads are downloading,
the GPUs are staying busy with SETI. Rosetta and Einstein
on the CPUs. I'm only running 1.0/.01 for Cache.

Maybe a smaller Cache works better in the present circumstances.

Chris Oliver
Avatar
Send message
Joined: 4 Jul 99
Posts: 39
Credit: 48,142,491
RAC: 33,398
United Kingdom
Message 1321855 - Posted: 30 Dec 2012, 0:26:17 UTC - in response to Message 1321803.

Best to ween the kitties off the weed in the new year.....



Ya know.........the kitties really don't care anymore.
They have given this poject EVERYTHING for many years now.


Yes, it pisses me off when things turn to sh/t.
I want my caches back.

It really does piss me off right now.. But, when the smoke clears, as it always does sooner or later.

You know who the first one back on track is?

Me.

Me. So, nobody best be calling me any names at this point.



____________

TBar
Volunteer tester
Send message
Joined: 22 May 99
Posts: 1080
Credit: 31,117,987
RAC: 85,203
United States
Message 1321856 - Posted: 30 Dec 2012, 0:27:22 UTC

As I recall, during the other 'Psychotic Server Episodes', the Uploads continued without any problems. It never took longer than a second or two to complete the upload. This is what has been going on since yesterday, around the time the Multibeam Shortie Storm began;

12/29/2012 6:58:31 PM | SETI@home | Computation for task ap_25no12ad_B5_P1_00295_20121227_30240.wu_1 finished
12/29/2012 6:58:31 PM | SETI@home | Starting task ap_25no12ad_B6_P0_00240_20121227_31377.wu_1 using astropulse_v6 version 604 (ati_opencl_100) in slot 2
12/29/2012 6:58:33 PM | SETI@home | Started upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 6:58:56 PM | | Project communication failed: attempting access to reference site
12/29/2012 6:58:56 PM | SETI@home | Temporarily failed upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0: connect() failed
12/29/2012 6:58:56 PM | SETI@home | Backing off 3 min 47 sec on upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 6:58:57 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 6:59:28 PM | SETI@home | Started upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:00:13 PM | | Project communication failed: attempting access to reference site
12/29/2012 7:00:13 PM | SETI@home | Temporarily failed upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0: connect() failed
12/29/2012 7:00:13 PM | SETI@home | Backing off 5 min 12 sec on upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:00:15 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 7:00:54 PM | SETI@home | Started upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:04:04 PM | | Project communication failed: attempting access to reference site
12/29/2012 7:04:04 PM | SETI@home | Temporarily failed upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0: transient HTTP error
12/29/2012 7:04:04 PM | SETI@home | Backing off 10 min 45 sec on upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:04:05 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 7:04:06 PM | SETI@home | Started upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:05:02 PM | | Project communication failed: attempting access to reference site
12/29/2012 7:05:02 PM | SETI@home | Temporarily failed upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0: connect() failed
12/29/2012 7:05:02 PM | SETI@home | Backing off 20 min 59 sec on upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:05:03 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 7:05:04 PM | SETI@home | Started upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:05:37 PM | SETI@home | Finished upload of ap_25no12ad_B5_P1_00295_20121227_30240.wu_1_0
12/29/2012 7:05:40 PM | SETI@home | Sending scheduler request: To fetch work.
12/29/2012 7:05:40 PM | SETI@home | Reporting 1 completed tasks, requesting new tasks for NVIDIA and ATI
12/29/2012 7:05:44 PM | SETI@home | Scheduler request completed: got 1 new tasks
12/29/2012 7:05:47 PM | SETI@home | Started download of 08oc12aa.31216.126844.12.10.140
12/29/2012 7:06:01 PM | SETI@home | Finished download of 08oc12aa.31216.126844.12.10.140
....

That's 7 minutes to upload one file. If that would have been last night during the height of the SS, I would have had another completed task before that one uploaded. In fact, I had around 8 stalled uploads at one point last night. If I got them all to remain active long enough, the report went through and I received downloaded files. If one was stalled, nothing worked. Things calmed down when the SS passed and the 4 minute MBs were replaced with 20+ minute MBs. Right now, I'm completing the Upload before the next task is finished.

The way I see it, the latest problem began when someone decided to release a Shortie Storm on a malfunctioning Upload Server. The Upload Server is still Borked.

Grant (SSSF)
Send message
Joined: 19 Aug 99
Posts: 5567
Credit: 51,650,456
RAC: 44,635
Australia
Message 1321887 - Posted: 30 Dec 2012, 1:37:12 UTC - in response to Message 1321856.

The way I see it, the latest problem began when someone decided to release a Shortie Storm on a malfunctioning Upload Server. The Upload Server is still Borked.

Even before the shorties, uploads & the Scheduler were both stuffed.
At least ince the shorties haved stopped arriving things aren't as bad as they were- i wouldn't say they've improved as things are still seriously screwed. But they're not as bad as they were (uploads still take ages & i'm still getting them backing up & Scheduler requests still timeout although the number of errors has dropped).
____________
Grant
Darwin NT.

Profile CLYDE
Volunteer tester
Avatar
Send message
Joined: 9 Aug 99
Posts: 801
Credit: 17,686,467
RAC: 34,477
United States
Message 1321908 - Posted: 30 Dec 2012, 2:43:45 UTC - in response to Message 1321803.

Ya know.........the kitties really don't care anymore.
They have given this poject EVERYTHING for many years now.


Yes, it pisses me off when things turn to sh/t.
I want my caches back.

It really does piss me off right now.. But, when the smoke clears, as it always does sooner or later.

You know who the first one back on track is?

Me.

Me. So, nobody best be calling me any names at this point.



Yes, it can be frustrating at times. But...

This is the only real attempt to find Intelligent Life, not just life, somewhere beyond this planet.

Despite the problems: Let's just 'Keep On Going'.

____________

TBar
Volunteer tester
Send message
Joined: 22 May 99
Posts: 1080
Credit: 31,117,987
RAC: 85,203
United States
Message 1321918 - Posted: 30 Dec 2012, 3:12:54 UTC - in response to Message 1321887.

The way I see it, the latest problem began when someone decided to release a Shortie Storm on a malfunctioning Upload Server. The Upload Server is still Borked.

Even before the shorties, uploads & the Scheduler were both stuffed.
At least ince the shorties haved stopped arriving things aren't as bad as they were- i wouldn't say they've improved as things are still seriously screwed. But they're not as bad as they were (uploads still take ages & i'm still getting them backing up & Scheduler requests still timeout although the number of errors has dropped).

I really haven't had any trouble with downloads. Other than the scheduler continually trying to give me all ATI MBs and nothing for the nVidia card. I just made the transition from all MBs to half APs & half MBs. Things are working well except for the Upload Server problem. Here's the latest one, 11 minutes and note how the scheduler didn't make a peep during that time. As soon as the upload completed, the scheduler kicked in and topped me off at 200 again.
12/29/2012 9:32:33 PM | SETI@home | Computation for task 08oc12ab.6744.8247.6.10.239_0 finished
12/29/2012 9:32:33 PM | SETI@home | Starting task 08oc12ab.6744.8247.6.10.208_1 using setiathome_enhanced version 609 (cuda23) in slot 3
12/29/2012 9:32:35 PM | SETI@home | Started upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:33:38 PM | SETI@home | Started download of 08oc12aa.13907.128071.13.10.99
12/29/2012 9:37:22 PM | | Project communication failed: attempting access to reference site
12/29/2012 9:37:22 PM | SETI@home | Temporarily failed upload of 08oc12ab.6744.8247.6.10.239_0_0: transient HTTP error
12/29/2012 9:37:22 PM | SETI@home | Backing off 3 min 56 sec on upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:37:24 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 9:37:27 PM | SETI@home | Started upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:38:38 PM | | Project communication failed: attempting access to reference site
12/29/2012 9:38:38 PM | SETI@home | Temporarily failed upload of 08oc12ab.6744.8247.6.10.239_0_0: transient HTTP error
12/29/2012 9:38:38 PM | SETI@home | Backing off 4 min 15 sec on upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:38:39 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 9:38:43 PM | SETI@home | Started upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:39:06 PM | | Project communication failed: attempting access to reference site
12/29/2012 9:39:06 PM | SETI@home | Temporarily failed upload of 08oc12ab.6744.8247.6.10.239_0_0: connect() failed
12/29/2012 9:39:06 PM | SETI@home | Backing off 9 min 57 sec on upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:39:07 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 9:39:11 PM | SETI@home | Started upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:39:13 PM | | Project communication failed: attempting access to reference site
12/29/2012 9:39:13 PM | SETI@home | Temporarily failed download of 08oc12aa.13907.128071.13.10.99: transient HTTP error
12/29/2012 9:39:13 PM | SETI@home | Backing off 6 min 0 sec on download of 08oc12aa.13907.128071.13.10.99
12/29/2012 9:39:14 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 9:39:15 PM | SETI@home | Started download of 08oc12aa.13907.128071.13.10.99
12/29/2012 9:39:51 PM | SETI@home | Finished download of 08oc12aa.13907.128071.13.10.99
12/29/2012 9:41:37 PM | | Project communication failed: attempting access to reference site
12/29/2012 9:41:37 PM | SETI@home | Temporarily failed upload of 08oc12ab.6744.8247.6.10.239_0_0: transient HTTP error
12/29/2012 9:41:37 PM | SETI@home | Backing off 22 min 6 sec on upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:41:38 PM | | Internet access OK - project servers may be temporarily down.
12/29/2012 9:41:40 PM | SETI@home | Started upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:43:47 PM | SETI@home | Finished upload of 08oc12ab.6744.8247.6.10.239_0_0
12/29/2012 9:43:48 PM | SETI@home | Sending scheduler request: To fetch work.
12/29/2012 9:43:48 PM | SETI@home | Reporting 1 completed tasks, requesting new tasks for NVIDIA and ATI
12/29/2012 9:43:54 PM | SETI@home | Scheduler request completed: got 1 new tasks
12/29/2012 9:43:56 PM | SETI@home | Started download of 08oc12ab.13879.15200.11.10.53
12/29/2012 9:44:10 PM | SETI@home | Finished download of 08oc12ab.13879.15200.11.10.53
....

1 · 2 · 3 · 4 . . . 25 · Next

Message boards : Number crunching : Panic Mode On (80) Server Problems?

Copyright © 2014 University of California