Panic Mode On (40) Server problems

Message boards : Number crunching : Panic Mode On (40) Server problems
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 7 · 8 · 9 · 10

AuthorMessage
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1044943 - Posted: 30 Oct 2010, 5:09:08 UTC

I just noticed some of the numbers on the server status page. Ready to Send is soaring, results in the field is dropping surprisingly fast.

I guess one good thing about this.. when everything comes back online, there will be PLENTY of work for everyone. I'm going to go ahead and predict 8 solid days at 94mbit on the network graph, assuming the new servers will handle that kind of stress for that long, which they certainly should.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1044943 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6325
Credit: 106,370,077
RAC: 121
Russia
Message 1044967 - Posted: 30 Oct 2010, 7:54:02 UTC

Why deleters are disabled? We can't look results anyway so better to delete all assimilated results/tasks to free disk spacebefore transition, not?

Also, completely disabling download servers not a way to go IMHO. Much better would be just to disable splitters. Then resends would provide some small work traffic that should not harm database server but would help to cleanup everything.
ID: 1044967 · Report as offensive
Profile Link
Avatar

Send message
Joined: 18 Sep 03
Posts: 834
Credit: 1,807,369
RAC: 0
Germany
Message 1044996 - Posted: 30 Oct 2010, 12:32:13 UTC - in response to Message 1044967.  

Why deleters are disabled? We can't look results anyway so better to delete all assimilated results/tasks to free disk spacebefore transition, not?

Also db_purge.x86_64 should be enabled to remove all validated workunits/results from the BOINC database before it will be copied to the new server.
ID: 1044996 · Report as offensive
DJStarfox

Send message
Joined: 23 May 01
Posts: 1066
Credit: 1,226,053
RAC: 2
United States
Message 1045002 - Posted: 30 Oct 2010, 13:42:50 UTC - in response to Message 1044996.  

Agreed that should disable the feeder.x86_64, and enable the deleters and purge. Makes no sense to assign work to clients if they can't download anything.
ID: 1045002 · Report as offensive
Claggy
Volunteer tester

Send message
Joined: 5 Jul 99
Posts: 4654
Credit: 47,537,079
RAC: 4
United Kingdom
Message 1045005 - Posted: 30 Oct 2010, 13:55:13 UTC - in response to Message 1045002.  
Last modified: 30 Oct 2010, 13:57:03 UTC

Agreed that should disable the feeder.x86_64, and enable the deleters and purge. Makes no sense to assign work to clients if they can't download anything.


I not so sure, but the feeder might need to be running for the scheduler to work, and then there's this changeset that appeared the other day: Changeset 22601

- scheduler/feeder: add a project config option <dont_send_jobs>.


If set, the feeder doesn't read jobs into shmem,
and the scheduler doesn't send jobs.
Intended for use when a project wants to process
a backlog of completed jobs and not issue more.


While there's no proof that this changeset has been applied here, it's the most obvious project that would require it,

Claggy
ID: 1045005 · Report as offensive
Profile Link
Avatar

Send message
Joined: 18 Sep 03
Posts: 834
Credit: 1,807,369
RAC: 0
Germany
Message 1045007 - Posted: 30 Oct 2010, 13:55:55 UTC - in response to Message 1045002.  

Work is not assigned to anyone... otherwise the results ready to send would be about 0 by now with the splitters turned off.
ID: 1045007 · Report as offensive
Profile Raistmer
Volunteer developer
Volunteer tester
Avatar

Send message
Joined: 16 Jun 01
Posts: 6325
Credit: 106,370,077
RAC: 121
Russia
Message 1045064 - Posted: 30 Oct 2010, 18:49:19 UTC - in response to Message 1045007.  

Work is not assigned to anyone... otherwise the results ready to send would be about 0 by now with the splitters turned off.

task to resend assigned even now. Lost task resend feature.
But it's impossible to download them.
ID: 1045064 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1045107 - Posted: 30 Oct 2010, 21:25:15 UTC

I'm getting ever-closer to having a cold room. Only about a day and a half of APs left to crunch.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1045107 · Report as offensive
Profile James Sotherden
Avatar

Send message
Joined: 16 May 99
Posts: 10436
Credit: 110,373,059
RAC: 54
United States
Message 1045123 - Posted: 30 Oct 2010, 23:07:53 UTC - in response to Message 1045107.  

I'm getting ever-closer to having a cold room. Only about a day and a half of APs left to crunch.


LOL, So thats why I had to jack the thermostat up a notch, I turned off the P4 and the i7. I did blow out all the dust bunnies. The i7 didnt seem to have much dust in it, First time I have blown it out in 18 months.
[/quote]

Old James
ID: 1045123 · Report as offensive
Robert Ribbeck
Avatar

Send message
Joined: 7 Jun 02
Posts: 644
Credit: 5,283,174
RAC: 0
United States
Message 1045227 - Posted: 31 Oct 2010, 15:17:18 UTC

Not only are the deleters and purging off
so are the "sah_assimilator's"

ID: 1045227 · Report as offensive
Profile Link
Avatar

Send message
Joined: 18 Sep 03
Posts: 834
Credit: 1,807,369
RAC: 0
Germany
Message 1045234 - Posted: 31 Oct 2010, 15:35:41 UTC - in response to Message 1045227.  

Well, two of them are on and they seem to be quick enough, yesterday we had 1.8M workunits waiting for assimilation, now it's down to 1.1M, so that's OK IMO.
ID: 1045234 · Report as offensive
Robert Ribbeck
Avatar

Send message
Joined: 7 Jun 02
Posts: 644
Credit: 5,283,174
RAC: 0
United States
Message 1045240 - Posted: 31 Oct 2010, 15:46:43 UTC - in response to Message 1045234.  

Well, two of them are on and they seem to be quick enough, yesterday we had 1.8M workunits waiting for assimilation, now it's down to 1.1M, so that's OK IMO.


Ya they came back on shortly after my post that they had been disabled

Gee Isn't that funny
ID: 1045240 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1045457 - Posted: 1 Nov 2010, 16:37:35 UTC

10 hours left on one AP and then I have a cold room for the next 3-4 weeks. The other cruncher finished its last AP early yesterday.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1045457 · Report as offensive
JLConawayII

Send message
Joined: 2 Apr 02
Posts: 188
Credit: 2,840,460
RAC: 0
United States
Message 1045489 - Posted: 1 Nov 2010, 20:26:09 UTC - in response to Message 1045457.  

10 hours left on one AP and then I have a cold room for the next 3-4 weeks. The other cruncher finished its last AP early yesterday.


Run something else, you'll still be helping the scientific community. Run some MW@home units or something.
ID: 1045489 · Report as offensive
Profile Fred J. Verster
Volunteer tester
Avatar

Send message
Joined: 21 Apr 04
Posts: 3252
Credit: 31,903,643
RAC: 0
Netherlands
Message 1045492 - Posted: 1 Nov 2010, 20:59:05 UTC - in response to Message 1045489.  
Last modified: 1 Nov 2010, 21:00:39 UTC

Last SETI MB WU's are Reported last Friday, only some ~100, now expired on 27october, are trying to UPLoad to SETI Bêta, which is still down.

Been crunching Einstein and GPUgrid on this host.

The 2 other QUAD's are doing a mix of Docking, CPDN, Leiden, MW, Collatz and DNETC (from time to time, it uses 2 GPU for 1 task, switched off, CPU tasks.)

SERVER Page isn't clear either, D'Load servers off, scheduling servers are on,
UPLoad servers and validators, also ON.
Probably to collect the last WU's, before everything, except 1 DataBase, is switched OFF, I think....

ID: 1045492 · Report as offensive
Robert Ribbeck
Avatar

Send message
Joined: 7 Jun 02
Posts: 644
Credit: 5,283,174
RAC: 0
United States
Message 1045493 - Posted: 1 Nov 2010, 21:04:56 UTC - in response to Message 1045492.  

Last SETI MB WU's are Reported last Friday, only some ~100, now expired on 27october, are trying to UPLoad to SETI Bêta, which is still down.

Been crunching Einstein and GPUgrid on this host.

The 2 other QUAD's are doing a mix of Docking, CPDN, Leiden, MW, Collatz and DNETC (from time to time, it uses 2 GPU for 1 task, switched off, CPU tasks.)

SERVER Page isn't clear either, D'Load servers off, scheduling servers are on,
UPLoad servers and validators, also ON.
Probably to collect the last WU's, before everything, except 1 DataBase, is switched OFF, I think....



Slow down
WTF
are you saying
ID: 1045493 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1045519 - Posted: 1 Nov 2010, 22:55:41 UTC - in response to Message 1045489.  

10 hours left on one AP and then I have a cold room for the next 3-4 weeks. The other cruncher finished its last AP early yesterday.


Run something else, you'll still be helping the scientific community. Run some MW@home units or something.

I wasn't saying it like everyone else does where they get mad and think they deserve to always have work. This is the only project that interests me. I support the current server situation. I'm fine with idle time, because I'm in it for the science, not the points.
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1045519 · Report as offensive
Previous · 1 . . . 7 · 8 · 9 · 10

Message boards : Number crunching : Panic Mode On (40) Server problems


 
©2025 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.