Panic Mode On (62) Server problems?

Message boards : Number crunching : Panic Mode On (62) Server problems?
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 10 · Next

AuthorMessage
Profile Dimly Lit Lightbulb 😀
Volunteer tester
Avatar

Send message
Joined: 30 Aug 08
Posts: 15399
Credit: 7,423,413
RAC: 1
United Kingdom
Message 1176099 - Posted: 6 Dec 2011, 16:17:45 UTC

Managed to pick up an astropulse last night, otherwise in about an hour it would've been backup project time.

Member of the People Encouraging Niceness In Society club.

ID: 1176099 · Report as offensive
AndrewM
Volunteer tester

Send message
Joined: 5 Jan 08
Posts: 369
Credit: 34,275,196
RAC: 0
Australia
Message 1176149 - Posted: 6 Dec 2011, 23:41:21 UTC - in response to Message 1176070.  

I'm still dreaming of the day when my GPU's don't run dry 2-3 times a week.

Steve


I'm still dreaming of the week when my GPU's don't run dry 2-3 times a day.


AndrewM
ID: 1176149 · Report as offensive
Profile arkayn
Volunteer tester
Avatar

Send message
Joined: 14 May 99
Posts: 4438
Credit: 55,006,323
RAC: 0
United States
Message 1176166 - Posted: 7 Dec 2011, 0:51:14 UTC

From empty yesterday on both machines, I currently have about 400 units on each after I re-enabled the proxy server.

ID: 1176166 · Report as offensive
tbret
Volunteer tester
Avatar

Send message
Joined: 28 May 99
Posts: 3380
Credit: 296,162,071
RAC: 40
United States
Message 1176179 - Posted: 7 Dec 2011, 1:48:08 UTC - in response to Message 1176166.  

From empty yesterday on both machines, I currently have about 400 units on each after I re-enabled the proxy server.


I really, really believe that if we knew WHY, we'd know something.

Maybe we'd know WHY.
ID: 1176179 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1176181 - Posted: 7 Dec 2011, 1:59:43 UTC

During the maintenance outage I noticed all of my "suck" downloads completed are great speed. I was seeing 800k on several.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1176181 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1176195 - Posted: 7 Dec 2011, 3:37:00 UTC

I've noticed that when connected direct, if you can get units, the download speed is quite good, up to 25KBps +, even though the proxy is still faster, these speeds are the fastest I've ever had direct from the project in 5 years.

I wonder if this means that despite what the Cricket graphs tell us, the actual network loading is not "super saturated" like it is when normally coming back from an outage, just "busy".

This, combined with the great difficulty getting work allocated (usually only one or two units at a time) means we could be looking at a Scheduler problem rather than a network overload.

All the scheduling processes are located on "bane". I wonder if this server is really being the "bane" of our lives ?

T.A.
ID: 1176195 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6534
Credit: 196,805,888
RAC: 57
United States
Message 1176204 - Posted: 7 Dec 2011, 4:26:58 UTC - in response to Message 1176195.  

I've noticed that when connected direct, if you can get units, the download speed is quite good, up to 25KBps +, even though the proxy is still faster, these speeds are the fastest I've ever had direct from the project in 5 years.

I wonder if this means that despite what the Cricket graphs tell us, the actual network loading is not "super saturated" like it is when normally coming back from an outage, just "busy".

This, combined with the great difficulty getting work allocated (usually only one or two units at a time) means we could be looking at a Scheduler problem rather than a network overload.

All the scheduling processes are located on "bane". I wonder if this server is really being the "bane" of our lives ?

T.A.

IIRC we are subject to the C10K problem. Maybe with the updates Matt was talking about this will no longer be an issue? As some software is not subject to this problem.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[
ID: 1176204 · Report as offensive
Starman
Avatar

Send message
Joined: 15 May 99
Posts: 204
Credit: 81,351,915
RAC: 25
Canada
Message 1176377 - Posted: 8 Dec 2011, 1:56:15 UTC

Is there anybody Home?

Looks like something is broken again! Can't report what work units I have completed. Not that it is slowing my decline in RAC by much.
ID: 1176377 · Report as offensive
Profile perryjay
Volunteer tester
Avatar

Send message
Joined: 20 Aug 02
Posts: 3377
Credit: 20,676,751
RAC: 0
United States
Message 1176378 - Posted: 8 Dec 2011, 2:03:55 UTC - in response to Message 1176377.  

I just managed to report a few, (28) but it took almost three minutes to complete. Also looks like the Cricket graphs are way down. Not completely dead but struggling.


PROUD MEMBER OF Team Starfire World BOINC
ID: 1176378 · Report as offensive
Profile Lint trap

Send message
Joined: 30 May 03
Posts: 871
Credit: 28,092,319
RAC: 0
United States
Message 1176385 - Posted: 8 Dec 2011, 2:44:18 UTC


I was getting only the front page for a while there. All other pages were reporting the project as down for maintenance. That was soon after 00:00 GMT, IIRC.

Just now I was able to report 90 completed tasks. No problems.

Lt

ID: 1176385 · Report as offensive
rob smith Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer moderator
Volunteer tester

Send message
Joined: 7 Mar 03
Posts: 22160
Credit: 416,307,556
RAC: 380
United Kingdom
Message 1176444 - Posted: 8 Dec 2011, 8:26:12 UTC

It looks as though the upload server has just gone off for a break. Also the backup server is reporting that its about 8hours behind the master, so things aren't too happy in the server room.
Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?
ID: 1176444 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13720
Credit: 208,696,464
RAC: 304
Australia
Message 1176447 - Posted: 8 Dec 2011, 8:44:17 UTC - in response to Message 1176444.  

It looks as though the upload server has just gone off for a break.

Yep.
Once again i'm buried under uploads that won't.
Grant
Darwin NT
ID: 1176447 · Report as offensive
Profile S@NL Etienne Dokkum
Volunteer tester
Avatar

Send message
Joined: 11 Jun 99
Posts: 212
Credit: 43,822,095
RAC: 0
Netherlands
Message 1176468 - Posted: 8 Dec 2011, 10:25:51 UTC

Let's not moan to hard all at once... Things have been picking up for the good for a while now.

The boys in the lab will probably just jumpstart the rigs again when they get in in the morning and all will be well

If they would just find a way to stop the "shortie"-storm I'd be a very happy cruncher ;-)
ID: 1176468 · Report as offensive
tbret
Volunteer tester
Avatar

Send message
Joined: 28 May 99
Posts: 3380
Credit: 296,162,071
RAC: 40
United States
Message 1176469 - Posted: 8 Dec 2011, 10:26:18 UTC - in response to Message 1176447.  

It looks as though the upload server has just gone off for a break.

Yep.
Once again i'm buried under uploads that won't.


It's 38 degrees F in Berkeley. Someone open a window.

No, I didn't mean they should jump.

ID: 1176469 · Report as offensive
Profile Belthazor
Volunteer tester
Avatar

Send message
Joined: 6 Apr 00
Posts: 219
Credit: 10,373,795
RAC: 13
Russia
Message 1176471 - Posted: 8 Dec 2011, 10:45:55 UTC - in response to Message 1176469.  


It's 38 degrees F in Berkeley.


Is it about 4 C? So cold in CA? I'm shocked!

ID: 1176471 · Report as offensive
JohnDK Crowdfunding Project Donor*Special Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 28 May 00
Posts: 1222
Credit: 451,243,443
RAC: 1,127
Denmark
Message 1176487 - Posted: 8 Dec 2011, 13:44:54 UTC - in response to Message 1176472.  

Wish I had such luck....
The way it's been working here, my top rig runs out of Seti due to repeated 'no work' responses and just a handful of issued tasks. It has a 0% share on Einstein, so when the GPU runs dry on Seti, it picks up several hours of Einstein work and persistently keeps trying to get work from Seti while that is being crunched. It manages to get something built up and goes back to it when the Einstein is done and then repeats the cycle.

Of course, with today's outage coming up, it's gonna be doing Einstein for the next 12 hours or more.

Only the slower hosts on the project could stay supplied with Seti work the way things are going right now.

The servers have actually held up reasonably well considering the shorty pounding...if we could get a day or two with some datasets split that did not contain 95% VHAR we might be able to get a leg up on things.

Guess it's your BOINC 6.12.* that sucks...

I stayed with 6.10.60 on my PCs, except when I started my old PC again I installed the new version thinking, it can't be that bad. But I ran out of work more than once and decided to go back to 6.10.60, and within a day or so I had a full cache and it have stayed that way.

So my question is, will the BOINC team fix that clear problem in future versions ?
ID: 1176487 · Report as offensive
Profile perryjay
Volunteer tester
Avatar

Send message
Joined: 20 Aug 02
Posts: 3377
Credit: 20,676,751
RAC: 0
United States
Message 1176495 - Posted: 8 Dec 2011, 15:01:20 UTC - in response to Message 1176472.  

Clyde,
If you set the resource share to zero, you only pick up work for that project when your main project is out of work. When that zero project has work it will run them until it finishes all of them it has downloaded and then if your main project has work it will start back on it. It shouldn't load so much of your backup project as to run into any problem with deadlines.


PROUD MEMBER OF Team Starfire World BOINC
ID: 1176495 · Report as offensive
j tramer

Send message
Joined: 6 Oct 03
Posts: 242
Credit: 5,412,368
RAC: 0
Canada
Message 1176498 - Posted: 8 Dec 2011, 15:52:40 UTC

same crap different day.....i cant get enough work to last more than a couple of hours....that why i left before....back to the same crap
ID: 1176498 · Report as offensive
Cruncher-American Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor

Send message
Joined: 25 Mar 02
Posts: 1513
Credit: 370,893,186
RAC: 340
United States
Message 1176500 - Posted: 8 Dec 2011, 16:01:47 UTC - in response to Message 1176495.  

Clyde,
If you set the resource share to zero, you only pick up work for that project when your main project is out of work. When that zero project has work it will run them until it finishes all of them it has downloaded and then if your main project has work it will start back on it. It shouldn't load so much of your backup project as to run into any problem with deadlines.


Awhile back, when I was running a couple of share 0 projects against SETI, to avoid unused cycles when SETI was bad, IIRC it D/L WUs for those projects 1 at a time per CPU/GPU so as to not build up a queue. Does BOINC still have the same behavior?
ID: 1176500 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51468
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1176503 - Posted: 8 Dec 2011, 16:35:13 UTC - in response to Message 1176495.  

Clyde,
If you set the resource share to zero, you only pick up work for that project when your main project is out of work. When that zero project has work it will run them until it finishes all of them it has downloaded and then if your main project has work it will start back on it. It shouldn't load so much of your backup project as to run into any problem with deadlines.

That's the way it's been working....and probably will continue to at the rate the rig is (or is not) getting Seti work. When 1 GPU goes idle, it downloads a batch of Einstein tasks, about 10-20, not sure exactly. It then finishes the Seti work on the remaining 3 GPUs and works on the batch of Einstein tasks until they have finished. If there is Seti work that has been downloaded in the interim, Einstein does not request new work and goes back to the Seti tasks when the Einstein is done and will stay on them until a GPU goes dry again.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1176503 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 · 7 . . . 10 · Next

Message boards : Number crunching : Panic Mode On (62) 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.