Panic Mode On (63) Server problems?

Message boards : Number crunching : Panic Mode On (63) Server problems?
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 9 · Next

AuthorMessage
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13371
Credit: 208,696,464
RAC: 304
Australia
Message 1178665 - Posted: 16 Dec 2011, 21:47:47 UTC


Time for major panicing- Scarecrow's Graphs link gives me a 404 error for the last 12 hours or so.
Grant
Darwin NT
ID: 1178665 · Report as offensive
BWX

Send message
Joined: 31 May 03
Posts: 36
Credit: 156,754,993
RAC: 24
United States
Message 1178677 - Posted: 16 Dec 2011, 22:06:23 UTC

Bizzare - my one rig with 1 GPU and an i7 is only getting GPU WU's, and my quad with 2 GPU's is only getting CPU WU's.

If I manually click 'Update', it will ask for the other (or both) type(s), but before too long, it only asks for the ones it is reporting, keeping the vicious cycle going.

Bad scheduling in the BOINC client?
ID: 1178677 · Report as offensive
bill

Send message
Joined: 16 Jun 99
Posts: 861
Credit: 29,352,955
RAC: 0
United States
Message 1178679 - Posted: 16 Dec 2011, 22:10:19 UTC - in response to Message 1178532.  

Oh well, back to AP only after the MB version test, and back to whining about No APs available :-)

Nothing is new under SETIs sun eh?

LOL

Edit: And the worst of it all, is knowing that I voluntarily put myself through this PITA....

Hehe....


It's Friday and you've finished your project.

Time for a wee dram of Macallen me'thinks.
ID: 1178679 · Report as offensive
Profile SciManStev Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Jun 99
Posts: 6634
Credit: 121,090,076
RAC: 0
United States
Message 1178682 - Posted: 16 Dec 2011, 22:21:20 UTC - in response to Message 1178681.  

Bizzare - my one rig with 1 GPU and an i7 is only getting GPU WU's, and my quad with 2 GPU's is only getting CPU WU's.

If I manually click 'Update', it will ask for the other (or both) type(s), but before too long, it only asks for the ones it is reporting, keeping the vicious cycle going.

Bad scheduling in the BOINC client?


On the SETI@home preferences page for the computer venue in question, untick either Use NVIDIA GPU, or "Use CPU ", whatever work you do not want for the moment.

That works perfect, after one more request, it stops requesting for the one that is unticked.

When you're satisfied with the amount of work of the type you want, just retick the unticked type.

I use that all the time in an effort to get as many GPU units as possible, but still maintain enough CPU units. Ticking and unticking what you want really helps.

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1178682 · Report as offensive
Profile HAL9000
Volunteer tester
Avatar

Send message
Joined: 11 Sep 99
Posts: 6533
Credit: 196,805,888
RAC: 57
United States
Message 1178686 - Posted: 16 Dec 2011, 22:30:51 UTC - in response to Message 1178665.  


Time for major panicing- Scarecrow's Graphs link gives me a 404 error for the last 12 hours or so.

If you go to the root domain you will see an apache test page. Looks like his/their webserver might have barfed.
SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the BP6/VP6 User Group today!
ID: 1178686 · Report as offensive
Grant (SSSF)
Volunteer tester

Send message
Joined: 19 Aug 99
Posts: 13371
Credit: 208,696,464
RAC: 304
Australia
Message 1178760 - Posted: 17 Dec 2011, 5:46:54 UTC
Last modified: 17 Dec 2011, 6:03:12 UTC

The upload problem makes it's return- uploads are accumulating again.

EDIT- Panic over, it's working again (for now at least).
Grant
Darwin NT
ID: 1178760 · Report as offensive
Profile rebest Project Donor
Volunteer tester
Avatar

Send message
Joined: 16 Apr 00
Posts: 1296
Credit: 45,357,093
RAC: 0
United States
Message 1178944 - Posted: 17 Dec 2011, 20:56:59 UTC

Wow. I take a few months off the boards and all h*** breaks loose.

OK, I've spent the past 15 minutes reading various NC threads. I've found vast quantities of gibberish, but very little useful information..

I'm trying to determine why, over the past 2 months, the RAC for my three machines has gone from around 25K to well under 10K. Now, I do not babysit my rigs, so I don't know the day to day availability of work. However, when I have checked, there appears to typically be 250K work units available on the server, but I have a pathetic few on my machine..

I'm also seeing a lot of gripes about BOINC 6.12. I have also noticed ridiculously high - and frequent - project backoffs and what appears to be a total disregard for my cache settings.

So, are the problems the result of inconsistent availability of work, or is 6.12 a piece of crap?

Thanks!!



Join the PACK!
ID: 1178944 · Report as offensive
bill

Send message
Joined: 16 Jun 99
Posts: 861
Credit: 29,352,955
RAC: 0
United States
Message 1178949 - Posted: 17 Dec 2011, 21:07:05 UTC - in response to Message 1178944.  

6.12.x = crap go to 6.10.58 or 6.10.60
ID: 1178949 · Report as offensive
Profile rebest Project Donor
Volunteer tester
Avatar

Send message
Joined: 16 Apr 00
Posts: 1296
Credit: 45,357,093
RAC: 0
United States
Message 1178953 - Posted: 17 Dec 2011, 21:24:22 UTC - in response to Message 1178949.  

6.12.x = crap go to 6.10.58 or 6.10.60


Ah. Very good. Will do.

Thanks for the reply.

Join the PACK!
ID: 1178953 · Report as offensive
Profile Donald L. Johnson
Avatar

Send message
Joined: 5 Aug 02
Posts: 8240
Credit: 14,654,533
RAC: 20
United States
Message 1178956 - Posted: 17 Dec 2011, 21:26:34 UTC - in response to Message 1178944.  

Wow. I take a few months off the boards and all h*** breaks loose.


Welcome back (8{)

So, are the problems the result of inconsistent availability of work, or is 6.12 a piece of crap?

6 of one, half-dozen of the other, plus another shorty storm.

I haven't "upgraded' to 6.12.xx, but most of the gripes seem to be about the 10x longer back-off times, which probably compound the usual distribution issues for the faster rigs.

My ancient G4s running 6.10 56 and 6.10.58 are not having any problems getting work or reporting completions. Plus I live in California, so the transfer route is pretty short and direct. YMMV
Donald
Infernal Optimist / Submariner, retired
ID: 1178956 · Report as offensive
j tramer

Send message
Joined: 6 Oct 03
Posts: 242
Credit: 5,412,368
RAC: 0
Canada
Message 1178959 - Posted: 17 Dec 2011, 21:37:05 UTC

back to the same crap.....as soon as i run out, i shut it off.....try again tomorrow
ID: 1178959 · Report as offensive
Profile SciManStev Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Jun 99
Posts: 6634
Credit: 121,090,076
RAC: 0
United States
Message 1178960 - Posted: 17 Dec 2011, 21:38:11 UTC - in response to Message 1178948.  



Well there are may reasons for dropping RAC's lately. I'd say the biggest reason is the flawed CreditNew scheme, which seems to give less and less credit/crunching hour the longer it is allowed to run.

My main reason was I just couldn't keep Piggy fed. The last three days have been excellent, and I am gaining ground very quickly! It takes over a day to catch RAC out of free fall, but it is really nice feeling the heat come off my rig, and being fed to the rest of the house. I did not buy enough oil to heat my house this year without the assistance of my rig. Crunching at full power (970 Watts from the tower, plus another 250 from the chiller)

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1178960 · Report as offensive
Profile rebest Project Donor
Volunteer tester
Avatar

Send message
Joined: 16 Apr 00
Posts: 1296
Credit: 45,357,093
RAC: 0
United States
Message 1178963 - Posted: 17 Dec 2011, 21:43:08 UTC - in response to Message 1178956.  
Last modified: 17 Dec 2011, 21:46:42 UTC

Wow. I take a few months off the boards and all h*** breaks loose.


Welcome back (8{)

So, are the problems the result of inconsistent availability of work, or is 6.12 a piece of crap?

I haven't "upgraded' to 6.12.xx, but most of the gripes seem to be about the 10x longer back-off times, which probably compound the usual distribution issues for the faster rigs.


Thanks for the info. The transfer retries are bad enough, but when you throw in a 7 or 8 hour project backoff on top of it....

I've set up for no new tasks. I'll clear my cache (which won't take long) and deep-six 6.12 and go back to 6.10.60.

So, what's this CreditNew thing about?

Join the PACK!
ID: 1178963 · Report as offensive
Profile SciManStev Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 20 Jun 99
Posts: 6634
Credit: 121,090,076
RAC: 0
United States
Message 1178968 - Posted: 17 Dec 2011, 22:02:33 UTC
Last modified: 17 Dec 2011, 22:04:15 UTC

Credit seems to be awarded now, based on the slowest host working on any given wu. If all fast hosts are your wingmen, then your credit will be lower. If you get a slower wingman, then your credit will be higher. With the recent limits of 50 per CPU, and 400 per GPU, it has been very tough for many rigs to keep fed. This coupled with a lot of shortie storms, has saturated the available bandwidth. That has caused hitting the retry button multiple times, over a long time to get 1, 2 minute work unit downloaded. Many rigs have continously run dry because of this.

The GPU Users Group has set up a system where the SETI staff has asked us for specific hardware, and we are doing fundraisers to get these specific items, and send them directly to Berkely. You can even use PayPal now to donate. My signature has the GPU Users group website in it if your are interested in donating.

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1178968 · Report as offensive
Cosmic_Ocean
Avatar

Send message
Joined: 23 Dec 00
Posts: 3027
Credit: 13,516,867
RAC: 13
United States
Message 1178973 - Posted: 17 Dec 2011, 22:30:18 UTC

You know, 6.2.19 has no problems getting work and doesn't have that project back-off "feature." You don't want it if you rely on GPUs though. :p
Linux laptop:
record uptime: 1511d 20h 19m (ended due to the power brick giving-up)
ID: 1178973 · Report as offensive
Profile rebest Project Donor
Volunteer tester
Avatar

Send message
Joined: 16 Apr 00
Posts: 1296
Credit: 45,357,093
RAC: 0
United States
Message 1179010 - Posted: 18 Dec 2011, 4:07:15 UTC - in response to Message 1178968.  

Credit seems to be awarded now, based on the slowest host working on any given wu. If all fast hosts are your wingmen, then your credit will be lower. If you get a slower wingman, then your credit will be higher. With the recent limits of 50 per CPU, and 400 per GPU, it has been very tough for many rigs to keep fed. This coupled with a lot of shortie storms, has saturated the available bandwidth. That has caused hitting the retry button multiple times, over a long time to get 1, 2 minute work unit downloaded. Many rigs have continously run dry because of this.

The GPU Users Group has set up a system where the SETI staff has asked us for specific hardware, and we are doing fundraisers to get these specific items, and send them directly to Berkely. You can even use PayPal now to donate. My signature has the GPU Users group website in it if your are interested in donating.

Steve


Hi, Steve. Thanks for the reply.

So, S@H has gone to full time, arbitrary workunit limits and the Cricket graph shows that we're MAXed out, as usual.

I have donated for years. As Mark will attest, I have also responded to the specific appeals for new equipment. However, it appears that very little is being done to address the bandwidth problem. New hard drives are nice, but I'll save my money for the day a plan is put forward that will finally punch a hole in the dam holding back the work at Berkeley.




Join the PACK!
ID: 1179010 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 64942
Credit: 55,293,173
RAC: 49
United States
Message 1179021 - Posted: 18 Dec 2011, 5:27:01 UTC - in response to Message 1179010.  

Credit seems to be awarded now, based on the slowest host working on any given wu. If all fast hosts are your wingmen, then your credit will be lower. If you get a slower wingman, then your credit will be higher. With the recent limits of 50 per CPU, and 400 per GPU, it has been very tough for many rigs to keep fed. This coupled with a lot of shortie storms, has saturated the available bandwidth. That has caused hitting the retry button multiple times, over a long time to get 1, 2 minute work unit downloaded. Many rigs have continously run dry because of this.

The GPU Users Group has set up a system where the SETI staff has asked us for specific hardware, and we are doing fundraisers to get these specific items, and send them directly to Berkely. You can even use PayPal now to donate. My signature has the GPU Users group website in it if your are interested in donating.

Steve


Hi, Steve. Thanks for the reply.

So, S@H has gone to full time, arbitrary workunit limits and the Cricket graph shows that we're MAXed out, as usual.

I have donated for years. As Mark will attest, I have also responded to the specific appeals for new equipment. However, it appears that very little is being done to address the bandwidth problem. New hard drives are nice, but I'll save my money for the day a plan is put forward that will finally punch a hole in the dam holding back the work at Berkeley.



Part of the problem is the DCF patch that Dr. A applied(which I've been told will not be unpatched) and yep 6.12.xx is crap, Yet Dr. A and Co I think want people to use 6.12.xx instead of the older and better 6.10.58...
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1179021 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1179028 - Posted: 18 Dec 2011, 7:14:40 UTC

One question. Where did the server named "maul" come from ?

T.A.
ID: 1179028 · Report as offensive
Profile Mike Special Project $75 donor
Volunteer tester
Avatar

Send message
Joined: 17 Feb 01
Posts: 33449
Credit: 79,922,639
RAC: 80
Germany
Message 1179039 - Posted: 18 Dec 2011, 8:54:16 UTC - in response to Message 1179028.  
Last modified: 18 Dec 2011, 8:54:43 UTC

One question. Where did the server named "maul" come from ?

T.A.


Maul was in the server closet long ago.
Just used for something different.
With each crime and every kindness we birth our future.
ID: 1179039 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51464
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1179076 - Posted: 18 Dec 2011, 16:27:17 UTC

Well, it's rather refreshing to have a break in the shorty storm. Most of my rigs have their cache limits hit, the only fly in the ointment.
But at least the tasks that are cached have some run time for the GPUs, not 80-90% 2 minute drills.

Now, about them there limits..........
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1179076 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 . . . 9 · Next

Message boards : Number crunching : Panic Mode On (63) Server problems?


 
©2022 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.