Panic Mode On (34) Server Problems

Message boards : Number crunching : Panic Mode On (34) Server Problems
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 5 · 6 · 7 · 8 · 9 · 10 · 11 · Next

AuthorMessage
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1008474 - Posted: 26 Jun 2010, 3:42:25 UTC - in response to Message 1008472.  
Last modified: 26 Jun 2010, 3:53:15 UTC

Not sure but it's funny that mine is only getting work for CPU, all my GPU requests have been turned down, been using the rescheduler to get GPU work.


This could be the reason. I know it's hard not to move units to the GPU's (been doing it myself) but it does screw up the allocations at the server end. I'm finding the rigs that have had units moved to the GPU's get a lot of CPU units and few if any GPU work. The rigs I've been able to keep my fingers out of get more GPU than CPU work but it's still only a few at a time.

Brodo

Oh that's nice to know, wonderful.

I just have a very warm QX6700 B1 cpu @ 3.4GHz on water and a GTX295 at stock speeds on air, It'll be interesting to see how much I get for each.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1008474 · 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: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1008475 - Posted: 26 Jun 2010, 3:51:11 UTC
Last modified: 26 Jun 2010, 3:51:33 UTC

Something has definitly changed. I have a multi-day cache, and my 480's are kicking out wu's in 7:02. Shorties in less than 2 minutes. I am pulling 745 watts at full throttle. CPU at 4.2339 GHz, 480's are at 830 MHz. It's go time!

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1008475 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1008479 - Posted: 26 Jun 2010, 3:57:50 UTC - in response to Message 1008475.  

Something has definitly changed. I have a multi-day cache, and my 480's are kicking out wu's in 7:02. Shorties in less than 2 minutes. I am pulling 745 watts at full throttle. CPU at 4.2339 GHz, 480's are at 830 MHz. It's go time!

Steve

Good deal Steve.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1008479 · Report as offensive
Profile soft^spirit
Avatar

Send message
Joined: 18 May 99
Posts: 6497
Credit: 34,134,168
RAC: 0
United States
Message 1008493 - Posted: 26 Jun 2010, 4:24:38 UTC

Still trying to fill up. At least I am not running idle now.
Janice
ID: 1008493 · 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: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1008502 - Posted: 26 Jun 2010, 4:48:04 UTC - in response to Message 1008493.  

Still trying to fill up. At least I am not running idle now.


I think now is a very good time to fill up. I am still downloading wu's. This is like it was before the changes. This is excellent!

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1008502 · Report as offensive
Profile skildude
Avatar

Send message
Joined: 4 Oct 00
Posts: 9541
Credit: 50,759,529
RAC: 60
Yemen
Message 1008504 - Posted: 26 Jun 2010, 4:49:32 UTC - in response to Message 1008502.  
Last modified: 26 Jun 2010, 4:55:45 UTC

no love for me and now work either

I stand corrected I got some earlier I just hadn't noticed and now I get a message

6/25/2010 11:39:46 PM SETI@home Message from server: This computer has finished a daily quota of 113 tasks)


that seems like a nice round number to me


In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope
ID: 1008504 · 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: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1008506 - Posted: 26 Jun 2010, 4:53:32 UTC

Mine started coming through after a full shut down and reboot. It couldn't hurt.

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1008506 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1008510 - Posted: 26 Jun 2010, 5:05:44 UTC

I'm waiting for Midnight, It's 10:05pm PDT as I type this, My PC might snag some, Maybe.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1008510 · Report as offensive
Profile soft^spirit
Avatar

Send message
Joined: 18 May 99
Posts: 6497
Credit: 34,134,168
RAC: 0
United States
Message 1008519 - Posted: 26 Jun 2010, 5:18:53 UTC - in response to Message 1008506.  

Mine started coming through after a full shut down and reboot. It couldn't hurt.

Steve


It was worth a shot. But it did not help.

the old friendly "Project has no tasks available".

Janice
ID: 1008519 · 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: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1008583 - Posted: 26 Jun 2010, 12:16:41 UTC

Well, after I went to sleep, my rig continued to download and adjust completion times. Now they are not too far off, and I am not about to run out of work. Before the recent changes went into effect, I was running a 6 day cache. Once my rig was drained, I switched to a 1 day cache so I could adapt to what ever happened. Still at empty, a few days later I switched to a 3 day cache. I was still empty, or with only a couple of hours work. My rig was tuned and waiting when I got the message of the extended outage. Many others were reporting that they were getting work, and of course some were not so I didn't think too much about it, and crunched for Einstein. Yesterday when I learned of the new scheduled outage, I thought what the heck, and set it for a 10 day cache, expecting to get a few wu's at best. I did a reboot, and tasks started coming in, in trucks! Every time I looked, more were downloading, and with all the trouble as of late, I wasn't about to stop it.

Now it looks like I have a 10 day cache, with completing times not to far off from reality. It may be a few days before my RAC settles out and starts to climb again, but I am very pleased that I am once again pumping out work like there is no tomorrow. So for this mega-cruncher, patience paid off and I am back in buisiness!

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1008583 · 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 1008591 - Posted: 26 Jun 2010, 13:06:01 UTC - in response to Message 1008583.  

I haven't messed with my cache yet. I'm afraid to. When this first started I was getting WUs that thought they would only take a couple of minutes but were really normal length. When I decided to do NNT it took me five days to run down what was supposed to be a three day cache. Since I did my detach the first WUs to come in thought they would take 16 hours each. They have now dropped to a guesstimate of 4 hours but that is still way high and both my CPUS and GPU are the same guesstimate (or close enough even though my GPU is at least twice as fast!). What I'm afraid of is hitting a bunch of shorties and throwing the to completion time way off in the other direction. I'm going to wait until I'm sure things have calmed down before I worry about maybe upping my cache.


PROUD MEMBER OF Team Starfire World BOINC
ID: 1008591 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1008620 - Posted: 26 Jun 2010, 14:56:12 UTC

The MB Splitters are now either not running or disabled.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1008620 · Report as offensive
Profile hiamps
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1008623 - Posted: 26 Jun 2010, 15:05:06 UTC - in response to Message 1008583.  

Well, after I went to sleep, my rig continued to download and adjust completion times. Now they are not too far off, and I am not about to run out of work. Before the recent changes went into effect, I was running a 6 day cache. Once my rig was drained, I switched to a 1 day cache so I could adapt to what ever happened. Still at empty, a few days later I switched to a 3 day cache. I was still empty, or with only a couple of hours work. My rig was tuned and waiting when I got the message of the extended outage. Many others were reporting that they were getting work, and of course some were not so I didn't think too much about it, and crunched for Einstein. Yesterday when I learned of the new scheduled outage, I thought what the heck, and set it for a 10 day cache, expecting to get a few wu's at best. I did a reboot, and tasks started coming in, in trucks! Every time I looked, more were downloading, and with all the trouble as of late, I wasn't about to stop it.

Now it looks like I have a 10 day cache, with completing times not to far off from reality. It may be a few days before my RAC settles out and starts to climb again, but I am very pleased that I am once again pumping out work like there is no tomorrow. So for this mega-cruncher, patience paid off and I am back in buisiness!

Steve

I can remember a time in the not so distant past that announcing you had a 10 day cache was like saying you were going to bring seti down, now it is all good. Funny how things change...
Official Abuser of Boinc Buttons...
And no good credit hound!
ID: 1008623 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1008626 - Posted: 26 Jun 2010, 15:09:06 UTC - in response to Message 1008623.  

Well, after I went to sleep, my rig continued to download and adjust completion times. Now they are not too far off, and I am not about to run out of work. Before the recent changes went into effect, I was running a 6 day cache. Once my rig was drained, I switched to a 1 day cache so I could adapt to what ever happened. Still at empty, a few days later I switched to a 3 day cache. I was still empty, or with only a couple of hours work. My rig was tuned and waiting when I got the message of the extended outage. Many others were reporting that they were getting work, and of course some were not so I didn't think too much about it, and crunched for Einstein. Yesterday when I learned of the new scheduled outage, I thought what the heck, and set it for a 10 day cache, expecting to get a few wu's at best. I did a reboot, and tasks started coming in, in trucks! Every time I looked, more were downloading, and with all the trouble as of late, I wasn't about to stop it.

Now it looks like I have a 10 day cache, with completing times not to far off from reality. It may be a few days before my RAC settles out and starts to climb again, but I am very pleased that I am once again pumping out work like there is no tomorrow. So for this mega-cruncher, patience paid off and I am back in buisiness!

Steve

I can remember a time in the not so distant past that announcing you had a 10 day cache was like saying you were going to bring seti down, now it is all good. Funny how things change...

I went for a 9 Day cache instead.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1008626 · 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: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1008628 - Posted: 26 Jun 2010, 15:16:00 UTC - in response to Message 1008623.  

I never tried it before now, and even then only because nothing was working any way. I think my impact on the system was minimal, and although I don't really know how long it will last, the fact that the splitters are down has become transparent to me. I can concentrate on learning, and doing what I can to make things better. The hard part is that I work an hour away. I leave home at 6:00 AM and get home at 5:30 or 6:00 in the evening. I'm finding it a bit hard to hit the books and continue learning code when the refrigerator is close by. Something about relaxing after work with a cold beer is just magical. Weekends are spent doing all the work I didn't have time for during the week. Things are settling down now, and I will make more time available. Now where did I put that refrigerator??

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1008628 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65709
Credit: 55,293,173
RAC: 49
United States
Message 1008662 - Posted: 26 Jun 2010, 16:53:45 UTC - in response to Message 1008628.  

I never tried it before now, and even then only because nothing was working any way. I think my impact on the system was minimal, and although I don't really know how long it will last, the fact that the splitters are down has become transparent to me. I can concentrate on learning, and doing what I can to make things better. The hard part is that I work an hour away. I leave home at 6:00 AM and get home at 5:30 or 6:00 in the evening. I'm finding it a bit hard to hit the books and continue learning code when the refrigerator is close by. Something about relaxing after work with a cold beer is just magical. Weekends are spent doing all the work I didn't have time for during the week. Things are settling down now, and I will make more time available. Now where did I put that refrigerator??

Steve

Maybe You mailed It somewhere? Like to one of the Ice Giants. ;)
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1008662 · Report as offensive
Josef W. Segur
Volunteer developer
Volunteer tester

Send message
Joined: 30 Oct 99
Posts: 4504
Credit: 1,414,761
RAC: 0
United States
Message 1008802 - Posted: 27 Jun 2010, 1:18:09 UTC - in response to Message 1008623.  

...
Now it looks like I have a 10 day cache, with completing times not to far off from reality. It may be a few days before my RAC settles out and starts to climb again, but I am very pleased that I am once again pumping out work like there is no tomorrow. So for this mega-cruncher, patience paid off and I am back in buisiness!

Steve

I can remember a time in the not so distant past that announcing you had a 10 day cache was like saying you were going to bring seti down, now it is all good. Funny how things change...

Like most choices, it's neither all good nor all bad. Having all S@H-only hosts go idle next Wednesday (as the BOINC defaults would do) would be a waste. OTOH, if all hosts were set to 10 day caches it would be disastrous. With luck staff would disable some of the servers before the whole system locked up and then put in some protective measure; what BOINC offers is a project config setting which simply limits the maximum number of tasks on each host. Then all hosts for which that limit was more than a 10 day cache would be unaffected but the most productive hosts would be severely limited. That's an impossible scenario of course, only a small fraction of users will even know about the new outage schedule and only some of those will react by increasing cache.

What any individual does to adapt to the changing circumstances is strictly their business, in my mind there's no blame unless they recommend that everyone else do the same and if that advice were widely accepted it would be harmful. A couple of years ago there was a user with one of the top hosts who was cherry-picking work by aborting what earned less credit per time on that host; I considered that slightly anti-social but not a heinous crime. I hope that some day BOINC will be mature enough to deliver the most suitable work to each host; I'd be very happy if Sten-Arne's ION would do the VHARs which cause my Willamette P4's DCF to increase by about 50%, for instance.
                                                                Joe
ID: 1008802 · 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: 6651
Credit: 121,090,076
RAC: 0
United States
Message 1008806 - Posted: 27 Jun 2010, 1:46:42 UTC - in response to Message 1008802.  

I appreciate what you said. I set mine as an experiment, and was quite supprised that it, or something worked. I don't know exactly how long it will last, but there is certainly a lot of work on my rig. I used to happily run a six day cache, and I don't mind reverting. I am hopeful that work will be available for everyone before and after the outage.

Steve
Warning, addicted to SETI crunching!
Crunching as a member of GPU Users Group.
GPUUG Website
ID: 1008806 · Report as offensive
FiveHamlet
Avatar

Send message
Joined: 5 Oct 99
Posts: 783
Credit: 32,638,578
RAC: 0
United Kingdom
Message 1008967 - Posted: 27 Jun 2010, 19:17:59 UTC
Last modified: 27 Jun 2010, 19:20:55 UTC

England bit the dust so time to take the flag down.
There we go.
Dave
ID: 1008967 · Report as offensive
Profile hiamps
Volunteer tester
Avatar

Send message
Joined: 23 May 99
Posts: 4292
Credit: 72,971,319
RAC: 0
United States
Message 1009102 - Posted: 28 Jun 2010, 4:54:55 UTC

Oh no, starting to see weird connection notices and the boards are on and off...or my internet is going down....
Official Abuser of Boinc Buttons...
And no good credit hound!
ID: 1009102 · Report as offensive
Previous · 1 . . . 5 · 6 · 7 · 8 · 9 · 10 · 11 · Next

Message boards : Number crunching : Panic Mode On (34) 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.