Panic Mode On (55) Server problems?

Message boards : Number crunching : Panic Mode On (55) Server problems?
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 4 · 5 · 6 · 7 · 8 · 9 · 10 · Next

AuthorMessage
Profile Sunny129
Avatar

Send message
Joined: 7 Nov 00
Posts: 190
Credit: 3,163,755
RAC: 0
United States
Message 1155634 - Posted: 24 Sep 2011, 13:58:01 UTC

well it appears that there are 2 distinct groups of people - those who are getting work, and those who aren't - but EVERYONE seems to fall under the "i have no idea what's going on" category. personally, i wish i fell into the category of those who are getting work, even if they generally have no idea how or why they're getting work...but unfortunately i'm in the other category...
ID: 1155634 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65858
Credit: 55,293,173
RAC: 49
United States
Message 1155636 - Posted: 24 Sep 2011, 14:03:53 UTC

I've been getting this and I'm all out of WU's...

1513 SETI@home 9/24/2011 1:55:09 AM Message from server: Your app_info.xml file doesn't have a usable version of Astropulse v505.
And I wish S@H would stop pushing Astro, I'm not interested in Astro.
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1155636 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155637 - Posted: 24 Sep 2011, 14:15:28 UTC - in response to Message 1155636.  

I've been getting this and I'm all out of WU's...

1513 SETI@home 9/24/2011 1:55:09 AM Message from server: Your app_info.xml file doesn't have a usable version of Astropulse v505.
And I wish S@H would stop pushing Astro, I'm not interested in Astro.

Kwitcherbellyachin' about AP....
You don't have an app loaded for it, you won't get any.
The kitties are ready for any Seti work that the servers will issue, and will gladly crunch it.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155637 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14656
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1155645 - Posted: 24 Sep 2011, 15:22:20 UTC - in response to Message 1155631.  

And my machines still have over 300 GPU units each and have downloaded more during the night.

Don't know what is going on.

I know what's going on.

1st....DA's Boinc server code faux pas.
2nd....splitter creation rate is way too low, and not ramping up.
3rd....cache limits are not working properly.
4th....shorty storm is killing us.

The perfect storm.

Only question is when somebody is gonna fix the first three.
The fourth is out of our control.

And 5th....the server storage areas are full of un-assimilated WUs.

My reading of it is:

#4 is causing #5
#5 is causing #2
#1 has to be fixed before #3

WRT #3 - I wonder if being above the limit for either work type prevents you getting work for both CPU and GPU?

I'm trying to test the boundaries here - up to 337 tasks in progress (plus one waiting to report), on a box which is requesting GPU work only, no SETI on the CPUs. The last assignment was 36 tasks, about 20 minutes ago.

Also, I wonder whether the 50-for-CPU limit is AP and MB separately, or both added together?

If you are below 400 GPU, but above 50 CPU, I wonder if it would help if you turned off 'Use CPU' in your web preferences for a couple of work fetch cycles? The first time you request work after that, the CPU will be told "don't ask again": the second time the computer will request GPU work only, and might - just might - get some. I'm not getting work every time I ask, so a bit of patience would be necessary - and it may not work at all.

But possibly worth a try?

[Note: the limits should be separate, but could easily have got mis-applied during the current FUBAR.
Setting 'Don't use CPU doesn't affect work already downloaded, only work fetch]
ID: 1155645 · 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 1155647 - Posted: 24 Sep 2011, 15:33:22 UTC - in response to Message 1155631.  

...
2nd....splitter creation rate is way too low, and not ramping up.
...

I suspect they've implemented something like a just in time production control. Aside from the notches at regular two hour intervals, the Cricket graph shows about as much download as is possible. We've seen those notches before, I guess they reflect some chron job which runs at those time, perhaps a reinitialization of some of the server code.

If that kind of control of the splitters is being used, it would be natural that there would be few units in the stockroom (Results ready to send). Not having to maintain a big stockroom is why that method reduces costs.
                                                                   Joe
ID: 1155647 · Report as offensive
Dave

Send message
Joined: 29 Mar 02
Posts: 778
Credit: 25,001,396
RAC: 0
United Kingdom
Message 1155650 - Posted: 24 Sep 2011, 15:45:12 UTC

Gone all scary quiet in here again. Makes life interesting.
ID: 1155650 · Report as offensive
W-K 666 Project Donor
Volunteer tester

Send message
Joined: 18 May 99
Posts: 19147
Credit: 40,757,560
RAC: 67
United Kingdom
Message 1155653 - Posted: 24 Sep 2011, 16:09:56 UTC - in response to Message 1155645.  

WRT#3 I can, with 99% certainty, say that any of the limits stops all d/loads.
For the last couple of days I have never had more that 150 in progress but have frequently had the limit message.
ID: 1155653 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155666 - Posted: 24 Sep 2011, 17:02:30 UTC - in response to Message 1155645.  

And my machines still have over 300 GPU units each and have downloaded more during the night.

Don't know what is going on.

I know what's going on.

1st....DA's Boinc server code faux pas.
2nd....splitter creation rate is way too low, and not ramping up.
3rd....cache limits are not working properly.
4th....shorty storm is killing us.

The perfect storm.

Only question is when somebody is gonna fix the first three.
The fourth is out of our control.

And 5th....the server storage areas are full of un-assimilated WUs.

My reading of it is:

#4 is causing #5
#5 is causing #2
#1 has to be fixed before #3

WRT #3 - I wonder if being above the limit for either work type prevents you getting work for both CPU and GPU?

I'm trying to test the boundaries here - up to 337 tasks in progress (plus one waiting to report), on a box which is requesting GPU work only, no SETI on the CPUs. The last assignment was 36 tasks, about 20 minutes ago.

Also, I wonder whether the 50-for-CPU limit is AP and MB separately, or both added together?

If you are below 400 GPU, but above 50 CPU, I wonder if it would help if you turned off 'Use CPU' in your web preferences for a couple of work fetch cycles? The first time you request work after that, the CPU will be told "don't ask again": the second time the computer will request GPU work only, and might - just might - get some. I'm not getting work every time I ask, so a bit of patience would be necessary - and it may not work at all.

But possibly worth a try?

[Note: the limits should be separate, but could easily have got mis-applied during the current FUBAR.
Setting 'Don't use CPU doesn't affect work already downloaded, only work fetch]

Any ideas are welcome about now, Richard, but I see that many of the work requests are already GPU only rather that GPU and CPU...and they are still garnering few downloads.

The few work requests that are successful often get only one WU in response.
I suspect the other Bonic 'safety' when the DCF gets too 'good'.
I have held off diddling with the flops BS...because I think we all were believing that this debacle would have been fixed by now.
Why the foot dragging and leaving known broken code in place is going on, I can only guess.

"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155666 · Report as offensive
AndyJ
Avatar

Send message
Joined: 17 Aug 02
Posts: 248
Credit: 27,380,797
RAC: 0
United Kingdom
Message 1155668 - Posted: 24 Sep 2011, 17:10:04 UTC

Hello,
Just a note to let you all know AndyJ is not very well right now. I am his nurse, and he has asked me to tell you, because you are the only friends he has left.
When I (and those very large burly guys) caught up with him, (quite a chase he gave us too!) and got the axe off him, and eventually got him to put some clothes on, he explained that he was looking for Anakin Bane and Vader. Anybody know who these people are?
Anyway, AndyJ has had a nice big blue pill, and is quiet for the moment.
Regards,
AndyJ`s Nurse.
ID: 1155668 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155670 - Posted: 24 Sep 2011, 17:12:24 UTC - in response to Message 1155668.  

Hello,
Just a note to let you all know AndyJ is not very well right now. I am his nurse, and he has asked me to tell you, because you are the only friends he has left.
When I (and those very large burly guys) caught up with him, (quite a chase he gave us too!) and got the axe off him, and eventually got him to put some clothes on, he explained that he was looking for Anakin Bane and Vader. Anybody know who these people are?
Anyway, AndyJ has had a nice big blue pill, and is quiet for the moment.
Regards,
AndyJ`s Nurse.

Best wishes for AndyJ....
At least he is not in panic land like some of us...LOL.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155670 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65858
Credit: 55,293,173
RAC: 49
United States
Message 1155674 - Posted: 24 Sep 2011, 17:17:53 UTC - in response to Message 1155668.  
Last modified: 24 Sep 2011, 17:18:07 UTC

Hello,
Just a note to let you all know AndyJ is not very well right now. I am his nurse, and he has asked me to tell you, because you are the only friends he has left.
When I (and those very large burly guys) caught up with him, (quite a chase he gave us too!) and got the axe off him, and eventually got him to put some clothes on, he explained that he was looking for Anakin Bane and Vader. Anybody know who these people are?
Anyway, AndyJ has had a nice big blue pill, and is quiet for the moment.
Regards,
AndyJ`s Nurse.

Anakin Bane and Vader are the name of servers(computers) here at Seti@Home, right now something needs attention there and the repair guy is I guess out. Welcome to Seti, Say Hi to AndyJ for US will You please?
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1155674 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155675 - Posted: 24 Sep 2011, 17:19:43 UTC - in response to Message 1155674.  

Hello,
Just a note to let you all know AndyJ is not very well right now. I am his nurse, and he has asked me to tell you, because you are the only friends he has left.
When I (and those very large burly guys) caught up with him, (quite a chase he gave us too!) and got the axe off him, and eventually got him to put some clothes on, he explained that he was looking for Anakin Bane and Vader. Anybody know who these people are?
Anyway, AndyJ has had a nice big blue pill, and is quiet for the moment.
Regards,
AndyJ`s Nurse.

Anakin Bane and Vader are the name of servers(computers) here at Seti@Home, right now something needs attention there and the repair guy is I guess out. Welcome to Seti, Say Hi to AndyJ for US will You please?

And when he wakes up, tell him the guy that he should be looking for is named
Anakin Anderson.

"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155675 · Report as offensive
Profile zoom3+1=4
Volunteer tester
Avatar

Send message
Joined: 30 Nov 03
Posts: 65858
Credit: 55,293,173
RAC: 49
United States
Message 1155679 - Posted: 24 Sep 2011, 17:26:51 UTC - in response to Message 1155675.  

Hello,
Just a note to let you all know AndyJ is not very well right now. I am his nurse, and he has asked me to tell you, because you are the only friends he has left.
When I (and those very large burly guys) caught up with him, (quite a chase he gave us too!) and got the axe off him, and eventually got him to put some clothes on, he explained that he was looking for Anakin Bane and Vader. Anybody know who these people are?
Anyway, AndyJ has had a nice big blue pill, and is quiet for the moment.
Regards,
AndyJ`s Nurse.

Anakin Bane and Vader are the name of servers(computers) here at Seti@Home, right now something needs attention there and the repair guy is I guess out. Welcome to Seti, Say Hi to AndyJ for US will You please?

And when he wakes up, tell him the guy that he should be looking for is named
Anakin Anderson.

Shouldn't that be Darth Anderson? ;)
The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's
ID: 1155679 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155682 - Posted: 24 Sep 2011, 17:29:47 UTC - in response to Message 1155679.  

Hello,
Just a note to let you all know AndyJ is not very well right now. I am his nurse, and he has asked me to tell you, because you are the only friends he has left.
When I (and those very large burly guys) caught up with him, (quite a chase he gave us too!) and got the axe off him, and eventually got him to put some clothes on, he explained that he was looking for Anakin Bane and Vader. Anybody know who these people are?
Anyway, AndyJ has had a nice big blue pill, and is quiet for the moment.
Regards,
AndyJ`s Nurse.

Anakin Bane and Vader are the name of servers(computers) here at Seti@Home, right now something needs attention there and the repair guy is I guess out. Welcome to Seti, Say Hi to AndyJ for US will You please?

And when he wakes up, tell him the guy that he should be looking for is named
Anakin Anderson.

Shouldn't that be Darth Anderson? ;)

LOL...quite so.

"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155682 · Report as offensive
Profile Dimly Lit Lightbulb 😀
Volunteer tester
Avatar

Send message
Joined: 30 Aug 08
Posts: 15399
Credit: 7,423,413
RAC: 1
United Kingdom
Message 1155689 - Posted: 24 Sep 2011, 17:38:56 UTC - in response to Message 1155645.  

*SNIP*
Also, I wonder whether the 50-for-CPU limit is AP and MB separately, or both added together?

*/SNIP*

I'm CPU only and have both MB and just a few APs. Total number of tasks on board is staying at 50.
ID: 1155689 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155690 - Posted: 24 Sep 2011, 17:41:51 UTC - in response to Message 1155689.  

*SNIP*
Also, I wonder whether the 50-for-CPU limit is AP and MB separately, or both added together?

*/SNIP*

I'm CPU only and have both MB and just a few APs. Total number of tasks on board is staying at 50.

I don't freakin' know...
My top rig has 73 total tasks on board....all except 3 or 4 for the CPU...
And it does get a single GPU WU now and again.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155690 · Report as offensive
Terror Australis
Volunteer tester

Send message
Joined: 14 Feb 04
Posts: 1817
Credit: 262,693,308
RAC: 44
Australia
Message 1155694 - Posted: 24 Sep 2011, 17:45:26 UTC - in response to Message 1155690.  

And it does get a single GPU WU now and again.


I am so friggin' jealous. Why can't I be as lucky as you :-S

T.A.
ID: 1155694 · Report as offensive
Richard Haselgrove Project Donor
Volunteer tester

Send message
Joined: 4 Jul 99
Posts: 14656
Credit: 200,643,578
RAC: 874
United Kingdom
Message 1155695 - Posted: 24 Sep 2011, 17:46:28 UTC

Well, on my GPU-only test rig, I've just managed to trigger the dreaded "reached a limit of tasks in progress" with a request for new work when there were exactly 400 GPU tasks in progress, and no work being reported.

I'll try bumping my head on the glass ceiling a couple of times as work finishes and gets reported, then go and add a CPU into the mix and see what happens then.
ID: 1155695 · Report as offensive
Dave

Send message
Joined: 29 Mar 02
Posts: 778
Credit: 25,001,396
RAC: 0
United Kingdom
Message 1155700 - Posted: 24 Sep 2011, 17:52:45 UTC

This is akin to thermal cycling. More hand-to-mouth than a hand-to-mouth thing.
ID: 1155700 · Report as offensive
kittyman Crowdfunding Project Donor*Special Project $75 donorSpecial Project $250 donor
Volunteer tester
Avatar

Send message
Joined: 9 Jul 00
Posts: 51469
Credit: 1,018,363,574
RAC: 1,004
United States
Message 1155701 - Posted: 24 Sep 2011, 17:56:39 UTC - in response to Message 1155700.  

This is akin to thermal cycling. More hand-to-mouth than a hand-to-mouth thing.

Right now Boinc is acting more like a brain retention module in anterior region thing.
"Freedom is just Chaos, with better lighting." Alan Dean Foster

ID: 1155701 · Report as offensive
Previous · 1 . . . 4 · 5 · 6 · 7 · 8 · 9 · 10 · Next

Message boards : Number crunching : Panic Mode On (55) 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.