Hannah (Apr 16 2009)


log in

Advanced search

Message boards : Technical News : Hannah (Apr 16 2009)

1 · 2 · 3 · 4 . . . 6 · Next
Author Message
Profile Matt Lebofsky
Volunteer moderator
Project administrator
Project developer
Project scientist
Avatar
Send message
Joined: 1 Mar 99
Posts: 1388
Credit: 74,079
RAC: 0
United States
Message 885910 - Posted: 16 Apr 2009, 21:39:09 UTC

Slow steady progress since the last tech news item. The science database continues to be massaged into shape from the past month of nastiness. It's working, but some indexes are still missing, and some queries are taking longer than we'd like. Sometime, probably next week, I'll turn the science status page updates back on - until then the numbers are old and/or flat out wrong.

We're narrowing down the cause of our data recorder woes to either the SATA card or the system itself. We're trying the former first. A new one is on order and we'll have to get it configured remotely (which is a lot easier than configuring a whole new system remotely).

We're also finding that we don't have the processing power we'd like. It seems like we lost a lot of active users over the past few months. I blame the recession. You could also blame Astropulse, I guess. In any case, we need more people. We're hoping the 10th anniversary buzz will help. And speaking of that, Jeff and I are putting all focus on the NTPCkr, just so we have something fun/new/interesting to present in time for any p.r. blitz. That means very little effort in systems/upgrades/etc. for the next 5-6 weeks. Simply don't have the time/manpower.

Sorry about the lull in tech news items. I was on vacation visiting 23 relatives. Many are under 5 years old, which meant a lot of them have colds, which meant I got sick immediately upon my return, earlier in the week.

- Matt

____________
-- BOINC/SETI@home network/web/science/development person
-- "Any idiot can have a good idea. What is hard is to do it." - Jeanne-Claude

OzzFan
Volunteer tester
Avatar
Send message
Joined: 9 Apr 02
Posts: 13541
Credit: 29,267,575
RAC: 15,431
United States
Message 885917 - Posted: 16 Apr 2009, 22:20:53 UTC

Lost a bit of power from me; electric bill got too high and the summertime heat is coming up quickly in the Windy City.

If you guys are speculating that the advent of AP is causing a loss of users, do you plan on digging further into this, or is there a plan to do something about it? Personally, I like AP and I would have hoped that MB could be phased out eventually (perhaps by looking for broadband and narrowband in the same WU with a single application, perhaps at the detriment of longer crunching time).
____________

Speedy
Volunteer tester
Avatar
Send message
Joined: 26 Jun 04
Posts: 643
Credit: 5,323,756
RAC: 6,163
New Zealand
Message 885921 - Posted: 16 Apr 2009, 22:26:18 UTC

Thanks for the update I'm sorry to hear you got sick, I hope you are better. Do you need help with AP or MB work more or do you just need more crunchers in general?
____________

Live in NZ y not join Smile City?

Rudy
Volunteer tester
Send message
Joined: 23 Jun 99
Posts: 189
Credit: 563,724
RAC: 35
Canada
Message 885938 - Posted: 16 Apr 2009, 23:37:01 UTC - in response to Message 885910.

We're also finding that we don't have the processing power we'd like. It seems like we lost a lot of active users over the past few months. I blame the recession. You could also blame Astropulse, I guess. In any case, we need more people.
- Matt


You could consider issuing the optimized astropulse r112 as stock. It increases the crunch speed considerably. Most windows machines should have a mimimum of SSE3, and those that don't probably should not be running AP anyway. Releasing a new stock AP would also fix all those clients that got corrupted AP.exe in March and are still erroring out every midnight.

DJStarfox
Send message
Joined: 23 May 01
Posts: 1040
Credit: 532,447
RAC: 19
United States
Message 885960 - Posted: 17 Apr 2009, 0:29:07 UTC - in response to Message 885938.

We're also finding that we don't have the processing power we'd like. It seems like we lost a lot of active users over the past few months. I blame the recession. You could also blame Astropulse, I guess. In any case, we need more people.
- Matt


You could consider issuing the optimized astropulse r112 as stock. It increases the crunch speed considerably. Most windows machines should have a mimimum of SSE3, and those that don't probably should not be running AP anyway.


Sounds like a new BOINC scheduler/server feature for the BOINC development team: Scheduler can choose science applications not only based on architecture but also processor features (req'd list of features). Obviously, sticking with simple things such as MMX, SSE2, SSE3 would be appropriate for an initial implementation.

If you like the idea, I'll head over to the development site and create a Trac ticket.

Oh, and if this is already implemented, why not use it?

DJStarfox
Send message
Joined: 23 May 01
Posts: 1040
Credit: 532,447
RAC: 19
United States
Message 885963 - Posted: 17 Apr 2009, 0:30:29 UTC - in response to Message 885910.

Horray for 10th anniversary. Make it BIG!!! Confetii and the works!

mattman6
Send message
Joined: 4 Jan 03
Posts: 7
Credit: 162,193
RAC: 0
United States
Message 886004 - Posted: 17 Apr 2009, 3:30:03 UTC - in response to Message 885917.

Lost a bit of power from me; electric bill got too high and the summertime heat is coming up quickly in the Windy City.


Same story here. I only crunch in the winter since I don't mind the added heat my computer generates, but in the summer I try not to fight the air conditioner.

Profile Paul D. Buck
Volunteer tester
Send message
Joined: 19 Jul 00
Posts: 3898
Credit: 1,158,042
RAC: 0
United States
Message 886039 - Posted: 17 Apr 2009, 5:44:48 UTC

One of the reasons I dropped out is the way the CUDA application could trash the system ... and sending VLAR units to me. I spend 5-6 times as long on them as for normals ... sorry ... I get paid better at GPU Grid ...

The other reasons are legion, and as they are all well known if you have been paying attention ... well ... you do the math ... you get what you pay for and loyalty has to be earned ... and once you have abused people enough, well, loyalty is a a cast-iron <female dog> to regain ...

Speedy
Volunteer tester
Avatar
Send message
Joined: 26 Jun 04
Posts: 643
Credit: 5,323,756
RAC: 6,163
New Zealand
Message 886060 - Posted: 17 Apr 2009, 8:33:06 UTC
Last modified: 17 Apr 2009, 8:40:54 UTC

We're also finding that we don't have the processing power we'd like

Is there a place with statistics so we can see the amount of data that has been processed over the past day/week/month? If people could see the amount of data that has been processed this might increase Setis data through put. I'm aware that there is a piece on wu completed in the last hour.How many work units a there on each tape? Knowing this I can work out how many tapes are completed in an hour or what % of a tape has been completed in the hour.
____________

Live in NZ y not join Smile City?

Profile Gundolf Jahn
Send message
Joined: 19 Sep 00
Posts: 3184
Credit: 356,746
RAC: 15
Germany
Message 886062 - Posted: 17 Apr 2009, 8:42:44 UTC - in response to Message 886060.

Is there a place with statistics so we can see the amount of data that has been processed over the past day/week/month? If people could see the amount of data that has been processed this might increase Setis data through put. I'm aware that there is a peice on wu completes in the last hour.

Perhaps the "Seti@home In Progress" chart from the scarecrow graphs over 90 days?

Gruß,
Gundolf
____________
Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

Speedy
Volunteer tester
Avatar
Send message
Joined: 26 Jun 04
Posts: 643
Credit: 5,323,756
RAC: 6,163
New Zealand
Message 886066 - Posted: 17 Apr 2009, 9:46:31 UTC - in response to Message 886062.


Perhaps the "Seti@home In Progress" chart from the scarecrow graphs over 90 days?

Thank you I forgot about scarecrow graphs
____________

Live in NZ y not join Smile City?

PhonAcq
Send message
Joined: 14 Apr 01
Posts: 1622
Credit: 21,938,034
RAC: 3,735
United States
Message 886109 - Posted: 17 Apr 2009, 22:34:59 UTC

Well, from here, other threads, and scarecrow's graphs, it looks like CUDA is a net negative for the project (decreasing computational capacity overall). Maybe those CUDA wu's should be backed off a bit, if that's possible. Same goes for the AP wu's, since that has been chaotic as well, frankly. I guess I'm suggesting limiting the number per day available until the systems can be stablized, and then turn the other sub-projects on 'slowly'.

Badly behaving servers infuriate users; so in this perspective Paul Buck is correct.

Profile RottenMutt
Avatar
Send message
Joined: 15 Mar 01
Posts: 992
Credit: 207,654,623
RAC: 2
United States
Message 886112 - Posted: 17 Apr 2009, 22:41:46 UTC - in response to Message 886109.

what happened today, the cricket graphs dropped out and project died????
____________

Speedy
Volunteer tester
Avatar
Send message
Joined: 26 Jun 04
Posts: 643
Credit: 5,323,756
RAC: 6,163
New Zealand
Message 886154 - Posted: 18 Apr 2009, 0:56:18 UTC - in response to Message 886112.

what happened today, the cricket graphs dropped out and project died????

Here's what happened
message from front page
April 17, 2009
Our boinc mysql database crashed last night. We are down for database table repairs. Everything is up & running now
____________

Live in NZ y not join Smile City?

Speedy
Volunteer tester
Avatar
Send message
Joined: 26 Jun 04
Posts: 643
Credit: 5,323,756
RAC: 6,163
New Zealand
Message 886190 - Posted: 18 Apr 2009, 3:40:09 UTC

Admin/Matt are you aware that tape 18mr09aa is 0.00 GB in size, or is this 0.00 GB due to the boinc mysql database crash last night? Thankyou to the person/people that repaired the mysql database last night.
____________

Live in NZ y not join Smile City?

Profile Misfit
Volunteer tester
Avatar
Send message
Joined: 21 Jun 01
Posts: 21790
Credit: 2,510,901
RAC: 0
United States
Message 886200 - Posted: 18 Apr 2009, 4:11:07 UTC - in response to Message 886039.

One of the reasons I dropped out is the way the CUDA application could trash the system ... and sending VLAR units to me. I spend 5-6 times as long on them as for normals ... sorry ... I get paid better at GPU Grid ...

That's why the VLAR autokill mod is so great.
____________

zoom314
Avatar
Send message
Joined: 30 Nov 03
Posts: 45731
Credit: 36,373,060
RAC: 8,456
Message 886420 - Posted: 19 Apr 2009, 17:44:55 UTC - in response to Message 886200.

One of the reasons I dropped out is the way the CUDA application could trash the system ... and sending VLAR units to me. I spend 5-6 times as long on them as for normals ... sorry ... I get paid better at GPU Grid ...

That's why the VLAR autokill mod is so great.

I like this one much better: MB_6.08_mod_CUDA_V11_VLARKill_refined.rar

No more -6 errors and no more backoffs for VLAR kills, This V11 is good or at least I haven't seen any in a while, Although I did have one CUDA WU that clocked out at just above 5 hours.
____________

elgar
Send message
Joined: 21 May 99
Posts: 69
Credit: 2,687,478
RAC: 0
United States
Message 886437 - Posted: 19 Apr 2009, 18:08:33 UTC

"It seems like we lost a lot of active users over the past few months."

Wow. This statement speaks volumes. Are the people running SETI really that obtuse?

nickth
Volunteer tester
Send message
Joined: 21 Jan 07
Posts: 8
Credit: 822,571
RAC: 0
United Kingdom
Message 886441 - Posted: 19 Apr 2009, 18:14:03 UTC - in response to Message 886109.

PhonAcq said.....
Well, from here, other threads, and scarecrow's graphs, it looks like CUDA is a net negative for the project (decreasing computational capacity overall). Maybe those CUDA wu's should be backed off a bit, if that's possible. Same goes for the AP wu's, since that has been chaotic as well, frankly. I guess I'm suggesting limiting the number per day available until the systems can be stablized, and then turn the other sub-projects on 'slowly'.

Badly behaving servers infuriate users; so in this perspective Paul Buck is correct.



This can not be done because the same work units that run on the CUDA application also run on cpu application as well so you can not back it off unless you shut down the whole project.

And for limiting work units well at the moment i think its set at 100 per cpu and 400 per CUDA unit per day so yes that could bee reduced but i don't thing it will help.

Profile SATAN
Avatar
Send message
Joined: 27 Aug 06
Posts: 835
Credit: 2,094,367
RAC: 0
United Kingdom
Message 886477 - Posted: 19 Apr 2009, 20:36:45 UTC

I've greatly reduced my contribution due to the lack of Mac OS support, the system seemingly going into a state of meltown.

I can understand concentrating on NTPCKr, but if the system is unable to cope with what the project currently has running. How is it going to cope running another piece of software in real time?
____________

1 · 2 · 3 · 4 . . . 6 · Next

Message boards : Technical News : Hannah (Apr 16 2009)

Copyright © 2014 University of California