Message boards :
Number crunching :
Can we agree on this?..........
Message board moderation
Author | Message |
---|---|
ksnash Send message Joined: 28 Nov 99 Posts: 402 Credit: 528,725 RAC: 0 |
It was not smart to put one variable for cache size and connect time. |
Timothy Maness Send message Joined: 3 Apr 99 Posts: 159 Credit: 4,929 RAC: 0 |
I think that's a consensus, but I could be wrong. <img border="0" src="http://boinc.mundayweb.com/one/stats.php?userID=567" /> <a href="http://boinc-doc.net/boinc-wiki/index.php?title=Main_Page">BOINC Wiki</a> <a href="http://www.tamaness.hostrocket.com/BOINC_team.html">Gideontech BOINC team!</a> |
ric Send message Joined: 16 Jun 03 Posts: 482 Credit: 666,047 RAC: 0 |
Can we agree on this?........... Why not? One definition of "successfully" boinccing could be having "all the time" enough work to keep the clients busy. If this targed can be reached with 1 or two or 99 settings, it's in my point of view, not so important. Usualy as less datas/tags = as less complicated. I'm ready for the try. happy chrunching <img src=\"http://www.setisynergy.com/images/stats/comb-207.jpg\"> multi Project BOINC Stats <a href=\"http://www.setisynergy.com/stats/teams.php\">Team</a> <a href=\"http://www.setisynergy.com/stats/individuals.php\">Individual</a> |
Tom Merrick Send message Joined: 10 Jun 03 Posts: 13 Credit: 4,924,150 RAC: 0 |
Also, cache size should be by project. Right now connect time is global - and it should be. But I would like to specify 7 days for seti and only 2 days for einstein. This is not only for the optimized clients, but also for the report deadlines which are much longer for seti than einstein. |
Red Squirrel Send message Joined: 3 Apr 99 Posts: 15 Credit: 4,395 RAC: 0 |
Also, cache size should be by project. Right now connect time is global - and it should be. But I would like to specify 7 days for seti and only 2 days for einstein. This is not only for the optimized clients, but also for the report deadlines which are much longer for seti than einstein. But that's what Resource Share is for. |
Tom Merrick Send message Joined: 10 Jun 03 Posts: 13 Credit: 4,924,150 RAC: 0 |
Also, cache size should be by project. Right now connect time is global - and it should be. But I would like to specify 7 days for seti and only 2 days for einstein. This is not only for the optimized clients, but also for the report deadlines which are much longer for seti than einstein. Not really. The resource share is mainly CPU and ESTIMATED quota. For example. I have the connect time set to 2 days. So it downloads about 4 7.5 hour Einstein jobs. It also downloads around 5 5.5 hour SETI jobs with my resource share set to 50/50. If you see from my credits I get about the same from both projects. But the SETI jobs finish in 3.5 hours with an unoptimized client and 1.5 hours with the optimized client. So I really only have about 7.5 hours of SETI on my computer at any time. If I increase my connect time any more I get into problems with the early dead lines in Einstein. One other interesting fact I have noticed is when Einstein goes into panic mode it will crunch all of the jobs and put it in no download state until SETI catches up. This causes problems because when SETI catches up Einstein then downloads ALL of the data sets that it needs at once with the same 7 day deadline rather than spacing them out some. So about 3 days later it goes into panic mode again due to the double connect time deadline approaching. |
Jim_ Send message Joined: 29 May 01 Posts: 14 Credit: 71,678 RAC: 0 |
If I increase my connect time any more I get into problems with the early dead lines in Einstein. One other interesting fact I have noticed is when Einstein goes into panic mode it will crunch all of the jobs and put it in no download state until SETI catches up. This causes problems because when SETI catches up Einstein then downloads ALL of the data sets that it needs at once with the same 7 day deadline rather than spacing them out some. So about 3 days later it goes into panic mode again due to the double connect time deadline approaching. What was the connect time value when you saw the scheduler switch to panic mode? Edit: How does the time estimate for Einstein compare to actual? I see 8.5 hour estimates for Einstein with actual time being about 11 hours. |
W-K 666 Send message Joined: 18 May 99 Posts: 19048 Credit: 40,757,560 RAC: 67 |
OSU I can see your problem but if they had two options one for connect time and one for WU cache, you could keep your 2 day connection time and increase your cache to 4 days of work. Then it wouldn't go into panic mode because of the 2 * connect time rule. Andy |
Tom Merrick Send message Joined: 10 Jun 03 Posts: 13 Credit: 4,924,150 RAC: 0 |
[quoteWhat was the connect time value when you saw the scheduler switch to panic mode?[/quote] It was still 2 days
The recent ones have been about right. I have seen some very bad ones at times (> 20 hours). Tom |
©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.