Message boards :
Number crunching :
to set up a Cache for BOINC SAH
Message board moderation
Author | Message |
---|---|
Sir Ulli Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 |
as on the last Problems with cogent it is time that you use the in build in Programms at BOINC to use a Cache what is a Cahe and who will i do this go to your Account, then View or edit general preferences, and under Network uage there is one Point Connect to network about every xxx Days where xx is somethink between 1-10, etermines size of work cache; maximum 10 days i think at Number between 3 or 5 is a good one and than you have no Probs if the Seti Servers are down for an Upgrade of for spezial Problems only for Info, and the People who are crying.... and not ...... Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © |
Pascal, K G Send message Joined: 3 Apr 99 Posts: 2343 Credit: 150,491 RAC: 0 |
Poor babies, I will maitain my 1 Day setting and not worry about it....... Semper Eadem So long Paul, it has been a hell of a ride. Park your ego's, fire up the computers, Science YES, Credits No. |
jimmyhua Send message Joined: 16 Apr 05 Posts: 97 Credit: 369,588 RAC: 0 |
Slowly working up to it. I started off with the default setting of 0.1 days. Upped it to 1 day. In the last outtage, one day was enough, but after servers came back up, it seemed like my settings got reset back to 0.1 days. I reset back to 1 day. This latest one, I've run out of work... Even though seti is not back up yet, I have set my que to 3 days. Hopefully it will work right when everything comes back up. Jimmy |
Captain Avatar Send message Joined: 17 May 99 Posts: 15133 Credit: 529,088 RAC: 0 |
> Poor babies, I will maitain my 1 Day setting and not worry about it....... > Back to Mr. Condescending our we Pascal? |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
The only way to guarantee that you can get past a server outage is to join several BOINC projects. Some of the projects have been down for more than a month, and with a deadlien of less than two weeks, there was no way that anyone coud have a queue long enough. There is also going to come a time (probably shortly after classic is shut down) that there will not be enough work to go around from S@H. The only way past is to join several BOINC projects. BOINC WIKI |
Sir Ulli Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 |
> The only way to guarantee that you can get past a server outage is to join > several BOINC projects. Some of the projects have been down for more than a > month, and with a deadlien of less than two weeks, there was no way that > anyone coud have a queue long enough. > > There is also going to come a time (probably shortly after classic is shut > down) that there will not be enough work to go around from S@H. The only way > past is to join several BOINC projects. > if you have a Cache for BOINC SAH i think not with a 3 Day Cache you can look at thit with a Smile.... Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
Hi I´m sorry John but thats not true. I´m running Seti over 4 years now and am boincing since 2003 and running only two times out of work with my 6-7 days of cache. I have einstein running only with 9% of share. Still have 35 seti units and 3 einstein units in cache. respectfully Mike With each crime and every kindness we birth our future. |
Prognatus Send message Joined: 6 Jul 99 Posts: 1600 Credit: 391,546 RAC: 0 |
|
Saenger Send message Joined: 3 Apr 99 Posts: 2452 Credit: 33,281 RAC: 0 |
I strongly support Sir Ulli regarding cache! BUT... Be aware of two things: 1. Cache is determined with the estimated crunching time, so for Seti-only on my puters it will always give me too little. (est: 6.6h, done:~4h). But some other projects estimate quite correct, like Einstein. 2. If you use the outdated 4.19, Boinc determines the amout for each project as if it's running solo. So for 4 projects and 5 days you'll get 20 days worth of crunching (minus the time concerning #1 ;) Gruesse vom Saenger For questions about Boinc look in the BOINC-Wiki |
Mike Send message Joined: 17 Feb 01 Posts: 34258 Credit: 79,922,639 RAC: 80 |
<blockquote>I strongly support Sir Ulli regarding cache! BUT... Be aware of two things: 1. Cache is determined with the estimated crunching time, so for Seti-only on my puters it will always give me too little. (est: 6.6h, done:~4h). But some other projects estimate quite correct, like Einstein. 2. If you use the outdated 4.19, Boinc determines the amout for each project as if it's running solo. So for 4 projects and 5 days you'll get 20 days worth of crunching (minus the time concerning #1 ;)</blockquote> The estimated time at einstein is oposite to setis. Seti eta 4.04 real 3.xx. Einstein eta 5.05 real 6.xx. My broteher is running 4.19 with a cache setting at 9 days and getting never more than 30 units what is enough for 3 days, so anything is wrong. greetz Mike With each crime and every kindness we birth our future. |
Dwarlock Send message Joined: 3 Apr 99 Posts: 27 Credit: 156,907 RAC: 0 |
I do a 6 day cache, mostly because I have some isolated systems that I have to run cable to manualy for updates. However with the recent number of noisy work units, I have had one system run through all 6 days in about 2 in the past. <img src="http://teamstarfire.org/boinc/summary.php?name=Dwarlock&team=Kansans%20Searching%20for%20a%20Planet%20With%20People%20More%20Interesting%20Than%20This%20One&seti=7832952&cbg=grey"> |
KB7RZF Send message Joined: 15 Aug 99 Posts: 9549 Credit: 3,308,926 RAC: 2 |
I've got my pref's set to 10 days, gives me 36 wu's, and my machine usually processes 1.5-2 wu's a day, so thats not too bad I guess. :-) Just my 2 cents worth. |
Steve Cressman Send message Joined: 6 Jun 02 Posts: 583 Credit: 65,644 RAC: 0 |
1.5 WUs * 14 days = 21 WUs I suggest you lower your 10 days to 5 or 6 days to avoid missing deadlines. 98SE XP2500+ @ 2.1 GHz Boinc v5.8.8 And God said"Let there be light."But then the program crashed because he was trying to access the 'light' property of a NULL universe pointer. |
Timcom99 Send message Joined: 30 Sep 04 Posts: 105 Credit: 8,927,290 RAC: 0 |
I like to keep a 5 Day Cache of Work Units. With 5 days I have never run out of Work during these outages. I usually can Crunch the 5 Days they give me in 3 1/2 Days cause their Bench Marks Under Rate my Computer. I doubt there will be a Shortage of Work Units to go around when the SETI Classic Crunchers move to SETI BOINC. They are Crunching the same Work Units now just using a Different Software. The Work Units will get sent out a Fewer number of times with SETI BOINC than with SETI Classic however. Soon hopefully the Southern Watch with the Parks Telescope will come online and we will have Tons More Work Units to Crunch. |
John McLeod VII Send message Joined: 15 Jul 99 Posts: 24806 Credit: 790,712 RAC: 0 |
May I remind everyone that there was a 2 week outage in classic when some nitwit stole the cable to the Space Science Lab. Also PP@H was down for 4 months last fall. LHC was also down for 4 months or so late last year and early this year. S@H BOINC will only give out work enough times to get the science done. If they run out of work, there will not be work available. So far, Berkeley is keeping ahead, but that may not continue forever. BOINC WIKI |
Bruno G. Olsen & ESEA @ greenholt Send message Joined: 15 May 99 Posts: 875 Credit: 4,386,984 RAC: 0 |
Yup, cache is good, but not the universal solution for all contingancies (propably misspelled, but you know what I mean ;)) Running two projects is getting close, but as I recall this may not even get you there. At one point three projects were down at the same time, all for different reasons and for different durations. |
Sir Ulli Send message Joined: 21 Oct 99 Posts: 2246 Credit: 6,136,250 RAC: 0 |
as on the Berkeley Outage, i push this thread Greetings from Germany NRW Ulli S@h Berkeley's Staff Friends Club m7 © |
Timcom99 Send message Joined: 30 Sep 04 Posts: 105 Credit: 8,927,290 RAC: 0 |
I usually keep a 5 Day Cache which I Crunch in about 3 1/2 Days. I might up this a bit after they get back online. I can Crunch a 7 Day supply in 5 Days or a 10 Day Supply in 7 Days. I might play it safe and be a 10 Day Cache guy. I will still be returning my Work Units a whole Week before they time out. |
KB7RZF Send message Joined: 15 Aug 99 Posts: 9549 Credit: 3,308,926 RAC: 2 |
Changed my preferences to have Seti cache to 6 days, and just joined einstein, and its set at 3 days. Resources set to having seti @ 90.91% and einstein doing 9.09. Does that sound ok? Running on a Intel Celeron 2.6ghz machine w/512mb ram. Any suggestions if that is ok or if I should tweak it would be most welcome. Thanks Jeremy KB7RZF |
Timcom99 Send message Joined: 30 Sep 04 Posts: 105 Credit: 8,927,290 RAC: 0 |
<blockquote>Changed my preferences to have Seti cache to 6 days, and just joined einstein, and its set at 3 days. Resources set to having seti @ 90.91% and einstein doing 9.09. Does that sound ok? Running on a Intel Celeron 2.6ghz machine w/512mb ram. Any suggestions if that is ok or if I should tweak it would be most welcome. Thanks Jeremy KB7RZF</blockquote> Hi KB7RZF: Better drop Einstein back to a Day or less if You are only giving them 9% of your resources. I think with Einstein You need to return your Work in 7 Days. With 3 days Cache and only 9% resources You will never make it. Reset Einstein to 0.5 Days and up their Share to a Full 50% until You have Crunched your 3 Days of Work Units You have Cached. Giving SETI a temporary 50% share will mean You will still be Ok by Returning your 6 Day Cache in 12 Days and SETI gives You a Full 14 Days to do so. |
©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.