Message boards :
Number crunching :
Oh, come on, is this normal?
Message board moderation
Author | Message |
---|---|
scsimodo Send message Joined: 15 May 03 Posts: 39 Credit: 51,577 RAC: 0 |
I've been an entuthiastic S@H cruncher for more than 1 year. Now I'm happily crunching P@H and LHC. Just for n ostalgic reasons I switched back to S@H for a few WUs. Well, what I see is the following: - Outages everywhere (since 2 days) - Pending credits are far behind - performance problems Is this normal? I've never seen this on other projects, ok some time with no work, yes, but not the problems I see on Seti. Did I just pick a bad time, or is this "business as usual"? |
[B^S] Paul@home Send message Joined: 20 Dec 99 Posts: 121 Credit: 1,885,420 RAC: 0 |
This is a particularly bad time.. but there is also an unfortunate familiarity about it!! There now appears to be light at the end of the tunnel (see the recent tech news + the odd post in this forum) so hopefully you'll get work again soon! cheers, Paul. Wanna visit BOINC Synergy? Click my stats! Join BOINC Synergy Team |
Toby Send message Joined: 26 Oct 00 Posts: 1005 Credit: 6,366,949 RAC: 0 |
Take a look around the message board and the technical news. Some unique problems have cropped up but the current outage should fix them if all goes according to plan. A member of The Knights Who Say NI! For rankings, history graphs and more, check out: My BOINC stats site |
scsimodo Send message Joined: 15 May 03 Posts: 39 Credit: 51,577 RAC: 0 |
This is a particularly bad time.. but there is also an unfortunate familiarity about it!! Well, the light at the end of the tunnel, could be the train heading towards you :-)) |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
Normal? No. Then again, LHC is out of work at this moment and Predictor was down tonight (they are only just back up). If some hardware wants to break down, it will. What you shouldn't do is compare relative small projects with Seti@Home. While LHC and PAH may have an increasing userbase, all they do is run that userbase, whereas Seti@Home runs two, Classic and Boinc. Yet if you want to read around, there's threads enough on this subject. It may cost you all week to get through them. Just don't be biased based upon another (smaller) project's up time. And don't forget who made Boinc. ;) If it weren't for Seti in the first place, you'd be knitting a wool sweater around now... :P |
ampoliros Send message Joined: 24 Sep 99 Posts: 152 Credit: 3,542,579 RAC: 5 |
You caught the tail end (hopefully) of several problems and outages that have culminated in an extended outage. Some things were out of the SetiBOINC staffs hands (outage to replace a faulty breaker or router or RAID disk), others were problems with BOINC (antiques/orphans) that only surfaced because the SetiBOINC project is much larger and has run much longer than the other BOINC projects. If your talking about the 'wating for validation' queue, there are some good discussions about it in this forum. For some perspective SetiClassic is 50 million behind, but that's not the point. Much like any program, there are problems found which result in patches, service packs, upgrades, etc. Unfortunautely, in order to do this maintenance, the SetiBOINC staff must shut down the scheduler. Prior to this, the validation queue may have been behind but it was not affecting anyones ability to do work. The only reason they are fixing this now in one big swoop is because the disks were filling up and it's easier to shut down the project and empty the disks then add storage and reconfigure the services. The problems that lead to this outage are being fixed as well, so we are all hopeful that once this outage is over, everything will return to 'normal'. Half of the BOINC projects are down anyway (for various reasons). BOINC is designed so that it's no big deal if some of them aren't working. [edit]bad grammer[/edit] 7,049 S@H Classic Credits |
scsimodo Send message Joined: 15 May 03 Posts: 39 Credit: 51,577 RAC: 0 |
Normal? No. LHC is out of work, but they announced it. P@H has obviously some problems with their scheduler (unfortunately on weekends). I have to live with it. Now to SETI! Some hardware breakdowns, true! But as non active member of SETi I recall at least 3 RAID crashes, database corruption, ancient results that are not deleted etc. No show stopper, but annoying though. I don't want to blame SETI, I'm just curious (and a liitle bothered) about the current state of SETI. |
[B^S] Paul@home Send message Joined: 20 Dec 99 Posts: 121 Credit: 1,885,420 RAC: 0 |
Backend services appear to be coming back up... Wanna visit BOINC Synergy? Click my stats! Join BOINC Synergy Team |
Jord Send message Joined: 9 Jun 99 Posts: 15184 Credit: 4,362,181 RAC: 3 |
Aren't we all, scsimodo. ;) At least I remember you from past times, and you're not even coming in to just flame everyone and their little sister, before gracefully leaving again. :) Addaboy (?) :) |
scsimodo Send message Joined: 15 May 03 Posts: 39 Credit: 51,577 RAC: 0 |
You caught the tail end (hopefully) of several problems and outages that have culminated in an extended outage. Some things were out of the SetiBOINC staffs hands (outage to replace a faulty breaker or router or RAID disk), others were problems with BOINC (antiques/orphans) that only surfaced because the SetiBOINC project is much larger and has run much longer than the other BOINC projects... I don't complain about the problem various BOINC projects have. I'm just curious about the massive outages at SETI. I'm aware that all projects suffer from low work, hardware failures, sporadic outages. That's one reason why BOINC is there. One project has problems, so crunch for the other. Now I now just hit a bad time at SETI. Well, I have to deal with it... |
scsimodo Send message Joined: 15 May 03 Posts: 39 Credit: 51,577 RAC: 0 |
Aren't we all, scsimodo. ;) Show me one, only ONE thread where I tried to flame! Showing my anger, ok, but I never tried to start a flamewar or to discrace someone (ok, except Guido Waldenmeier) |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
I've been an entuthiastic S@H cruncher for more than 1 year. Now I'm happily crunching P@H and LHC. Just for n ostalgic reasons I switched back to S@H for a few WUs. Is this normal? Why, yes it is very normal. In another thread azwoody asked if there has been a two week period that the SETI/BOINC project has gone without a problem . . . no one answered, and I don't remember that happening . . . so it seems that this is, indeed normal. That doesn't mean I'm leaving, or flaming anyone. And I am well aware that the Berkeley folks are running this operation "where no one has gone before", but I am convinced that this project is in a state (an extended state) of development, with new versions of BOINC and SETI@home coming out regularly, new and/or different hardware configurations, and like this instance, the occasional massive glitch that has to be dealt with before it crashes the whole mess. Is this normal? Until the project achieves some long term (six-months at least) stability, this is, and will be normal. OK. Flame away. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
Server status page says that the services are up. The upload/download servers are still (intentionally) off. |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
The upload/download servers are still (intentionally) off. Then the Status page should say the scheduler is disabled, shouldn't it? |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
Is this normal? Why, yes it is very normal. In another thread azwoody asked if there has been a two week period that the SETI/BOINC project has gone without a problem . . . no one answered, and I don't remember that happening . . . so it seems that this is, indeed normal. Does anyone have logs that would prove, conclusively, that azwoody's statement is either true or not? If I look at BOINC Stats the graphs certainly tend upward quite nicely, but is the resolution good enough to tell for sure? |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
The upload/download servers are still (intentionally) off. The scheduler is a different machine. |
Qui-Gon Send message Joined: 15 May 99 Posts: 2940 Credit: 19,199,902 RAC: 11 |
The upload/download servers are still (intentionally) off. I'm sure you're right, but when I try to connect the message I get is, "8/25/2005 12:38:50 PM SETI@home No schedulers responded", yet the Status page says the "Scheduler" is running. |
[B^S] Paul@home Send message Joined: 20 Dec 99 Posts: 121 Credit: 1,885,420 RAC: 0 |
It is possible they are blocking comms to the scheduler... but want it running. They may want it running for some reason (is it needed to pickup WUs that fail to get a quorum and need to be sent again... or does the feeder do that?) cheers, Paul. Wanna visit BOINC Synergy? Click my stats! Join BOINC Synergy Team |
Shaktai Send message Joined: 16 Jun 99 Posts: 211 Credit: 259,752 RAC: 0 |
I'm sure you're right, but when I try to connect the message I get is, "8/25/2005 12:38:50 PM SETI@home No schedulers responded", yet the Status page says the "Scheduler" is running. It will take a little while for things to settle down. After an extended outage like we had, there are probably a million boxes all trying at once to upload and download work. Give it three or four hours and just let your client manage the process. Be patient, the spice, er a, work will flow again. Team MacNN - The best Macintosh team ever. |
1mp0£173 Send message Joined: 3 Apr 99 Posts: 8423 Credit: 356,897 RAC: 0 |
The upload/download servers are still (intentionally) off. I haven't read the code to see how the schedulers work, but I've seen Matt comment that there are a limited number of simultaneous CGIs allowed to run on each machine, and remember that every single cruncher out here probably has a reason to want to talk to the mother ship. We'll see a couple of days of "ratty comms" as a result of the downtime. |
©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.