Seti@Home optimized science apps and information
Optimized Seti@Home apps => Discussion Forum => Topic started by: kevin6912 on 24 Jun 2010, 10:12:04 pm
-
Looks like some decisions were made. Next week would be 6/29/2010.
SETI home page quote:
"Starting next week our servers will be turned off for several days a week - initially, Tuesday through Thursday. This will allow us to devote more time to scientific development."
Regards,
Kevin
-
... plus the usual unscheduled outages from Friday to Monday :-\
-
I guess keeping the work buffer maxxed-out should allow us to bridge the outages (scheduled and not). :(
"Starting next week ..."
I think they started a week early! ::)
-
This must be a good time for internal maintenance. For some time I've been unable to change SETI's priority on my PCs. I have a preference set in BOINCstats for the fraction I want SETI to process but that is not propagating to the PC's even via manual sync.
There are no local prefs set; I use only BOINC to set them. On SETI the setting was 25; now it is double hyphen (--) achieved by entering zero. But SETI was stuck at 95 on my pc's.
I have zero WU right now, let my cache empty to see if I could fix this. Can someone offer a hint where to look for the stuck bit?
Thanks
Art
-
...
For some time I've been unable to change SETI's priority on my PCs. I have a preference set in BOINCstats for the fraction I want SETI to process but that is not propagating to the PC's even via manual sync.
...
Can someone offer a hint where to look for the stuck bit?
BOINCstats/BAM! "Project resources" settings will work only for your new computers, attached through BAM!
To change SETI@home fraction setting for your already connected to BAM! computer, you need to enter into "Host list" and set required value there and sync.
-
WoW, I recived 7 messages total today about snv rep update in Berkeley. Looks like they keep their promises and work hard on new code (updates go into "seti_science" part of SVN)
-
"Projects are down due to a database machine crash.
The machine that serves the BOINC database crashed this evening. It will be recovering overnight. 3 Jul 2010 4:17:58 UTC "
Ha ha I knew it would not be long, after a three day planned outage we are now hit by an unplanned outage when we should be restocking our work supply before the next planned 3 day outage.
-
Came back up at midnight, at least that is when I started getting work.
-
I have to say one thing, since everything came back online there were frequent times where the connection would become maxed out and then drop back to normal, now we've been pegged since the WU limit was released and no crashes or anything. I'm quite proud of how well it actually held up this time.
-
Hrm, buffers cranked to
11 10. Tnk should do it... I had hoped to keep them smaller, especially on my work PC, but I'd rather keep work flowing.
-
Well here we go again, after another planned outage we have another unplanned outage. If the guys at SETI worked in the real world they would have been sacked long ago!!!!
-
Well here we go again, after another planned outage we have another unplanned outage. If the guys at SETI worked in the real world they would have been sacked long ago!!!!
If the guys at SETI worked in the real world they'd be getting salaries at least 5 times higher, have a team of technicians working to maintain the hardware, have a stockroom full of new spares under warranty, etc. etc. etc. Under those circumstances there would be no need for the new outage schedule, and internal unplanned outages would be rare.
The cause of the outage extension is external, outside the area they control, on campus or between there and the PAIX. Considering volunteers couldn't be bothered to donate enough to get decent bandwidth down the hill to campus, the thought of an alternative connection path is really out in the "pie in the sky" category. We live in an imperfect world, either the goals of the project are sufficient to outweigh temporary inconveniences or not, a matter of personal opinion. I'm in for the long term, whenever they can deliver work I have equipment to do some of it.
Joe
-
Uploads are going in now. Shouldn't be long before they open up for us to report and then download new work. We are on our way!
-
Well here we go again, after another planned outage we have another unplanned outage. If the guys at SETI worked in the real world they would have been sacked long ago!!!!
Hmm. It actually works better with these regular outages. I have no problems whatsoever to get my computers filled.
-
Well here we go again, after another planned outage we have another unplanned outage. If the guys at SETI worked in the real world they would have been sacked long ago!!!!
Hmm. It actually works better with these regular outages. I have no problems whatsoever to get my computers filled.
Likewise, I'm getting regular work. But through no action on my part, my SETI cache is much smaller than it used to be. Perhaps with fewer available download days SETI has become more watchful about the number of WU sent to each PC?
-
Watching the recovery this time was interesting.
The Cricket graphs basically showed the Scdeuler coming online clearing the "waiting to report" and starting of the uploads. Then when downloads were enabled, the bulk of the traffic could be used for downloads.
So now getting the quota points during the weekend set to allow fillin gof larger caches.
-
Likewise, I'm getting regular work. But through no action on my part, my SETI cache is much smaller than it used to be. Perhaps with fewer available download days SETI has become more watchful about the number of WU sent to each PC?
The times are just off by a bit. So the reported times are a bit higher than the actual time. They are in the process of fixing it for some time now.