Discussion in 'World Community Grid (WCG)' started by Chicken Patty, Feb 20, 2009.
Indeed they do
If I recall correctly the default settings are "Run according to preferences" and the default preference is to stop computation if the CPU is above 25% load. Maybe there was another program running, a virus scan maybe?
A virus scan will do it. I've also had things crap out when Nvidia decided to tell me there was a new driver update.
Was just passing by, haha.
Came in to see how things going.
Nice to see all of you still kickin'.
I have a strange problem with BOINC for some time now:
It runs jobs that have a deadline of 10+ days away as "High Priority", but most of the time decides to ignore jobs that have, for example, 20 hours left to meet the deadline. It ignores those and starts running them when they are more than a day overdue.
Because of that, I am starting to get jobs with the status "too late" (which BOINC still decides to complete).
What's the friggin' deal? Why is it prioritizing stuff which doesn't need it, yet decides not to prioritize stuff which definitely needs it?
Which version of BOINC manager are you using? The one that is available on the WCG website has some trouble with this problem as you are describing it. The later development versions (7.0.40 or later) are much better at scheduling job properly. The latest version(s) can be found thru the BOINCStats website
*Note- there are other sites with all versions but I don't know which sites those are atm
Also, you may need to reduce your work buffers in "computing preferences" - this will also help
I am using version 7.0.60
I did experiment with work buffer size (as I noticed it affects this). If reduce my buffer, it goes OK for a short while and then stars working even worse than before. Thus, I am afraid to reduce the buffer any more
(got it down from 3 to 0.75 days during my experimenting)
Gonna update to 7.0.62 and see if that solves it.
Okay! Got all my parts today!
Gonna start assembling a new cruncher as soon as I finish my lunch.
This is the FX-8, right?
Not yet a "great job" . Things were going slow. The new cruncher probably won't be online till friday. But hey, at least I finally got my "main" rig running with a new mobo... sort of.
And yes - the main rig runs with a FX-8320, while the new cruncher is going to use a FX-8120
I'm thinking of bringing my 7870 rig back over to WCg from F@H because it seems to be pretty much worthless over there. Since gpu WUs are winding down, I want to take advantage of them while I can. I'll probably do it tonight.
A future tense great job then
I'm doing a bit of downsizing and I have a system to give away to a lucky WCG cruncher. You'll get an AMD Athlon 64 X2 4400+, HP pull motherboard, 2GB RAM, and an 80GB HDD. I think it's good for about 600 PPD. PM me if you're interested--all I'm charging is shipping. United States only please
AM2 or 939?
Meanwhile, in Lithuania...
The graphics card will go away after I'm done with it. Gonna sit in a case when I scavenge one off from my university's scrap.
AM2. At least I think so--I'm using DDR2 RAM in it
awesome im glad you are donating it to another member so it will continue on crunching.
I don't need this many systems but it's still useful. I may have more goodies soon
I have a bit of a problem - BOINC refuses to run Help Conquer Cancer CPU variant. My manager shows: this. Currently, my CPU is half idle due to lack of load. If this goes on, I am going to be out of CPU jobs in less than half an hour and only have the HCC GPU variant running and CPU staying idle
My work buffer is crammed with unfinished HCC CPU variant jobs, so it refuses to get any more "unstuck" work.
What the hell?
I suspect this was the underlying cause of my previous [unsolved] scheduling problem - the jobs it strangely delayed were HCC CPU jobs.
Temporarily disable the other CPU WUs in your WCG profile to see what happens.
Wihoo, goodies in the way
Got it working again! The cause was a borken app_config.xml file
8350 Vishera feedback
So I changed my platform from Intel QX9770 to AMD 8350 Vishera just recently. I thought from 4 cores to 8 cores would be nice.
However, it seems the 8350 is slower computing WUs than the old (2008) QX9770...
Is it me, or is the "Vishera" kind of slow?
As long as it's taking less than double the time per single WU compared to QX9770, it's working faster, as it does two times more WUs in parallel.
EDIT: My second cruncher is now online
How long have you had the Vishera and are you referring to the finishing of gpu wu's or the cpu in general?
Bulldozer/Piledriver (i.e.Vishera) chips really shine when you load up all of the cores. Give it a lot of work to do and see how it does for you
Most WUs take a little longer.
For example, the Clean Energy Project is taking closer to 13 hours on the 8350 while on the QX9770 it took only up to 10 hours...
On the 8350, the Fight Aids @ Home WUs are all taking 40+ hours - every single one of them.
On the 9770 it only took something like 5 or 6 hours.
This makes me wonder if it there isn't something wrong with my CPU...
Separate names with a comma.