Message boards :
Number crunching :
Why would CPU tasks have Way different times?
Message board moderation
Author | Message |
---|---|
hiamps Send message Joined: 23 May 99 Posts: 4292 Credit: 72,971,319 RAC: 0 |
I clicked to show all tasks and noticed about 1/2 my CPU tasks have a completion time of 50:36 and the other half are at 63:39:39. They are all setiathome_enhanced 6.03 . Seems a bit odd to me, wish they would do away with it, or tone it WAY down... Official Abuser of Boinc Buttons... And no good credit hound! |
Hellsheep Send message Joined: 12 Sep 08 Posts: 428 Credit: 784,780 RAC: 0 |
I clicked to show all tasks and noticed about 1/2 my CPU tasks have a completion time of 50:36 and the other half are at 63:39:39. They are all setiathome_enhanced 6.03 . Seems a bit odd to me, wish they would do away with it, or tone it WAY down... I'm seeing this too. The times aren't that high for me though. They're more like 50 minutes to 7 hours. Although they still seem to complete all in about 2-3 hours regardless of their estimated. I was thinking maybe it could be related to the way the estimation works now, i remember seeing Richard say it may take up to 300+ tasks before the estimation becomes somewhat stable. - Jarryd |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
I had ests. from the same "tape" as 5:xx:xx and 145:xx:xx for the same computer. Because of the latter, I couldn't get any WUs for days before the current "3"-day outage. I think the system is really bollixed up, and the crunching is going to be screwed up for a very long time - weeks, maybe. Unless, of course, the Monster Ego who made the changes to the server side relents, and goes back to the older server-side code. |
Hellsheep Send message Joined: 12 Sep 08 Posts: 428 Credit: 784,780 RAC: 0 |
I had ests. from the same "tape" as 5:xx:xx and 145:xx:xx for the same computer. Because of the latter, I couldn't get any WUs for days before the current "3"-day outage. Of course going back to the old code would fix a few things, although i disagree that it's the best option. Lets hope with this new outage schedule we can get all the bugs worked out and the servers running in a decent manner. :) - Jarryd |
Sutaru Tsureku Send message Joined: 6 Apr 07 Posts: 7105 Credit: 147,663,825 RAC: 5 |
I have still the flops entries in my app_info.xml file and all estimates are fine. AFAIK, current the server flops are disabled again. If they will be enabled again, I need to look if still buggy (not interact well with ReSchedule tool) and if then I need to adjust the values in my app_info.xml file. |
kittyman Send message Joined: 9 Jul 00 Posts: 51468 Credit: 1,018,363,574 RAC: 1,004 |
It's anybody's guess what the Boinc code is going to be like when the servers come back up.... We got no notice when the bomb was dropped on us the last time, and I suspect we will have no notice other than by empirical evidence of our own discovery when 'corrections' have been made. "Freedom is just Chaos, with better lighting." Alan Dean Foster |
Cruncher-American Send message Joined: 25 Mar 02 Posts: 1513 Credit: 370,893,186 RAC: 340 |
It's anybody's guess what the Boinc code is going to be like when the servers come back up.... Amen, brother!!!!! |
Josef W. Segur Send message Joined: 30 Oct 99 Posts: 4504 Credit: 1,414,761 RAC: 0 |
I clicked to show all tasks and noticed about 1/2 my CPU tasks have a completion time of 50:36 and the other half are at 63:39:39. They are all setiathome_enhanced 6.03 . Seems a bit odd to me, wish they would do away with it, or tone it WAY down... If that's host 5185924, the "in progress" tasks listed show sent dates from before, during and after the period estimates were being adjusted server-side. I expect it has actually done much of that work during the outage, but having a 1:75 ratio of estimates implies there's still a mix. The original estimates from the splitter can vary over a 1:5 range depending on AR. The transition to server-side adjustment should have caused another 1:5, yielding 1:25. But since the averaging for anonymous platform was "messed up" there's another factor of 3 in there someplace, and that could be even worse on some hosts. Joe |
©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.