Advanced search

Message boards : Number crunching : Asymptotic progress on a task

Author Message
HillFamily
Send message
Joined: 2 Mar 09
Posts: 6
Credit: 584,337,082
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 9963 - Posted: 19 May 2009 | 11:14:05 UTC

My computer's been trying to finish a task called '428-GIANNI_FB-0-100-RND4078_1 using acemd version 664' for 4-5 days now. Normal tasks seem to take about a day. It makes good progress for a while, then seems to get stuck and none of the figures move, even though the task shows as running.

Exiting/relaunching the BOINC manager gets it back unstuck for a while. Finally over the course of 10-12 exits/relaunches throughout the weekend, it finally finished.

I don't normally pay that close attention to a task's progress, so I don't know if this a Normal Thing or not, nor do I know whether it would have gotten unstuck by itself.

When I exit/relaunch the % complete will drop slightly. For instance, it may read 99.321%, then when I relaunch, it'll read that upon load, but within a few seconds it'll drop to 99.210%.

Is this something I should keep an eye on, or is there something strange going on with that type of task?

Thanks,
Eugene

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 9974 - Posted: 19 May 2009 | 19:54:56 UTC - in response to Message 9963.

Unfortunately you witnessed a bug recently introduced. I don't think we have properly organized info on it yet, except that it has been reported for 6.4.7 and 6.6.x .. and may not depend on the BOINC version at all.

Is it this task? It seems like it ran for ~3 days (if you run 24/7) but registers 50k seconds, just shy of 14h, which is your normal time.

However, it's normal that if you restart BOINC you loose some progress - the client is resuming from the last checkpoint.

MrS
____________
Scanning for our furry friends since Jan 2002

HillFamily
Send message
Joined: 2 Mar 09
Posts: 6
Credit: 584,337,082
RAC: 0
Level
Lys
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 9996 - Posted: 20 May 2009 | 11:32:04 UTC - in response to Message 9974.

Yes that's the job that eventually finished. I probably was running about 18 hrs/day.
Thanks for your response.
--Eugene

ExtraTerrestrial Apes
Volunteer moderator
Volunteer tester
Avatar
Send message
Joined: 17 Aug 08
Posts: 2705
Credit: 1,311,122,549
RAC: 0
Level
Met
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 10027 - Posted: 21 May 2009 | 11:14:32 UTC

Mh, so we have a reported runtime of 14h versus an estimated runtime of 3*18h = 54h. That's 4 times longer, just as in the case of "the classical 6.6.20 bug"..

MrS
____________
Scanning for our furry friends since Jan 2002

Post to thread

Message boards : Number crunching : Asymptotic progress on a task

//