can it really be so slow ?

Message boards : Number crunching : can it really be so slow ?

To post messages, you must log in.

1 · 2 · 3 · Next

AuthorMessage
mahdia

Send message
Joined: 26 Sep 05
Posts: 7
Credit: 316,259
RAC: 0
Message 603 - Posted: 27 Sep 2005, 13:14:19 UTC

Hello.
I added this interesting project (one of three current protein projects, I think) to two of my computers.
One of these two is a P3 1 GHz, the other a P4 3 GHz HT. On the last one all works fine, but the good old P3 seems to stuck on 1% since 10 hours. The displayed speed is ca. 5 MFlOPs, which is realy slow. The same computer reaches over 600 MFlOPs with Einstein@home, so I wonder, why the rosetta application is so slow ?


ID: 603 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile [B^S] Paul@home
Avatar

Send message
Joined: 18 Sep 05
Posts: 34
Credit: 393,096
RAC: 0
Message 606 - Posted: 27 Sep 2005, 13:25:08 UTC

Hi, what version of the science app did you get? you should be running 4.77 for windows. anything other than that is very slow...

Paul
Wanna visit BOINC Synergy team site? Click below!

Join BOINC Synergy Team
ID: 606 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Nightbird

Send message
Joined: 17 Sep 05
Posts: 70
Credit: 32,418
RAC: 0
Message 608 - Posted: 27 Sep 2005, 13:32:09 UTC - in response to Message 603.  

Hello.
I added this interesting project (one of three current protein projects, I think) to two of my computers.
One of these two is a P3 1 GHz, the other a P4 3 GHz HT. On the last one all works fine, but the good old P3 seems to stuck on 1% since 10 hours. The displayed speed is ca. 5 MFlOPs, which is realy slow. The same computer reaches over 600 MFlOPs with Einstein@home, so I wonder, why the rosetta application is so slow ?


Right, too slow specially when i read "but the good old P3 seems to stuck on 1% since 10 hours"
Can you suspend this wu (not abort) and begin a new wu ?



ID: 608 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mahdia

Send message
Joined: 26 Sep 05
Posts: 7
Credit: 316,259
RAC: 0
Message 610 - Posted: 27 Sep 2005, 13:36:32 UTC - in response to Message 606.  

Hi, what version of the science app did you get? you should be running 4.77 for windows. anything other than that is very slow...

Paul

4.77 is used.

ID: 610 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mahdia

Send message
Joined: 26 Sep 05
Posts: 7
Credit: 316,259
RAC: 0
Message 611 - Posted: 27 Sep 2005, 13:37:57 UTC - in response to Message 608.  

[/quote]
Right, too slow specially when i read "but the good old P3 seems to stuck on 1% since 10 hours"
Can you suspend this wu (not abort) and begin a new wu ?
[/quote]
Suspended, let's look, how the next one will do the job.

ID: 611 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile adrianxw
Avatar

Send message
Joined: 18 Sep 05
Posts: 653
Credit: 11,840,739
RAC: 42
Message 612 - Posted: 27 Sep 2005, 13:39:57 UTC

Do not expect the % complete figure to move in anything like a linear fashion. It tends to move in jumps. It often appears "stuck" at certain percentages, 83.3% for example is a regular "sticking point". It is not in fact stuck, it suddenly lurches forward after a stage has been completed.
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.
ID: 612 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mahdia

Send message
Joined: 26 Sep 05
Posts: 7
Credit: 316,259
RAC: 0
Message 613 - Posted: 27 Sep 2005, 13:47:29 UTC - in response to Message 612.  

Do not expect the % complete figure to move in anything like a linear fashion. It tends to move in jumps. It often appears "stuck" at certain percentages, 83.3% for example is a regular "sticking point". It is not in fact stuck, it suddenly lurches forward after a stage has been completed.


I know this well. But stucking on 1% for over 10 hours is not normal. It seems to be a problem with this particular workunit. The next one I started 5 minutes ago reached 8.33 % just this moment. Seems to work well.
I wonder if the workunits should behave so different? Perhaps they do ohter sort of computation, but I don't think so.
Perhaps there was a drunken bit in ;-)
ID: 613 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Nightbird

Send message
Joined: 17 Sep 05
Posts: 70
Credit: 32,418
RAC: 0
Message 614 - Posted: 27 Sep 2005, 13:49:08 UTC - in response to Message 612.  
Last modified: 27 Sep 2005, 13:50:29 UTC

Do not expect the % complete figure to move in anything like a linear fashion. It tends to move in jumps. It often appears "stuck" at certain percentages, 83.3% for example is a regular "sticking point". It is not in fact stuck, it suddenly lurches forward after a stage has been completed.

Time moves like that :
1 % -> 8.33 % -> 16.67 % -> 25 % etc etc -> 91.67 % -> 100 %
You have a first step at 1 % then 12 * 8.33 %
But you don't need 10 hours to do 1 % !

The next one I started 5 minutes ago reached 8.33 % just this moment.

This one seems "normal".


ID: 614 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Shaktai
Avatar

Send message
Joined: 21 Sep 05
Posts: 56
Credit: 575,419
RAC: 0
Message 615 - Posted: 27 Sep 2005, 13:49:21 UTC

If you are running multiple projects on that computer, you may also need to extend the time between project switching from 60 minutes to 120-180 minutes. Otherwise it will never reach the next checkpoint before switches and will always revert back to the last checkpoint.

As an additional, I had one work unit get stuck at at 1%. A simple restart of BOINC fixed it, and the work unit finished fine.


Team MacNN - The best Macintosh team ever.
ID: 615 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mahdia

Send message
Joined: 26 Sep 05
Posts: 7
Credit: 316,259
RAC: 0
Message 616 - Posted: 27 Sep 2005, 13:53:55 UTC - in response to Message 615.  

If you are running multiple projects on that computer, you may also need to extend the time between project switching from 60 minutes to 120-180 minutes. Otherwise it will never reach the next checkpoint before switches and will always revert back to the last checkpoint.

But doesn't that matter when applications are stay in memory while preempted ?


As an additional, I had one work unit get stuck at at 1%. A simple restart of BOINC fixed it, and the work unit finished fine.


This is an option, if you can watch the computer in regular intervals. But all my BOINCing hosts are on 7/24 and BOINC runs in service mode with no one is logged on .
ID: 616 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mrwizer
Avatar

Send message
Joined: 18 Sep 05
Posts: 23
Credit: 507,085
RAC: 0
Message 639 - Posted: 27 Sep 2005, 17:03:52 UTC - in response to Message 615.  



As an additional, I had one work unit get stuck at at 1%. A simple restart of BOINC fixed it, and the work unit finished fine.


I have also had a WU stick at 1%. Normally I get WU done in a bit over 2 hours, but one was stuck on 1% after 3 hours.
ID: 639 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Nuadormrac

Send message
Joined: 27 Sep 05
Posts: 37
Credit: 202,469
RAC: 0
Message 646 - Posted: 27 Sep 2005, 17:56:52 UTC - in response to Message 616.  
Last modified: 27 Sep 2005, 17:57:25 UTC

If you are running multiple projects on that computer, you may also need to extend the time between project switching from 60 minutes to 120-180 minutes. Otherwise it will never reach the next checkpoint before switches and will always revert back to the last checkpoint.

But doesn't that matter when applications are stay in memory while preempted ?


Yes, if the preferences are set to keep the projects in memory, then when it resumes (baring a shutdown, system crash, or the like) it will pick up where it left off. I'm actually set to switch each 50 mins with "leave in memory" and all seems well. I'm assuming like others that you might have had a bad WU or something.

BTW, it might be a good idea to give the WU ID so when a project manager per chance sees this thread, they'll have an idea of what unit was giving problems here.
ID: 646 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
mahdia

Send message
Joined: 26 Sep 05
Posts: 7
Credit: 316,259
RAC: 0
Message 687 - Posted: 28 Sep 2005, 7:34:14 UTC

Hello !

Today morning when I came into my office, another workunit on the fast P4 was stucked at 1% since 5 hours. After a restart of the whole boinc service all works fine and the particular WU is at 16,67 % right now.

It seems, that there is a bug in the rosetta application, which causes under circumstances the application to be stucked. Looks like an invalid ending condition of a loop or so (this is only a assumption).

Regards, mahdia
ID: 687 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Pconfig

Send message
Joined: 26 Sep 05
Posts: 6
Credit: 56,254
RAC: 0
Message 689 - Posted: 28 Sep 2005, 10:43:27 UTC

Yeah i've had my first 'stuck at 1%' wu today too. It was there for 10 hours, after restarting boinc the calculation time starts all over.

I think this is a serious bug that will hang up alot of systems as people can't watch all there computers all the time ...
Proud member of the Dutch Power Cows
ID: 689 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Keith E. Laidig
Volunteer moderator
Project developer
Avatar

Send message
Joined: 1 Jul 05
Posts: 154
Credit: 117,189,961
RAC: 0
Message 700 - Posted: 28 Sep 2005, 15:14:56 UTC - in response to Message 689.  

Yeah i've had my first 'stuck at 1%' wu today too. It was there for 10 hours, after restarting boinc the calculation time starts all over.

I think this is a serious bug that will hang up alot of systems as people can't watch all there computers all the time ...


Thanks for the feedback. I'll make sure DK learns of your problems upon his return. We want to R@H to run smoothly and provide appropriate credit for work done, so we'll track the trouble down.

ID: 700 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Bok

Send message
Joined: 17 Sep 05
Posts: 54
Credit: 3,514,973
RAC: 0
Message 702 - Posted: 28 Sep 2005, 15:28:17 UTC - in response to Message 700.  

I've had a number of these. Only on windows though. I just tend to abort them via Boincview.

I'll watch out and make a note of the id if I get another one.

Bok

Free-DC

Stats for all projects

Custom Stats
ID: 702 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Divide Overflow

Send message
Joined: 17 Sep 05
Posts: 82
Credit: 921,382
RAC: 0
Message 736 - Posted: 29 Sep 2005, 3:48:51 UTC
Last modified: 29 Sep 2005, 3:53:33 UTC

I've just hit my first one of these 1% WU's. I let it run for three hours, almost an hour past my longest previous WU completion time. The estimated time to completion was at 5 hours plus and increasing. I shut down BOINC and restarted my my computer. After restarting BOINC, the same WU resumed from scratch, but this time it seems to be behaving properly. Percentage complete indicators are increasing by 1/12 and the estimated time to completion is decreasing.
ID: 736 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile [B@H] Ray
Avatar

Send message
Joined: 20 Sep 05
Posts: 118
Credit: 100,251
RAC: 0
Message 785 - Posted: 29 Sep 2005, 21:44:07 UTC

Why have I not had a problem yet? My system should be reinstalled, to many people adding programs, updateing programs, removeing programs, gotten so it takes quite a while to start up. BUT no problem with Rosetta@home.

Maby Rosatta needs more systems like this.

I am going to reformat the HD and reinstall Windows, hope that I keep not having problems with Rosetta.
Ray


Pizza@Home Rays Place Rays place Forums
ID: 785 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Padanian

Send message
Joined: 27 Sep 05
Posts: 14
Credit: 15,190
RAC: 0
Message 864 - Posted: 1 Oct 2005, 9:28:52 UTC

I had a WU stuck at 83.66% for 5 hours (2.8GHz P4). It suddenly jumped to 91.66, and then back to 83.66% per 5 or 6 times. I had to kill it.
ID: 864 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Shaktai
Avatar

Send message
Joined: 21 Sep 05
Posts: 56
Credit: 575,419
RAC: 0
Message 872 - Posted: 1 Oct 2005, 15:19:16 UTC - in response to Message 864.  
Last modified: 1 Oct 2005, 15:21:00 UTC

I had a WU stuck at 83.66% for 5 hours (2.8GHz P4). It suddenly jumped to 91.66, and then back to 83.66% per 5 or 6 times. I had to kill it.


Are you running more then one project on that computer, so that it is switching projects. If so, under general preferences try increasing the "time" between project switches to 120 minutes instead of the default 60. From 83.33% on, the calculations are more complex and take longer. Another option is to "leave in memory". I had this same problem, and increasing the time between switches solved it.


Team MacNN - The best Macintosh team ever.
ID: 872 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
1 · 2 · 3 · Next

Message boards : Number crunching : can it really be so slow ?



©2024 University of Washington
https://www.bakerlab.org