Welcome to MilkyWay@home

New BOINC Manager = no MWAH WUs

Message boards : Number crunching : New BOINC Manager = no MWAH WUs
Message board moderation

To post messages, you must log in.

Previous · 1 · 2

AuthorMessage
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3315
Credit: 519,950,829
RAC: 21,429
Message 54032 - Posted: 15 Apr 2012, 14:43:50 UTC - in response to Message 54030.  

Thanks Floyd that solved it for me too! :-)


To expand:
This change didn't make sense to me which is why I didn't try it before you suggested it. It seems like there is a limit to how many tasks you can have locally in buffer/queue and I was already there. Even after changing it, the number of tasks hasn't changed. The value was set to 0.10 (days) which I would think is 2.4 hours. At max, I only see about 40 tasks which at an estimated 1:15 would only be 50 minutes of work so at first glance it already seemed set higher than what was being queued locally.

Anyway, I changed it to 0.50 and voila, it's working. Thanks again.


You MUST read the words under the settings CAREFULLY in Version 7! The first one says "minimum work buffer" This translates to how LOW your cache can get before it asks for more work! The second setting says "max additional work buffer" this translates to how MUCH work you can have at any one time at the MOST!!! In Ver 7 Boinc will NOT ask for more work until it gets to the minimum setting and will only get as much as is needed to reach the max work setting! What seems to work best is to just reverse your Boinc 6 settings.
ID: 54032 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
floyd

Send message
Joined: 13 Sep 11
Posts: 17
Credit: 3,263,835
RAC: 0
Message 54033 - Posted: 15 Apr 2012, 16:20:45 UTC - in response to Message 54030.  

This change didn't make sense to me which is why I didn't try it before you suggested it. It seems like there is a limit to how many tasks you can have locally in buffer/queue and I was already there.


Now that you describe it in more detail it doesn't make sense to me either. There is a limit of 40 MilkyWay GPU tasks per GPU, but you should have been able to get a new task every time one is returned. Perhaps the whole refetch mechanism isn't activated if you don't get the buffer filled to minimum in the first place? That would be "unexpected behaviour" at least. But glad it works now.
ID: 54033 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
greg_be

Send message
Joined: 18 Aug 09
Posts: 122
Credit: 20,694,647
RAC: 4,385
Message 54042 - Posted: 15 Apr 2012, 23:09:57 UTC - in response to Message 54032.  

Thanks Floyd that solved it for me too! :-)


To expand:
This change didn't make sense to me which is why I didn't try it before you suggested it. It seems like there is a limit to how many tasks you can have locally in buffer/queue and I was already there. Even after changing it, the number of tasks hasn't changed. The value was set to 0.10 (days) which I would think is 2.4 hours. At max, I only see about 40 tasks which at an estimated 1:15 would only be 50 minutes of work so at first glance it already seemed set higher than what was being queued locally.

Anyway, I changed it to 0.50 and voila, it's working. Thanks again.


You MUST read the words under the settings CAREFULLY in Version 7! The first one says "minimum work buffer" This translates to how LOW your cache can get before it asks for more work! The second setting says "max additional work buffer" this translates to how MUCH work you can have at any one time at the MOST!!! In Ver 7 Boinc will NOT ask for more work until it gets to the minimum setting and will only get as much as is needed to reach the max work setting! What seems to work best is to just reverse your Boinc 6 settings.


So what equals the best communication value to the server?
Something like what Mike said?
I just did something like a 2 day and 3 day on the settings and got a huge amount of tasks that flooded my screen. I set things back to 0 and 0 but think it should be something like 5 and 0? What gives me of between 3 and 5 days of total work? But also communicates to the server a few times a day?

ID: 54042 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Mike Simicsak

Send message
Joined: 14 Jan 11
Posts: 5
Credit: 51,204,877
RAC: 0
Message 54044 - Posted: 16 Apr 2012, 4:58:26 UTC - in response to Message 54042.  

Assumption: You are running more projects than MW@H? I am too. I run 4. SETI in particular will flood you with a ton of tasks. As has already been stated MW@H is limited to 40 tasks per GPU. I am assuming it's another project that flooded you with WU's.

I was running 4 projects. I had two projects set to "No New Tasks". I was wondering if there was a chance that the "No New Tasks" was bleeding over from Collatz/PrimeGrid onto MW@H which is now my main project. I removed one and with the improved scheduling in ver7 I am trying to accept tasks from all Projects again. For me it seemed like the "minimum work buffer" solved it although I was kind of doing both things at the same time. I'm still curious if the "No New Tasks" code is implicated?

Do you have "No New Tasks" set for any other project? Can you try (temporarily) removing all other projects and then adjust the "minimum work buffer" setting to see if when MW@H is by itself that resolves it?

If you haven't already another thing I did was remove MW@H, restart BOINC, and then add it back in just to make sure there were no ver 6.x to 7.x upgrade issues. It was after this that I found and adjusted the minimum buffer setting.
ID: 54044 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3315
Credit: 519,950,829
RAC: 21,429
Message 54047 - Posted: 16 Apr 2012, 11:16:10 UTC - in response to Message 54042.  

Thanks Floyd that solved it for me too! :-)


To expand:
This change didn't make sense to me which is why I didn't try it before you suggested it. It seems like there is a limit to how many tasks you can have locally in buffer/queue and I was already there. Even after changing it, the number of tasks hasn't changed. The value was set to 0.10 (days) which I would think is 2.4 hours. At max, I only see about 40 tasks which at an estimated 1:15 would only be 50 minutes of work so at first glance it already seemed set higher than what was being queued locally.

Anyway, I changed it to 0.50 and voila, it's working. Thanks again.


You MUST read the words under the settings CAREFULLY in Version 7! The first one says "minimum work buffer" This translates to how LOW your cache can get before it asks for more work! The second setting says "max additional work buffer" this translates to how MUCH work you can have at any one time at the MOST!!! In Ver 7 Boinc will NOT ask for more work until it gets to the minimum setting and will only get as much as is needed to reach the max work setting! What seems to work best is to just reverse your Boinc 6 settings.


So what equals the best communication value to the server?
Something like what Mike said?
I just did something like a 2 day and 3 day on the settings and got a huge amount of tasks that flooded my screen. I set things back to 0 and 0 but think it should be something like 5 and 0? What gives me of between 3 and 5 days of total work? But also communicates to the server a few times a day?


Why do you want it to contact "the server a few times a day" if you have plenty of work? Remember a server only has so many ports available to it, and every time we connect to it it uses a port, then closes it and opens it for the next person. But it is not like you and I each have our own internet wire connected to the server, we MUST share with others. I have 1.25 in the first box and 0.10 in the second box, this gives me just over one days worth of work and that works fine for me. Now I DO babysit them EVERYDAY, meaning I manually log onto each and every one and send back any completed units. On two pc's I have auto send done thru a cc_config.xml file though, so I just check them to make sure they are still running. The file setting returns a unit as soon as it is completed, this is helpful on gpugrid, for example, where you get more credits for returning a unit within certain time frames.
ID: 54047 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Nick

Send message
Joined: 1 Nov 09
Posts: 21
Credit: 18,543,759
RAC: 0
Message 54061 - Posted: 18 Apr 2012, 22:34:04 UTC - in response to Message 54047.  

Can you tell me where I can get a sample cc_config.xml file?

ID: 54061 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Nick

Send message
Joined: 1 Nov 09
Posts: 21
Credit: 18,543,759
RAC: 0
Message 54062 - Posted: 18 Apr 2012, 22:38:04 UTC - in response to Message 53994.  

My machine was happily crunching away steady numbers of WUs until I over-installed BOINC 7.0.25.

Same thing here. No WUs since installing 7.0.25
ID: 54062 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3315
Credit: 519,950,829
RAC: 21,429
Message 54065 - Posted: 19 Apr 2012, 11:03:22 UTC - in response to Message 54061.  

Can you tell me where I can get a sample cc_config.xml file?


Here is a simple one that is desinged to use all of the gpu's in a system, for those that have more than one gpu in a single pc:

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

ONLY open the file in NOTEPAD in Windows, do NOT use a word processing program as it adds stuff that makes the file useless!!! Save the file as cc_config.xml but as TYPE of txt file! This web page
http://boinc.berkeley.edu/wiki/Client_configuration

will give you all the available options. BE CAREFUL on what you use and what you don't. In most you will see a zero/1 in the middle, zero means OFF and 1 ON or do it. The option above says YES use all gpus found. If I were to add:
<report_results_immediately>1</report_results_immediately> then if would say YES return all results immediately. You must stop Boinc and then restart it after you make any changes to enable the changes.
ID: 54065 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Nick

Send message
Joined: 1 Nov 09
Posts: 21
Credit: 18,543,759
RAC: 0
Message 54071 - Posted: 19 Apr 2012, 16:11:44 UTC - in response to Message 54065.  

Thanks for your help.
ID: 54071 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Matt Arsenault
Volunteer moderator
Project developer
Project tester
Project scientist

Send message
Joined: 8 May 10
Posts: 576
Credit: 15,979,383
RAC: 0
Message 54082 - Posted: 20 Apr 2012, 21:15:34 UTC - in response to Message 53994.  

Should be fixed
ID: 54082 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile cornel
Avatar

Send message
Joined: 28 Feb 09
Posts: 38
Credit: 10,200,014
RAC: 0
Message 54083 - Posted: 20 Apr 2012, 21:32:10 UTC
Last modified: 20 Apr 2012, 21:39:09 UTC

Oh my, thank you.

It seems HD4770 also supports only OpenCL 1.0.

Again, thanks for fixing that scheduler. :)


Oh, and Combat Marmot, could you tell me how you managed those 2784 GFLOPS on your card? I can only reach about 2200 GFlops on mine.
ID: 54083 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Combat Marmot

Send message
Joined: 9 May 10
Posts: 4
Credit: 103,752,795
RAC: 0
Message 54092 - Posted: 21 Apr 2012, 12:15:07 UTC

Matt - happy to report that I received MW ATi WUs today. Thanks for fixing it.

Cornel, I don't think I've done anything special to my card. The clock is at 860 MHz and memory at 975 MHz. Isn't the MFLOPs figure the total between the CPU and GPU?
ID: 54092 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
candido

Send message
Joined: 17 Nov 10
Posts: 4
Credit: 5,057,590
RAC: 0
Message 54114 - Posted: 22 Apr 2012, 23:15:24 UTC

Now I'm getting work too!
thanks
ID: 54114 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile John Black

Send message
Joined: 3 May 10
Posts: 74
Credit: 1,532,760
RAC: 0
Message 54120 - Posted: 23 Apr 2012, 17:22:26 UTC
Last modified: 23 Apr 2012, 17:26:17 UTC

Hi Guys,

at first this new run produced wus that errored out as noted by msattler in another thread started by greg_be. That problem now seems to be solved however all my MW@H wus NBody and PS now seem to be running at high priority.
Guys I have only two CPU cores and I have set my prefs to run MW@H and SETI 50/50. What is now happening is that SETI is being effectively excluded by MW@H which is running wus with a report date 13 days hence in high priority.
I am not confident about "applying patches" or any other such fix so I have been suspending all but one MW@H task to enable SETI to run. This means that I have to set my alarm to go off each time the current MW@H task is due to complete so that I can enable the next task. THIS IS NOT A VIABLE OPTION in the long term.

CAN SOMEBODY PLEASE sort out the glitch in the 7.0.25 BOINC Manager so that normal working can be resumed.

Thanks
John

p.s. I am replying to the above but did not realise that this was in NEWS and perhaps some others in this thread should be in NUMBER CRUNCHING
ID: 54120 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3315
Credit: 519,950,829
RAC: 21,429
Message 54123 - Posted: 24 Apr 2012, 10:09:39 UTC - in response to Message 54120.  

Hi Guys,

at first this new run produced wus that errored out as noted by msattler in another thread started by greg_be. That problem now seems to be solved however all my MW@H wus NBody and PS now seem to be running at high priority.
Guys I have only two CPU cores and I have set my prefs to run MW@H and SETI 50/50. What is now happening is that SETI is being effectively excluded by MW@H which is running wus with a report date 13 days hence in high priority.
I am not confident about "applying patches" or any other such fix so I have been suspending all but one MW@H task to enable SETI to run. This means that I have to set my alarm to go off each time the current MW@H task is due to complete so that I can enable the next task. THIS IS NOT A VIABLE OPTION in the long term.

CAN SOMEBODY PLEASE sort out the glitch in the 7.0.25 BOINC Manager so that normal working can be resumed.

Thanks
John

p.s. I am replying to the above but did not realise that this was in NEWS and perhaps some others in this thread should be in NUMBER CRUNCHING


Reduce your cache size so you store fewer units on your pc and connect to MWAH more often, Seti too for that matter. Boinc thinks you are close to not finishing the work in time for the deadline is the problem, lower the amount of work and Boinc will be okay again. In the extreme you can always abort some of the Seti or MWAH units, they will just get sent out to someone else. They only count against you a little bit, and not in credits, and as long as you don't start making it a habit to abort units you should be okay. If you DO start aborting units regularly you will get fewer units to crunch as the Server will think you have an unreliable pc and reduce the amount of units you can have per day.
ID: 54123 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 · 2

Message boards : Number crunching : New BOINC Manager = no MWAH WUs

©2024 Astroinformatics Group