Welcome to MilkyWay@home

The computer has reached a limit on tasks in progess

Message boards : Number crunching : The computer has reached a limit on tasks in progess
Message board moderation

To post messages, you must log in.

AuthorMessage
mrchips
Avatar

Send message
Joined: 31 Oct 10
Posts: 15
Credit: 280,797,444
RAC: 0
Message 60602 - Posted: 15 Dec 2013, 20:54:32 UTC

I Have 2 dual-core machines an only get 6 tasks max on each.
I then get a limit on task error.
What do I need to change to get more than 6 tasks?
The total runtime for these are only 3-4 hours.
ID: 60602 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3315
Credit: 519,949,340
RAC: 22,061
Message 60606 - Posted: 16 Dec 2013, 11:56:12 UTC - in response to Message 60602.  

I Have 2 dual-core machines an only get 6 tasks max on each.
I then get a limit on task error.
What do I need to change to get more than 6 tasks?
The total runtime for these are only 3-4 hours.


You can't, as soon as you finish some of those units you will get more, it is NOT a limit per day, it is a limit for running at one time. It helps prevent those from really fast pc's and gpu's from grabbing a bazillion units and leaving none for the rest of us. The make work Server can keep up as long as the need is steady, but for someone to come in and grab a few thousand units would be detrimental to us little guys.
ID: 60606 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Link
Avatar

Send message
Joined: 19 Jul 10
Posts: 578
Credit: 18,845,239
RAC: 856
Message 60634 - Posted: 20 Dec 2013, 13:31:46 UTC - in response to Message 60606.  

It helps prevent those from really fast pc's and gpu's from grabbing a bazillion units and leaving none for the rest of us. The make work Server can keep up as long as the need is steady, but for someone to come in and grab a few thousand units would be detrimental to us little guys.

That's not the real reason IIRC.

New work units are made based on the results returned, so if someone would have let's say a 7 days cache, he would be returning results for workunits, which are sort of out of date, because since those workunits were made, millions of new results have been returned and the model of the galaxy has been improved based on them.

That's also why there are just about 500 results ready to send for each application. It would not be a problem to have there several times more, but that's how the project tries to send out workunits based on as current version of the model as possible.

At least that's how I remember the official explanation for the low limits, which btw. result in even shorter work buffer for the GPUs. Those 3 WUs per CPUs last actually quite long compared to that.
ID: 60634 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile mikey
Avatar

Send message
Joined: 8 May 09
Posts: 3315
Credit: 519,949,340
RAC: 22,061
Message 60639 - Posted: 21 Dec 2013, 12:07:35 UTC - in response to Message 60634.  

It helps prevent those from really fast pc's and gpu's from grabbing a bazillion units and leaving none for the rest of us. The make work Server can keep up as long as the need is steady, but for someone to come in and grab a few thousand units would be detrimental to us little guys.

That's not the real reason IIRC.

New work units are made based on the results returned, so if someone would have let's say a 7 days cache, he would be returning results for workunits, which are sort of out of date, because since those workunits were made, millions of new results have been returned and the model of the galaxy has been improved based on them.

That's also why there are just about 500 results ready to send for each application. It would not be a problem to have there several times more, but that's how the project tries to send out workunits based on as current version of the model as possible.

At least that's how I remember the official explanation for the low limits, which btw. result in even shorter work buffer for the GPUs. Those 3 WUs per CPUs last actually quite long compared to that.


Darn it, it is only 7AM and I have already learned my one new thing for today!!! THANK YOU Link for setting me straight.
ID: 60639 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile dskagcommunity
Avatar

Send message
Joined: 26 Feb 11
Posts: 170
Credit: 205,557,553
RAC: 0
Message 60657 - Posted: 26 Dec 2013, 19:09:45 UTC - in response to Message 60602.  
Last modified: 26 Dec 2013, 19:13:32 UTC

I Have 2 dual-core machines an only get 6 tasks max on each.
I then get a limit on task error.
What do I need to change to get more than 6 tasks?
The total runtime for these are only 3-4 hours.


See it that way, thats way better than my 30-45 Minutes Cache on GPU :P The daily 24h Disconnect on the 3G Connection for this machine is the greatest enemy for that, because the router sometimes does not recognize that in the first minutes to reconnect ;)

But the last explanination was good there so i know now too why it is so low :)
DSKAG Austria Research Team: http://www.research.dskag.at



ID: 60657 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote

Message boards : Number crunching : The computer has reached a limit on tasks in progess

©2024 Astroinformatics Group