How come the remaining workunits are going up not down?

Message boards : Number crunching : How come the remaining workunits are going up not down?

To post messages, you must log in.

AuthorMessage
carlosi

Send message
Joined: 4 Oct 11
Posts: 2
Credit: 54,430
RAC: 0
Message 2134 - Posted: 10 Oct 2011, 18:30:48 UTC
Last modified: 10 Oct 2011, 18:31:49 UTC

Hi, How come the remaining workunits are going up not down? A few days ago was at 71711 today is at 71774. the # of workunits: 2880000 has not change at all. Also can someone explain to me why there are 2 in progress field with different numbers? The page is in the Server status page
Thanks

Carlos
ID: 2134 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile TJM
Project administrator
Project developer
Project scientist
Avatar

Send message
Joined: 25 Aug 07
Posts: 843
Credit: 70,472,323
RAC: 388,749
Message 2137 - Posted: 24 Oct 2011, 12:35:43 UTC - in response to Message 2134.  
Last modified: 24 Oct 2011, 12:36:30 UTC

This is due to the way the server handles workunits.
They're marked as "in progress" when the workunit enters BOINC server queue (it's the first "in progress" value).
There are some unlucky workunits which do not come back (timeout, errors, etc) so when they hit their timeout (it's 40 days) they're reinserted back to the "available" pool.
The second "in progress" counts the workunits which are in "sent" state. In theory, the first "in progress" should be equal to the "results ready to send" + second "in progress", however most of the time it's a bit off, because not all of the backend daemons work in realtime.
M4 Project homepage
M4 Project wiki
ID: 2137 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote

Message boards : Number crunching : How come the remaining workunits are going up not down?




Copyright © 2017 TJM