REC calculation to determine project work fetch, NFS returns 100 times greater than expected PRIO

log in

Advanced search

Message boards : Questions/Problems/Bugs : REC calculation to determine project work fetch, NFS returns 100 times greater than expected PRIO

Author Message
marmot
Send message
Joined: 10 Nov 15
Posts: 3
Credit: 1,098,862
RAC: 4
Message 1650 - Posted: 22 Feb 2016, 18:52:33 UTC

From a post at BOINC main forums that the spam engine won't let me link...

Here is the current situation from <work_fetch_debug>1</work_fetch_debug>:

2/22/2016 11:37:55 AM | MindModeling@Beta | [work_fetch] share 0.324
2/22/2016 11:37:55 AM | NFS@Home | [work_fetch] share 0.001


2/22/2016 11:37:55 AM | MindModeling@Beta | [work_fetch] REC 3509.462 prio -1.989 can request work
2/22/2016 11:37:55 AM | NFS@Home | [work_fetch] REC 1949.634 prio -263.383 can request work


Resource shares assigned:

MindModeling = 240
NFS = 001

There are open cores waiting for work and sitting idle because the client decides MindModeling is not the highest priority unless manually updated.

Mindmodeling rarely gives any work, maybe 20,000 WU once a week.
NFS has been sending a steady stream and this machine has been doing NFS WU's for 3 weeks steadily. In order to get any work on MM I have to manually update every few minutes to get past NFS lock on work.

So NFS's prio is about 2 to 3 orders of magnitude greater (abs()) than it should be. NFS prio should be down around -2.63 to -0.263 as it's already at 45% of maximum potential RAC while MindModeling reports -1.9 prio and is at 3-7% of potential RAC from days on end of no work.

How is NFS returning such extreme prio numbers? Is there an error in the server side expected credit information?

Greg
Project administrator
Send message
Joined: 26 Jun 08
Posts: 582
Credit: 223,912,432
RAC: 15,881
Message 1651 - Posted: 23 Feb 2016, 0:57:01 UTC

Here's a link to the thread for future reference:
https://boinc.berkeley.edu/dev/forum_thread.php?id=10839

I'll follow the responses there.

marmot
Send message
Joined: 10 Nov 15
Posts: 3
Credit: 1,098,862
RAC: 4
Message 1652 - Posted: 23 Feb 2016, 23:55:30 UTC - in response to Message 1651.

Am I going to find an answer to this?

I'm going to have to drop NFS or limit it to a single core 2GB virtual machine if I can't get work fetch from other projects while it runs.

Greg
Project administrator
Send message
Joined: 26 Jun 08
Posts: 582
Credit: 223,912,432
RAC: 15,881
Message 1653 - Posted: 25 Feb 2016, 0:17:39 UTC - in response to Message 1652.

From the discussion at the BOINC forum, it appears that NFS@Home is the lowest priority. Have you tried adding, say, Einstein@Home and see if it is fetched before NFS@Home?

marmot
Send message
Joined: 10 Nov 15
Posts: 3
Credit: 1,098,862
RAC: 4
Message 1656 - Posted: 3 Mar 2016, 5:28:52 UTC - in response to Message 1653.
Last modified: 3 Mar 2016, 5:29:56 UTC

From the discussion at the BOINC forum, it appears that NFS@Home is the lowest priority. Have you tried adding, say, Einstein@Home and see if it is fetched before NFS@Home?


I have 31 projects available (set to fetch no new work) on that machine and was trying to get work for 3 projects but NFS always dominated forcing me to manually fetch Mind Modeling and vLHC work. I posted the work fetch diagnostics.

I moved some NFS work into 2 VM's and have given up on this project for shared use in any client. I'm not a noob to running these WU's and am skilled at getting various projects to send me WU's. It took me some diagnostics to track it down to NFS as something is not behaving as expected.

Message boards : Questions/Problems/Bugs : REC calculation to determine project work fetch, NFS returns 100 times greater than expected PRIO


Home | My Account | Message Boards