Message boards :
Number crunching :
New bizarre Work Units
Message board moderation
Author | Message |
---|---|
![]() Send message Joined: 3 Oct 06 Posts: 101 Credit: 8,972,814 RAC: 2 |
Dear project Team! You started new batch of WUs with: 1. minimum quorum - 3; 2. initial replication - 3; 3. max # of error/total/success tasks - also 3, 3, 3!!! or ZERO possible attempts, if at least 1 task in WU will end with computation error, download error, user abort etc. So, just take a look to first results of this new "strategy"... I observed my nearest host only: - from 32 tasks 6 got status "Cancelled by server" or "Completed, can't validate"; - 20% WUs are lost; - approx. 170 hours of CPU time is wasted (in 6 lost WUs)... Is this simple mistake by changing server's configuration (still hope for this)? Or is this Your new strategy? If so, I do not understand You... ![]() |
Send message Joined: 29 May 11 Posts: 10 Credit: 2,810,862 RAC: 0 ![]() ![]() |
|
Send message Joined: 12 Jul 11 Posts: 857 Credit: 1,619,050 RAC: 0 |
Thanksi, I was about to reply, and I have just posted a Mea Culpa to "Host messing up tons of results". Eric. |
Send message Joined: 12 Jul 11 Posts: 857 Credit: 1,619,050 RAC: 0 |
I will try and get help to stop the project, cancel WUs, and restart. I can't do this myself, and it is a holiday weekend. We shall see. Eric. |
![]() Send message Joined: 3 Oct 06 Posts: 101 Credit: 8,972,814 RAC: 2 |
please see http://lhcathomeclassic.cern.ch/sixtrack/forum_thread.php?id=3921&postid=27279 Thank You, I know this problem. And I have NOTHING against raised quorum - the project knows here better than all of us, which level of confidence is mandatory at the moment. I just reported an another problem. Eric, Thank You for fast reply. Happy Holiday! ![]() |
![]() Send message Joined: 3 Oct 06 Posts: 101 Credit: 8,972,814 RAC: 2 |
Dublis... :-) |
©2023 CERN