Message boards :
Number crunching :
Nbody 1.80 errors
Message board moderation
Author | Message |
---|---|
Send message Joined: 10 Feb 09 Posts: 52 Credit: 16,291,993 RAC: 0 |
Some errors with the new app: <search_application> milkyway_nbody 1.80 Windows x86_64 double OpenMP, Crlibm </search_application> |
Send message Joined: 21 Aug 18 Posts: 59 Credit: 5,350,675 RAC: 0 |
Don't worry about this error. It occurred because the newer application is trying to use an older lua file from an older optimization. These older runs have been taken down, but there is a backlog of older jobs that need to be pushed through. Thankfully, since these older jobs only take about 2 seconds to error out, they should clear up over the next few weeks. -Eric |
Send message Joined: 8 Oct 07 Posts: 52 Credit: 5,959,457 RAC: 6,030 |
Resetting the project will correct this problem. |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
Don't worry about this error. It occurred because the newer application is trying to use an older lua file from an older optimization. These older runs have been taken down, but there is a backlog of older jobs that need to be pushed through. Thankfully, since these older jobs only take about 2 seconds to error out, they should clear up over the next few weeks. I don't understand why you guys didn't just name the new lua file something like lua-1-80 and force it to download with the new tasks, that way the people still running the old tasks are file and the new people are good too as they are using the new lua file. Then when version 1.81 comes out you are back to the standard lua name and just rotate thru or keep the new naming convention going forward. It seems you are causing problems for users needlessly and as you know less than 10% of users ever come to the Forums at all they just quit and move to another project when they have problems. |
Send message Joined: 16 Dec 07 Posts: 37 Credit: 25,864,938 RAC: 5,535 |
I know not too worry about them cause we're not wasting CPU time. But after noticing a few errors yesterday checking the thread and then examining the n-body work queued up on the machine. You do have workunits for 1.76 being run on 1.80. I thought Boinc could support two versions of the same science app anyway so that you could have n-body 1.76 and n-body 1.80. As you migrated the app towards more robust output, but still had earlier work that had been previously generated. I like finishing resends and it seems a waste of that earlier modelled run if you're going to let them error out on the server, Even if you're discarding a trivial fraction of the run. |
©2024 Astroinformatics Group