Message boards : Number crunching : Project encountered internal error: shared memory
Previous · 1 · 2
Author | Message |
---|---|
Keith T. Send message Joined: 1 Mar 07 Posts: 58 Credit: 34,135 RAC: 0 |
If you still have tasks onboard and are not already using the max "Target CPU run time", then you can change this in https://boinc.bakerlab.org/rosetta/prefs.php?subset=project to a higher value. Max is 24 hours, the default is 3 hours. BTW there are 3 threads with identical titles about this problem! If you have run out of Rosetta work there are plenty of other BOINC projects out there. |
dutchie Send message Joined: 14 Oct 07 Posts: 2 Credit: 445,340 RAC: 0 |
HELLO EVERYBODY. UNFURTIANLY,I HAD AN ATTACK OF TROJAN HORSSES.AFTER THAT I REMOVED THEM,I HAVE THE NEXT PROBLEM,THE SERVER SAYS: PROJECT ENCOUNTERED INTERNAL ERROR:SHARED MEMORY. AND THE PROJECT STOPPED AND GETTIN NO NEW WORK. DOES SOMEBODY KNOW THIS PROBLEM,OR KNOW WHAT TO DO. I HOPE SO -RENE--HOLLAND |
Path7 Send message Joined: 25 Aug 07 Posts: 128 Credit: 61,751 RAC: 0 |
Technical news Hi all, At first best wishes to DK, hopefully he will recover soon. Rhiju, thank you a lot for fixing the servers. I'm glad the servers are up & running. Happy crunching, Path7. |
Melissa Chung Send message Joined: 30 Nov 07 Posts: 1 Credit: 4,973 RAC: 0 |
If you still have tasks onboard and are not already using the max "Target CPU run time", then you can change this in https://boinc.bakerlab.org/rosetta/prefs.php?subset=project to a higher value. Max is 24 hours, the default is 3 hours. I manage to set up my general settings to 8 days worth of work but I am not clear in what is "Target CPU run time" and Task onboard? Would you please explain? Thank you for your time. |
Mod.Sense Volunteer moderator Send message Joined: 22 Aug 06 Posts: 4018 Credit: 0 RAC: 0 |
Melissa, at the time that was posted, the project was unable to give out new work, so "on-board" meaning that your computer already has them in the list waiting to run. The target CPU runtime is a Rosetta preference that you can set to meet your needs. By default tasks ship with a 3hr preference. This means that Rosetta tries to end the task once it has run for 3 hours of CPU time. It does this by calculating more and more models until the desired runtime it reached. Some models take more then an hour to complete, so if your perference is only to run for an hour, you will often see it taking longer then that. This is because you have to complete a model in order to have anything useful to report back. If you already have 8 days of work, and BOINC thinks your machine runs 24hrs a day, then you don't want to be adjusting your runtime preference. You see the preference takes effect when you update to the project. So your existing 8 days worth of 3hr work units would immediately become about 20 days worth of 10 hour work units if you changed your preference to 10hrs for example. And 20 days is too long, the deadlines are 10 days. If you want to adjust your runtime perference, you should first reduce the amount of work you have already loaded ("onboard") on your machine. It will take BOINC some time to properly understand how long the tasks run, and to show an accurate estimate on the runtime. And it cannot order new work accurately until it's gotten the picture that the tasks will run at your new runtime preference. Basically, set the runtime preference once, and forget it. Then later inxcrease your number of days of work to keep on hand. And adjust it gradually over a week or so. Rosetta Moderator: Mod.Sense |
Luuklag Send message Joined: 13 Sep 07 Posts: 262 Credit: 4,171 RAC: 0 |
omg how anoying is this, i wasn't able to read for a few days, but always keep a baclog of 3/4 days of rosetta work, since my pc isn't crunching that much';) but now im reading and i have to read like 20 times the same message, and 20 times the same question what to do with it. but then if you go a little higher you can read 1 or 2 posters saying its a server issue, and still people asking what to do. then i would say read things above and find out you just have to wait till the server is back online. |
Message boards :
Number crunching :
Project encountered internal error: shared memory
©2025 University of Washington
https://www.bakerlab.org