![]() With a 16GB drive you don't have a lot of breathing room. In this work we present our experience from the first year of theSkyNet Pan-STARRS1 Optical Galaxy Survey (POGS) project. I'm not sure how many threads on your 4P, but divide the threads by 3 and that's how many gigabytes of free space you need just to run. Sure sounds like disk space for the running slots is the problem. Any ideas? Or should I start its own thread? ![]() I guess if I can't get it figured out, I will kick it over the pogs, as I don't need to bunker for that. That is the only thing I can think of, is it is running out of space. All 3 installs are exactly the same ( appliance) but the 2P's are on 32GB SSD's while the 4P is on a 16GB. I am not sure what the issue is, as it has done so fine on both of the 2P blades. description: Correspond with other users on the pogs message boards. sajinor I am having an issue "bunkering" on my 4P. projectdir: /var/lib/boinc-client/projects/. I don't think queued work units are very large it's just that they replicate their large database (downloaded once) into each slot when it runs. I had to move a bunch of stuff off the running partition so I could free up enough space for 32 slots, and even then I don't know if 16GB of RAM will be enough for 32 tasks. Not only that, but each "slot" (the working directory for each running task) takes about 340MB of disk space. As a heads-up, Rosetta tasks are taking from 250 - 500MB of RAM, each!
0 Comments
Leave a Reply. |