Message boards : Number crunching : Size of FAMOUS on HD
Message board moderation
Author | Message |
---|---|
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
What are the disk space requirements for one FAMOUS model? I have two running right now and they\'ve already used 2.5GB! |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Moved to it\'s own thread. I\'m running 4 of them on the beta site, and they\'re taking up 900Megs. I\'ve just uploaded a about 50 trickles (9K each), plus 7 zips, (4.5Megs each), and with these, plus the files from several test versions, the folder was 980Megs. This was measured using Windows Properties on the beta project folder. So: Q1 At which point are you measuring 2.5Gigs? Q2 If at the same point, what else is in the projects folder? |
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
Q1 At which point are you measuring 2.5Gigs? Q3 Have you enabled graphics recording to disk? If you have the model will be creating a 73kb .cpdn file in its tmp directory for every timestep. If that\'s the case you\'ll see the line \"recording...\" above the model time line in the graphics. Recording is enabled and disabled using <CTRL>Q and .cpdn files can be deleted without affecting the model. "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
Send message Joined: 31 Dec 07 Posts: 1152 Credit: 22,363,583 RAC: 5,022 |
Hi, Everyone: Backup size doesn’t seem excessive to me. A backup made today, containing 4 Famous models (1 very early stage and 3 “waiting to runâ€) plus 1 very late stage AM3P WU and the backup took up only about 800 MB’s. This is not excessively large. I am used to 2 late stage AM3Ps plus thier replacements taking up about 1.2 GBs. |
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
[root@jasper boinc]# du -h -s projects/climateprediction.net/ 1.8G projects/climateprediction.net/ And that\'s after one model failed overnight and reported already. So, I only have one model using 1.8G right now. |
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
|
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
Found this in the stdout file. What does it mean? dlopen error: /var/lib/boinc/projects/climateprediction.net/famous_se_6.01_i686-pc-linux-gnu.so: cannot restore segment prot after reloc: Permission denied Other climate models have fun just fine for a long time. |
Send message Joined: 5 Sep 04 Posts: 7629 Credit: 24,240,330 RAC: 0 |
Q2 Not sure what to look for? In this case, models that are no longer being crunched, but which have a folder in the climateprediction.net folder. Which would be the 2 listed in your 2nd last post: The 1st one failed, and the 2nd one was aborted. The stderr out list for both of them contain messages about disk problems. A) HD close to failure? B) Another running program using the HD at the same time as the climate model wanted it? In any case, if those 2 that you listed are no longer in the Tasks tab, then delete the folders. As for your latest post, this may also be due to HD contention. Backups: Here |
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
If you look at the 1st error, it\'s something unrelated to the disk. dlopen error: /var/lib/boinc/projects/climateprediction.net/famous_se_6.01_i686-pc-linux-gnu.so: cannot restore segment prot after reloc: Permission denied The disk problems were because it was out of space because each models generates gigabytes of information, far more than a typical climate model. Is this by design? And if so, how much space is needed per model? Is not, then all this debugging files it generates needs to be quelled. I reset the project and will try one more FAMOUS model, but if there\'s a way to run these models without the gigabytes of files generated, let me know. |
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
Found this in the stdout file. What does it mean? From this search it looks like it\'s related to the way shared libraries which require text relocation are handled in Security-Enhanced Linux when enforcing is enabled. There are a couple of options:
"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
===snip=== Thanks for your help Thyme. I implemented a policy so the whole climateprediction.net folder is in that SEL context. Also ran the chcon so took effect immediately. Will that stop the runaway created-files? The new task that I\'m testing is here: 11386757 |
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
Thanks for your help Thyme. I implemented a policy so the whole climateprediction.net folder is in that SEL context. Also ran the chcon so took effect immediately. Will that stop the runaway created-files? I hope so :) Nobody reported the problem in CPDN Beta, so I\'m guessing it\'s directly related to the SELinux security policy. Assuming your changes fix the problem it would help others if you posted the details to this sticky. "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
Send message Joined: 27 Jan 07 Posts: 300 Credit: 3,288,263 RAC: 26,370 |
|
Send message Joined: 5 Aug 04 Posts: 1283 Credit: 15,824,334 RAC: 0 |
Thanks! "The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer |
©2024 cpdn.org