| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: Clean Energy Project phase 2 Mac and Windows Released |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 92
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Do these read/write numbers look right (right two columns, in bytes)? My HD really gets hammered at the start, especially when multiples start at the same time.
 Including the startup, the 3 running tasks have read/written about 11 GB in the last half hour or so. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
That's nomal, CEP2 has high disk usage.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
OK, I finally have my new drive setup complete. Physically, I have 3 1.5 TB drives. Using Intel's Matrix Storage, I have a 1.5 TB RAID-5 volume for data and a 2 TB RAID-0 volume for a scratch disk, to which I've moved the BOINC data directory. Write-back cache is enabled on both drives. I could not believe how fast a new CEP2 block initialized on that drive. It probably took 3 seconds (maybe less) until the % meter started climbing. With this new setup, I am back to 6 simultaneous units on my Core i7 with HT. Disk access issues should be a thing of the past.
If anyone is wondering how to move the data dir, just kill the BOINC process, create the new data dir, set permissions, move everything over, then do a search in the registry for C:\ProgramData\BOINC and replace that with your new data dir. Start BOINC and it should pick right back up in the new directory. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi,
I am losing computation time when the boinc client is suspended or the PC is rebooted. Today, 5 hours computation time disappeared. Looking at the very long computation time needed for clean energy the task will be stumpled back many times... |
||
|
|
KWSN - A Shrubbery
Master Cruncher Joined: Jan 8, 2006 Post Count: 1585 Status: Offline |
This project has some rather long checkpoint intervals. Setting your profile to "leave applications in memory when suspended" will keep you from losing time when the client is suspended. A reboot will take it back to the previous checkpoint.
----------------------------------------The only other recourse is to check the progress of the work unit before suspending or rebooting and wait until it has recently finished a checkpoint. Good luck.  Distributed computing volunteer since September 27, 2000 |
||
|
|
l_mckeon
Senior Cruncher Joined: Oct 20, 2007 Post Count: 439 Status: Offline Project Badges: 
|
This project has some rather long checkpoint intervals. Setting your profile to "leave applications in memory when suspended" will keep you from losing time when the client is suspended. A reboot will take it back to the previous checkpoint. The only other recourse is to check the progress of the work unit before suspending or rebooting and wait until it has recently finished a checkpoint. Good luck. I don't crunch this project because of its unfriendly habits, but would Hibernate work? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello lochner,
I am losing computation time when the boinc client is suspended or the PC is rebooted. Today, 5 hours computation time disappeared. Looking at the very long computation time needed for clean energy the task will be stumpled back many times... This can be an irritating project. You absolutely need to select 'Leave Application In Memory When Suspended' in your profile. If you need to reboot or power-down daily, I suggest you run other projects than this one. The idea is to be happy about the contribution you are making to science. Avoid anything that frustrates you. Lawrence |
||
|
|
BoincST
Cruncher Joined: Feb 25, 2010 Post Count: 12 Status: Offline Project Badges: 
|
This project has some rather long checkpoint intervals. Setting your profile to "leave applications in memory when suspended" will keep you from losing time when the client is suspended. A reboot will take it back to the previous checkpoint. The only other recourse is to check the progress of the work unit before suspending or rebooting and wait until it has recently finished a checkpoint. Good luck. I don't crunch this project because of its unfriendly habits, but would Hibernate work? Yes, that works. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Dear lochner,
yes, the CEP2 checkpointing intervals are rather long, but LAIM in combination with hybernation/sleep rather than shutdown will bypass this problem. If those are no viable options for you, then you may indeed be happier with crunching another project. We are very sorry about that, but at the end of the day you should contribute in a way which is satisfying for you. Best wishes Your Harvard CEP team |
||
|
|
nittany85
Cruncher United States of America Joined: Apr 29, 2007 Post Count: 17 Status: Offline Project Badges:
|
I have been aware of the long time intervals between checkpoints but my machine is currently working on a job that has not had a checkpoint for almost 11 hours of CPU time. This seems a bit extreme and is much longer than the 4-6 hour intervals that I have seen in the past.
Is this job an exception to the rule or is this a trend that we will be seeing in the future? The workunit name is E203204_104_C.27.C22H15NOSSi2.005875551.1set1d06. |
||
|
|
|