Heavy load on Harddisk with "Autodock with mgltools scripts"

Message boards : Number crunching : Heavy load on Harddisk with "Autodock with mgltools scripts"

To post messages, you must log in.

AuthorMessage
Profile [B^S] BOINC-SG
Avatar

Send message
Joined: 22 Apr 09
Posts: 8
Credit: 1,078
RAC: 0
Message 64 - Posted: 29 Apr 2009, 21:04:44 UTC

Every time a WU starts (currently version 1.13) my harddisk is working very hard, in the meantime the CPU is idle for appr. 8-12 seconds...

The WUs arent that long anyway, so lots of time is wasted, imo...

Is the app unzipping something?


Cheers!


My NEW BOINC-Site

Why people joined BOINC Synergy...
ID: 64 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Jack Shultz
Avatar

Send message
Joined: 10 Apr 09
Posts: 503
Credit: 120,150
RAC: 0
Message 66 - Posted: 29 Apr 2009, 22:35:19 UTC - in response to Message 64.  

yup, I caused lots of pain when I tried to make it so it did not have to unzip every time. I was trying to do it once in the project directory. This simple approach won. Current workunits are short I'm going to change that soon, and use a genetic algorithm to optimize the docking results. That should bring it up to 1.5 hours.
ID: 66 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile [B^S] BOINC-SG
Avatar

Send message
Joined: 22 Apr 09
Posts: 8
Credit: 1,078
RAC: 0
Message 88 - Posted: 2 May 2009, 14:59:56 UTC

Update: Still the same with 1.21 - WU took 3 mins, heavy load!


My NEW BOINC-Site

Why people joined BOINC Synergy...
ID: 88 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Rebirther
Avatar

Send message
Joined: 1 May 09
Posts: 2
Credit: 476
RAC: 0
Message 122 - Posted: 6 May 2009, 22:11:28 UTC

Yes, its bad for non checkpointing apps, restarting from last scratch because all running projects getting "no heartbeat from core client"
ID: 122 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Jack Shultz
Avatar

Send message
Joined: 10 Apr 09
Posts: 503
Credit: 120,150
RAC: 0
Message 123 - Posted: 7 May 2009, 2:56:37 UTC - in response to Message 122.  

hmm, I'll see what we can do regarding checkpointing...
ID: 123 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Rebirther
Avatar

Send message
Joined: 1 May 09
Posts: 2
Credit: 476
RAC: 0
Message 245 - Posted: 12 May 2009, 14:48:23 UTC

Also with 1.23, the uncompress process takes too much load on disk (several 100MB). Result of no heartbeat from coreclient still present. (6.6.28 win32)
ID: 245 · Rating: 0 · rate: Rate + / Rate - Report as offensive
Profile Tim Turner
Avatar

Send message
Joined: 1 May 09
Posts: 570
Credit: 184,322
RAC: 0
Message 246 - Posted: 12 May 2009, 15:20:18 UTC - in response to Message 245.  
Last modified: 12 May 2009, 15:21:37 UTC

boinc manager and core are strained with 4 drug's running; backing off 1 cpu so that the other 3 can dedicate to boinc and let 1 core handle operations of pc and manager. i would say, if you have a raptor or 10 k+ hard drive you wouldn't see a huge problem but i only wish i had 1.

i got a new rig in a newegg wish list.
ID: 246 · Rating: 0 · rate: Rate + / Rate - Report as offensive

Message boards : Number crunching : Heavy load on Harddisk with "Autodock with mgltools scripts"


©2017 All rights reserved | Design by Digital BioPharm Ltd