SMP Assignment server

[anko1]

I know this isn't likely to attract much attention over the long weekend, but I thought I'd mention that my dual core SMP client has received 2 3064 WUs in a row.

First one:

Code: Select all

[11:22:37] + No unsent completed units remaining.
[11:22:37] - Preparing to get new work unit...
[11:22:37] + Attempting to get work packet
[11:22:37] - Will indicate memory of 2046 MB
[11:22:37] - Connecting to assignment server
[11:22:37] Connecting to http://assign.stanford.edu:8080/
[11:22:37] Posted data.
[11:22:37] Initial: 40AB; - Successful: assigned to (171.64.65.63).
[11:22:37] + News From Folding@Home: Welcome to Folding@Home
[11:22:37] Loaded queue successfully.
[11:22:37] Connecting to http://171.64.65.63:8080/
[11:22:38] Posted data.
[11:22:38] Initial: 0000; - Receiving payload (expected size: 609129)
[11:22:39] - Downloaded at ~594 kB/s
[11:22:39] - Averaged speed for that direction ~491 kB/s
[11:22:39] + Received work.
[11:22:39] Trying to send all finished work units
[11:22:39] + No unsent completed units remaining.
[11:22:39] + Closed connections
[11:22:39] 
[11:22:39] + Processing work unit
[11:22:39] Work type a1 not eligible for variable processors
[11:22:39] Core required: FahCore_a1.exe
[11:22:39] Core found.
[11:22:39] Using generic mpiexec calls
[11:22:39] Working on queue slot 09 [May 20 11:22:39 UTC]
[11:22:39] + Working ...
[11:22:39] - Calling 'mpiexec -np 4 -channel auto -host 127.0.0.1 FahCore_a1.exe -dir work/ -suffix 09 -checkpoint 15 -verbose -lifeline 4924 -version 623'

[11:22:40] 
[11:22:40] *------------------------------*
[11:22:40] Folding@Home Gromacs SMP Core
[11:22:40] Version 1.74 (March 10, 2007)
[11:22:40] 
[11:22:40] Preparing to commence simulation
[11:22:40] - Ensuring status. Please wait- Created dyn
[11:22:40] - Files status OK
[11:22:40] - Expanded 608617 -> 3266341 (decompressed 536.6 percent)
[11:22:40] - Starting from initial work packet
[11:22:40] 
[11:22:40] Project: 3064 (Run 1, Clone 435, Gen 14)

And just recently:

Code: Select all

[07:39:00] Folding@home Core Shutdown: FINISHED_UNIT
[07:39:00] 
[07:39:00] Folding@home Core Shutdown: FINISHED_UNIT
[07:39:03] CoreStatus = 64 (100)
[07:39:03] Unit 9 finished with 21 percent of time to deadline remaining.
[07:39:03] Updated performance fraction: 0.446877
[07:39:03] Sending work to server
[07:39:03] Project: 3064 (Run 1, Clone 435, Gen 14)


[07:39:03] + Attempting to send results [May 23 07:39:03 UTC]
[07:39:03] - Reading file work/wuresults_09.dat from core
[07:39:03]   (Read 1832082 bytes from disk)
[07:39:03] Connecting to http://171.64.65.63:8080/
[07:39:04] Posted data.
[07:39:04] Initial: 0000; - Uploaded at ~895 kB/s
[07:39:05] - Averaged speed for that direction ~652 kB/s
[07:39:05] + Results successfully sent
[07:39:05] Thank you for your contribution to Folding@Home.
[07:39:05] + Number of Units Completed: 78

[07:39:09] - Warning: Could not delete all work unit files (9): Core returned invalid code
[07:39:09] Trying to send all finished work units
[07:39:09] + No unsent completed units remaining.
[07:39:09] - Preparing to get new work unit...
[07:39:09] + Attempting to get work packet
[07:39:09] - Will indicate memory of 2046 MB
[07:39:09] - Connecting to assignment server
[07:39:09] Connecting to http://assign.stanford.edu:8080/
[07:39:09] Posted data.
[07:39:09] Initial: 40AB; - Successful: assigned to (171.64.65.63).
[07:39:09] + News From Folding@Home: Welcome to Folding@Home
[07:39:09] Loaded queue successfully.
[07:39:09] Connecting to http://171.64.65.63:8080/
[07:39:10] Posted data.
[07:39:10] Initial: 0000; - Receiving payload (expected size: 608217)
[07:39:11] - Downloaded at ~593 kB/s
[07:39:11] - Averaged speed for that direction ~512 kB/s
[07:39:11] + Received work.
[07:39:11] Trying to send all finished work units
[07:39:11] + No unsent completed units remaining.
[07:39:11] + Closed connections
[07:39:11] 
[07:39:11] + Processing work unit
[07:39:11] Work type a1 not eligible for variable processors
[07:39:11] Core required: FahCore_a1.exe
[07:39:11] Core found.
[07:39:11] Using generic mpiexec calls
[07:39:11] Working on queue slot 00 [May 23 07:39:11 UTC]
[07:39:11] + Working ...
[07:39:11] - Calling 'mpiexec -np 4 -channel auto -host 127.0.0.1 FahCore_a1.exe -dir work/ -suffix 00 -checkpoint 15 -verbose -lifeline 4924 -version 623'

[07:39:12] 
[07:39:12] *------------------------------*
[07:39:12] Folding@Home Gromacs SMP Core
[07:39:12] Version 1.74 (March 10, 2007)
[07:39:12] 
[07:39:12] Preparing to commence simulation
[07:39:12] - Ensuring status. Please wait- Created dyn
[07:39:12] - Files status OK
[07:39:12] - Expanded 607705 -> 3255941 (decompressed 535.7 percent)
[07:39:12] - Starting from initial work packet
[07:39:12] 
[07:39:12] Project: 3064 (Run 0, Clone 224, Gen 36)

From prior discussions, iirc, I should go ahead and let the unit finish even though there's no way I'll finish before deadline?

[codysluder]

If you are certain that you cannot meet the final deadline, you should discard the WU manually. If you can finish it before the final deadline, you should keep working on it.

One of the SMP servers has been down and was recently fixed. That may be related.

Also, you should check to see which servers might be down. Try to avoid downloading additional WUs that you won't be able to finish.

[anko1]

Hi, codysluder. Thanks for the confirmation of the protocol. I wasn't very accurate in my post (running out the door for the weekend ), but yes, I meant that I could finish by the final deadline, but not the preferred.