Trying my hand at SMP

[bubba]

Well I set the test bench up with the SMP client. Running Q6600 @ 3.0Ghz. Never done it before, but looks like I've got it setup right. Right now it looks like its averaging 15min steps.

Not that I will catch some of you guys, but I have fallen 13 places to GPU and SMP clients don't feel much like falling any more

[Darkstar]

Sweet! we can use some more PPD's..... way to go.....

[stopthekilling77]

Awesome! I'm trying out the GPU client at the moment, seems to be stable for me!

[KnightRid]

Spiffy!!!

Mike

MORE POWER BABY

[bubba]

Should turn in my first SMP unit around 2 this afternoon lets hope I have all the flags set right

[dicecca112]

you want -smp -verbosity 9 -forceasm -advmethods

[bubba]

don't have the "-forceasm -advmethods" part, so might not have any points for the weekend.

[dicecca112]

no you'll be fine

forceasm protects you from loosing WUs if the client ends because of lose of network connectivity or accidental shutdown

advmethods allows you to get the A2 core which utilizes the whole CPU better, the client with A1 really uses only 80%.

[bubba]

Well the points showed up, looks like I did something right, first SMP WU got me 1,920.

Now if I can talk Nate into upgrading the test bench video card

[vbironchef]

bubba , and all folders, just want to let you know that it's not easy to restart the computer and start the SMP client. Try to make a shortcut so you can start the SMP client with ease. http://fahwiki.net/index.php/How_do_I_a ... _client%3F This should ease your pain.

[bubba]

tis the way I have it

[bubba]

Ok, does this mean it sent in the WU and not able to get a new one, or has failed to do both?

Code: Select all

[07:30:53] 
[07:30:53] Finished Work Unit:
[07:30:53] - Reading up to 21193200 from "work/wudata_03.arc": Read 21193200
[07:30:53] - Reading up to 532060 from "work/wudata_03.xtc": Read 532060
[07:30:53] goefile size: 0
[07:30:53] logfile size: 203301
[07:30:53] Leaving Run
[07:30:56] - Writing 21934933 bytes of core data to disk...
[07:30:56]   ... Done.
[07:30:56] - Failed to delete work/wudata_03.sas
[07:30:56] - Failed to delete work/wudata_03.goe
[07:30:56] Warning:  check for stray files
[07:30:56] - Shutting down core
[07:32:56] 
[07:32:56] Folding@home Core Shutdown: FINISHED_UNIT
[07:32:56] 
[07:32:56] Folding@home Core Shutdown: FINISHED_UNIT
[07:32:59] CoreStatus = 64 (100)
[07:32:59] Unit 3 finished with 84 percent of time to deadline remaining.
[07:32:59] Updated performance fraction: 0.834406
[07:32:59] Sending work to server
[07:32:59] Project: 2665 (Run 1, Clone 989, Gen 39)


[07:32:59] + Attempting to send results [July 25 07:32:59 UTC]
[07:32:59] - Reading file work/wuresults_03.dat from core
[07:32:59]   (Read 21934933 bytes from disk)
[07:32:59] Connecting to http://171.64.65.64:8080/
[07:39:45] - Autosending finished units... [July 25 07:39:45 UTC]
[07:39:45] Trying to send all finished work units
[07:39:45] - Already sending work
[07:39:45] + Sent 0 of 1 completed units to the server
[07:39:45] - Autosend completed
[07:52:59] Posted data.
[08:12:59] Initial: 0000; + Could not connect to Work Server (results)
[08:32:59]     (171.64.65.64:8080)
[08:32:59] + Retrying using alternative port
[08:32:59] Connecting to http://171.64.65.64:80/
[08:53:00] Posted data.
[09:13:00] Initial: 0000; 

[Darkstar]

looks like it sent the work, as it says posted data..maybe one of the folding experts here know, with the new Beta's its hard to tell whats going on sometimes. You might want to post it in the folding forum too to see what they have to say. I didnt see anything about "attempting to get work" or downloading...any more data following what you posted?

[geokilla]

Uhh, I'm not sure if it did or not. Cus it says sent 0 of 1 completed units to the server. In my console based versions, if the WU did get sent, it'll state what the upload speed was, etc.

With everyone jumping on the SMP bandwagon, I'm getting the urge to do the same as well. But sadly, I system will not make the deadlines in time

[bubba]

Still not sending....

Code: Select all

[07:29:52] Completed 250000 out of 250000 steps  (100 percent)
[07:29:52] Writing final coordinates.
[07:29:53] Past main M.D. loop
[07:29:53] Will end MPI now
[07:30:53] 
[07:30:53] Finished Work Unit:
[07:30:53] - Reading up to 21193200 from "work/wudata_03.arc": Read 21193200
[07:30:53] - Reading up to 532060 from "work/wudata_03.xtc": Read 532060
[07:30:53] goefile size: 0
[07:30:53] logfile size: 203301
[07:30:53] Leaving Run
[07:30:56] - Writing 21934933 bytes of core data to disk...
[07:30:56]   ... Done.
[07:30:56] - Failed to delete work/wudata_03.sas
[07:30:56] - Failed to delete work/wudata_03.goe
[07:30:56] Warning:  check for stray files
[07:30:56] - Shutting down core
[07:32:56] 
[07:32:56] Folding@home Core Shutdown: FINISHED_UNIT
[07:32:56] 
[07:32:56] Folding@home Core Shutdown: FINISHED_UNIT
[07:32:59] CoreStatus = 64 (100)
[07:32:59] Unit 3 finished with 84 percent of time to deadline remaining.
[07:32:59] Updated performance fraction: 0.834406
[07:32:59] Sending work to server
[07:32:59] Project: 2665 (Run 1, Clone 989, Gen 39)


[07:32:59] + Attempting to send results [July 25 07:32:59 UTC]
[07:32:59] - Reading file work/wuresults_03.dat from core
[07:32:59]   (Read 21934933 bytes from disk)
[07:32:59] Connecting to http://171.64.65.64:8080/
[07:39:45] - Autosending finished units... [July 25 07:39:45 UTC]
[07:39:45] Trying to send all finished work units
[07:39:45] - Already sending work
[07:39:45] + Sent 0 of 1 completed units to the server
[07:39:45] - Autosend completed
[07:52:59] Posted data.
[08:12:59] Initial: 0000; + Could not connect to Work Server (results)
[08:32:59]     (171.64.65.64:8080)
[08:32:59] + Retrying using alternative port
[08:32:59] Connecting to http://171.64.65.64:80/
[08:53:00] Posted data.
[09:13:00] Initial: 0000; + Could not connect to Work Server (results)
[09:33:00]     (171.64.65.64:80)
[09:33:00] - Error: Could not transmit unit 03 (completed July 25) to work server.
[09:33:00] - 1 failed uploads of this unit.
[09:33:00]   Keeping unit 03 in queue.
[09:33:00] Trying to send all finished work units
[09:33:00] Project: 2665 (Run 1, Clone 989, Gen 39)


[09:33:00] + Attempting to send results [July 25 09:33:00 UTC]
[09:33:00] - Reading file work/wuresults_03.dat from core
[09:33:00]   (Read 21934933 bytes from disk)
[09:33:00] Connecting to http://171.64.65.64:8080/
[09:53:00] Posted data.
[10:13:00] Initial: 0000; + Could not connect to Work Server (results)
[10:33:00]     (171.64.65.64:8080)
[10:33:00] + Retrying using alternative port
[10:33:00] Connecting to http://171.64.65.64:80/
[10:53:00] Posted data.
[11:13:00] Initial: 0000; - Unknown packet returned from server, expected ACK for results
[11:30:24] - Error: Could not transmit unit 03 (completed July 25) to work server.
[11:30:24] - 2 failed uploads of this unit.


[11:30:24] + Attempting to send results [July 25 11:30:24 UTC]
[11:30:24] - Reading file work/wuresults_03.dat from core
[11:30:24]   (Read 21934933 bytes from disk)
[11:30:24] Connecting to http://171.64.122.86:8080/
[11:50:25] Posted data.
[12:10:25] Initial: 0000; + Could not connect to Work Server (results)
[12:30:25]     (171.64.122.86:8080)
[12:30:25] + Retrying using alternative port
[12:30:25] Connecting to http://171.64.122.86:80/
[12:30:25] - Couldn't send HTTP request to server
[12:30:25]   (Got status 503)
[12:30:25] + Could not connect to Work Server (results)
[12:30:25]     (171.64.122.86:80)
[12:30:25]   Could not transmit unit 03 to Collection server; keeping in queue.
[12:30:25] + Sent 0 of 1 completed units to the server
[12:30:25] - Preparing to get new work unit...
[12:30:25] + Attempting to get work packet
[12:30:25] - Will indicate memory of 2044 MB
[12:30:25] - Connecting to assignment server
[12:30:25] Connecting to http://assign.stanford.edu:8080/
[12:30:25] Posted data.
[12:30:25] Initial: 40AB; - Successful: assigned to (171.64.65.64).
[12:30:25] + News From Folding@Home: Welcome to Folding@Home
[12:30:26] Loaded queue successfully.
[12:30:26] Connecting to http://171.64.65.64:8080/
[12:30:31] Posted data.
[12:30:31] Initial: 0000; - Receiving payload (expected size: 4851152)
[12:30:48] - Downloaded at ~278 kB/s
[12:30:48] - Averaged speed for that direction ~304 kB/s
[12:30:48] + Received work.
[12:30:48] Trying to send all finished work units
[12:30:48] Project: 2665 (Run 1, Clone 989, Gen 39)


[12:30:48] + Attempting to send results [July 25 12:30:48 UTC]
[12:30:48] - Reading file work/wuresults_03.dat from core
[12:30:48]   (Read 21934933 bytes from disk)
[12:30:48] Connecting to http://171.64.65.64:8080/

Anyway around this and not lose the work?????

[DMB2000uk]

Is it actually working on other things at the minute?

If it is, you don't need to worry too much as the client will still try to send back what it has done occasionally until it is able to get through to stanford's servers.

Dan

[bubba]

Its not working on anything, just setting there and trying to send the data.

[DMB2000uk]

Hmm, well that sucks.

If you know internet connectivity is fine, then the only thing you can really do is wait it out.

Dan

[bubba]

Its connecting ok, I made the post with the code from that machine.

Tis a waiting game now... sucks too I was getting some nice points form it.

[geokilla]

I'm not sure if restarting your computer would cause you to lose your WU that you just finished, but if it doesn't, restart your computer.