Projects 6892 : Can't send results on Linux 32 bits

[martiou]

I got 12 of these WUs since 19/08/2011 19:24:15 UTC from server 171.67.108.53:80.
Clients are at work and I'm behind a proxy. So only port 80 is open, not port 8080.

I got 4 WUs on a Windows client v6.23. I finished 3 (the 4th is being calculated) and I send the results ((Run 72, Clone 2, Gen 1), (Run 483, Clone 9, Gen 1), (Run 533, Clone 13, Gen 0)).
Each time, client tries to send results to 171.67.108.53:8080, got status 504, and send results to 171.67.108.53:80.

But I got 8 WUs on 3 Linux 32 bits clients v6.02. I finished 5 (the 3 others are being calculated) and I can't send the results ((Run 606, Clone 9, Gen 0), (Run 527, Clone 8, Gen 1), (Run 513, Clone 9, Gen 0), (Run 941, Clone 12, Gen 0), (Run 308, Clone 5, Gen 0)).
Each time, client tries to send results to 171.67.108.53:8080, got status 504, and tries again to 171.67.108.53:8080, sometimes it tries to 171.65.103.160:8080 but it never tries on alternative port 80.

Is it possible to do something so I can send these WUs ?

Thanks for your help.

Martiou

[bruce]

According to serverstat, port 80 is open on 171.67.108.53 and when I open http://171.67.108.53 I do not get an error so I believe the server is configured correctly. I don't understand why your client is not trying port 80. Please post a segment of the log showing a typical upload attempt.

Which client version are you running?

Were these WUs downloaded through the same proxy or have they been moved from a non-proxy environment to a proxy environment?

[martiou]

Sorry for the late response but I was on holidays.

I still can't send the results and I have lost several.

I know that port 80 is open onserver 171.67.108.53 because my Windows client (v6.23) sent results by this port.

All the Linux clients which can't send results are in v6.02 because of Linux 32 bits.
They downloaded the WUs through the same proxy.

Here a segment of log

Code: Select all

[15:12:02] Connecting to http://assign2.stanford.edu:80/
[15:12:03] Posted data.
[15:12:03] Initial: 43AB; - Successful: assigned to (171.67.108.53).
[15:12:03] + News From Folding@Home: Welcome to Folding@Home
[15:12:03] Loaded queue successfully.
[15:12:03] Connecting to http://171.67.108.53:80/
[15:12:04] Posted data.
[15:12:04] Initial: 0000; - Receiving payload (expected size: 663774)
[15:12:07] - Downloaded at ~216 kB/s
[15:12:07] - Averaged speed for that direction ~292 kB/s
[15:12:07] + Received work.
[15:12:07] + Closed connections
[15:12:07] 
[15:12:07] + Processing work unit
[15:12:07] Core required: FahCore_78.exe
[15:12:07] Core found.
[15:12:07] Working on Unit 06 [August 30 15:12:07]
[15:12:07] + Working ...
[15:12:07] - Calling './FahCore_78.exe -dir work/ -suffix 06 -checkpoint 3 -verbose -lifeline 18874 -version 602'

[15:12:07] 
[15:12:07] *------------------------------*
[15:12:07] Folding@Home Gromacs Core
[15:12:07] Version 1.90 (March 8, 2006)
[15:12:07] 
[15:12:07] Preparing to commence simulation
[15:12:07] - Looking at optimizations...
[15:12:07] - Created dyn
[15:12:07] - Files status OK
[15:12:07] - Expanded 663262 -> 3332352 (decompressed 502.4 percent)
[15:12:07] - Starting from initial work packet
[15:12:07] 
[15:12:07] Project: 6892 (Run 15, Clone 0, Gen 18)
[15:12:07] 
[15:12:07] Assembly optimizations on if available.
[15:12:07] Entering M.D.
[15:12:14] Protein: ALZHEIMER DISEASE AMYLOID
[15:12:14] 
[15:12:14] Writing local files
[15:12:48] Extra SSE boost OK.
[15:12:49] Writing local files
[15:12:49] Completed 0 out of 250000 steps  (0%)
.
.
.
[16:55:53] Completed 247500 out of 250000 steps  (99%)
[16:58:54] Timered checkpoint triggered.
[17:01:54] Timered checkpoint triggered.
[17:04:55] Timered checkpoint triggered.
[17:07:55] Timered checkpoint triggered.
[17:10:56] Timered checkpoint triggered.
[17:11:27] Writing local files
[17:11:27] Completed 250000 out of 250000 steps  (100%)
[17:11:27] Writing final coordinates.
[17:11:27] Past main M.D. loop
[17:12:27] 
[17:12:27] Finished Work Unit:
[17:12:27] - Reading up to 548184 from "work/wudata_06.arc": Read 548184
[17:12:27] - Reading up to 491892 from "work/wudata_06.xtc": Read 491892
[17:12:27] goefile size: 0
[17:12:27] logfile size: 21580
[17:12:27] Leaving Run
[17:12:30] - Writing 1067564 bytes of core data to disk...
[17:12:30] Done: 1067052 -> 1017036 (compressed to 95.3 percent)
[17:12:30]   ... Done.
[17:12:31] - Shutting down core
[17:12:31] 
[17:12:31] Folding@home Core Shutdown: FINISHED_UNIT
[17:12:32] CoreStatus = 64 (100)
[17:12:32] Unit 6 finished with 97 percent of time to deadline remaining.
[17:12:32] Updated performance fraction: 0.967396
[17:12:32] Sending work to server
[17:12:32] - Read packet limit of 540015616... Set to 524286976.


[17:12:32] + Attempting to send results
[17:12:32] - Reading file work/wuresults_06.dat from core
[17:12:32]   (Read 1017548 bytes from disk)
[17:12:32] Connecting to http://171.67.108.53:8080/
[17:15:32] - Couldn't send HTTP request to server
[17:15:32]   (Got status 504)
[17:15:32] + Could not connect to Work Server (results)
[17:15:32]     (171.67.108.53:8080)
[17:15:32] - Error: Could not transmit unit 06 (completed August 31) to work server.
[17:15:32] - 1 failed uploads of this unit.
[17:15:32]   Keeping unit 06 in queue.
[17:15:32] Trying to send all finished work units
[17:15:32] - Read packet limit of 540015616... Set to 524286976.


[17:15:32] + Attempting to send results
[17:15:32] - Reading file work/wuresults_00.dat from core
[17:15:32]   (Read 1014737 bytes from disk)
[17:15:32] Connecting to http://171.67.108.53:8080/
[17:18:32] - Couldn't send HTTP request to server
[17:18:32]   (Got status 504)
[17:18:32] + Could not connect to Work Server (results)
[17:18:32]     (171.67.108.53:8080)
[17:18:32] - Error: Could not transmit unit 00 (completed August 25) to work server.
[17:18:32] - 33 failed uploads of this unit.
[17:18:32] - Read packet limit of 540015616... Set to 524286976.


[17:18:32] + Attempting to send results
[17:18:32] - Reading file work/wuresults_00.dat from core
[17:18:32]   (Read 1014737 bytes from disk)
[17:18:32] Connecting to http://171.65.103.160:8080/
[17:21:32] - Couldn't send HTTP request to server
[17:21:32]   (Got status 504)
[17:21:32] + Could not connect to Work Server (results)
[17:21:32]     (171.65.103.160:8080)
[17:21:32]   Could not transmit unit 00 to Collection server; keeping in queue.
[17:21:32] - Read packet limit of 540015616... Set to 524286976.


[17:21:32] + Attempting to send results
[17:21:32] - Reading file work/wuresults_01.dat from core
[17:21:32]   (Read 1017452 bytes from disk)
[17:21:32] Connecting to http://171.67.108.53:8080/
[17:24:32] - Couldn't send HTTP request to server
[17:24:32]   (Got status 504)
[17:24:32] + Could not connect to Work Server (results)
[17:24:32]     (171.67.108.53:8080)
[17:24:32] - Error: Could not transmit unit 01 (completed August 26) to work server.
[17:24:32] - 23 failed uploads of this unit.
[17:24:32] - Read packet limit of 540015616... Set to 524286976.


[17:24:32] + Attempting to send results
[17:24:32] - Reading file work/wuresults_01.dat from core
[17:24:32]   (Read 1017452 bytes from disk)
[17:24:32] Connecting to http://171.65.103.160:8080/
[17:27:32] - Couldn't send HTTP request to server
[17:27:32]   (Got status 504)
[17:27:32] + Could not connect to Work Server (results)
[17:27:32]     (171.65.103.160:8080)
[17:27:32]   Could not transmit unit 01 to Collection server; keeping in queue.
[17:27:32] - Read packet limit of 540015616... Set to 524286976.

Thank you for your help

Martiou

[martiou]

Finally, I could send the results of some Wus by transferring them at home and then sending them from a virtual machine on my computer.
But I lost 6 results.

It seems the issue is only with v6.02 because I can send results from Windows client v6.23 and from Linux client 64 bits v6.29.

Is it possible on your side to stop giving 6892 Wus to Linux v6.04 port 80 clients ?
I still get 6892 WUs occasionnaly on theses clients and I have to transfer them at home to send the results.
I may therefore be improper handling and delete unsent results.

Thank you for your help

Martiou

[bruce]

It's not really clear whether this is a Linux-32 vs. Linux-64 problem or if it's a Port 80 proxy problem. I've notified the Pande Group about the issue.

[martiou]

I think it's not a Linux-32 vs Linux 64 but v6.02 vs post v6.02.

It's not a port 80 proxy problem because my Windows client v6.23 and my Linux 64 clients v6.29 are at work behind a proxy as my Linux 32 clients v6.02.
And only Linux 32 clients v6.02 can't send results fro these WUs.

[bruce]

It's not likely that v6.02 will be updated. At this point all development work is going into V7. Hopefully there will be a version that works on your 32-bit version of Linux. I know there are a number of unresolved issues that are probably related to this and I'm not sure how far along the developers are in fixing them for the next beta or for the eventual open release.