Page 1 of 1

v7 default linux failing on A3 cores?

PostPosted: Mon Jul 08, 2013 1:19 pm
by k1wi
I the installed the latest 32bit v7 client on my ubuntu build, running a Celeron G540 for 2k PPD on A4s. However, whenever it downloads an A3 work unit the following error always occurs.

Code: Select all
******************************* Date: 2013-07-08 *******************************
11:20:47:WU00:FS00:0xa4:Completed 1940000 out of 2000000 steps  (97%)
11:31:57:WU00:FS00:0xa4:Completed 1960000 out of 2000000 steps  (98%)
11:31:58:WU01:FS00:Connecting to assign3.stanford.edu:8080
11:31:59:WU01:FS00:News: Welcome to Folding@Home
11:31:59:WU01:FS00:Assigned to work server 128.143.231.202
11:31:59:WU01:FS00:Requesting new work unit for slot 00: RUNNING cpu:2 from 128.143.231.202
11:31:59:WU01:FS00:Connecting to 128.143.231.202:8080
11:32:00:WU01:FS00:Downloading 3.67MiB
11:32:04:WU01:FS00:Download complete
11:32:04:WU01:FS00:Received Unit: id:01 state:DOWNLOAD error:NO_ERROR project:8572 run:0 clone:3 gen:104 core:0xa3 unit:0x0000053c0a3b1e5951225b33286fe337
11:43:17:WU00:FS00:0xa4:Completed 1980000 out of 2000000 steps  (99%)
11:54:26:WU00:FS00:0xa4:Completed 2000000 out of 2000000 steps  (100%)
11:54:27:WU00:FS00:0xa4:DynamicWrapper: Finished Work Unit: sleep=10000
11:54:37:WU00:FS00:0xa4:
11:54:37:WU00:FS00:0xa4:Finished Work Unit:
11:54:37:WU00:FS00:0xa4:- Reading up to 3558984 from "00/wudata_01.trr": Read 3558984
11:54:37:WU00:FS00:0xa4:trr file hash check passed.
11:54:37:WU00:FS00:0xa4:- Reading up to 530284 from "00/wudata_01.xtc": Read 530284
11:54:37:WU00:FS00:0xa4:xtc file hash check passed.
11:54:37:WU00:FS00:0xa4:edr file hash check passed.
11:54:37:WU00:FS00:0xa4:logfile size: 85051
11:54:37:WU00:FS00:0xa4:Leaving Run
11:54:39:WU00:FS00:0xa4:- Writing 4182243 bytes of core data to disk...
11:54:40:WU00:FS00:0xa4:Done: 4181731 -> 3925087 (compressed to 93.8 percent)
11:54:40:WU00:FS00:0xa4:  ... Done.
11:55:49:WU00:FS00:0xa4:- Shutting down core
11:55:49:WU00:FS00:0xa4:
11:55:49:WU00:FS00:0xa4:Folding@home Core Shutdown: FINISHED_UNIT
11:55:53:WU00:FS00:FahCore returned: FINISHED_UNIT (100 = 0x64)
11:55:53:WU00:FS00:Sending unit results: id:00 state:SEND error:NO_ERROR project:10450 run:866 clone:0 gen:89 core:0xa4 unit:0x0000007c0a3b1e7550a53aa3fa6507be
11:55:53:WU00:FS00:Uploading 3.74MiB to 171.64.65.81
11:55:53:WU00:FS00:Connecting to 171.64.65.81:8080
11:55:53:WU01:FS00:Starting
11:55:53:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 01 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:55:53:WU01:FS00:Started FahCore on PID 8920
11:55:53:WU01:FS00:Core PID:8924
11:55:53:WU01:FS00:FahCore 0xa3 started
11:55:53:WARNING:WU01:FS00:FahCore returned: FAILED_3 (255 = 0xff)
11:55:54:WU01:FS00:Starting
11:55:54:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 01 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:55:54:WU01:FS00:Started FahCore on PID 8927
11:55:54:WU01:FS00:Core PID:8931
11:55:54:WU01:FS00:FahCore 0xa3 started
11:55:54:WARNING:WU01:FS00:FahCore returned: FAILED_3 (255 = 0xff)
11:55:59:WU00:FS00:Upload 81.80%
11:56:01:WU00:FS00:Upload complete
11:56:01:WU00:FS00:Server responded WORK_ACK (400)
11:56:01:WU00:FS00:Final credit estimate, 1265.00 points
11:56:01:WU00:FS00:Cleaning up
11:56:54:WU01:FS00:Starting
11:56:54:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 01 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:56:54:WU01:FS00:Started FahCore on PID 8968
11:56:54:WU01:FS00:Core PID:8972
11:56:54:WU01:FS00:FahCore 0xa3 started
11:56:54:WARNING:WU01:FS00:FahCore returned: FAILED_3 (255 = 0xff)
11:57:54:WU01:FS00:Starting
11:57:54:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 01 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:57:54:WU01:FS00:Started FahCore on PID 9022
11:57:54:WU01:FS00:Core PID:9026
11:57:54:WU01:FS00:FahCore 0xa3 started
11:57:54:WARNING:WU01:FS00:FahCore returned: FAILED_3 (255 = 0xff)
11:58:54:WU01:FS00:Starting
11:58:54:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 01 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:58:54:WU01:FS00:Started FahCore on PID 9057
11:58:54:WU01:FS00:Core PID:9061
11:58:54:WU01:FS00:FahCore 0xa3 started
11:58:54:WARNING:WU01:FS00:FahCore returned: FAILED_3 (255 = 0xff)
11:58:54:WARNING:WU01:FS00:Too many errors, failing
11:58:54:WU01:FS00:Sending unit results: id:01 state:SEND error:FAILED project:8572 run:0 clone:3 gen:104 core:0xa3 unit:0x0000053c0a3b1e5951225b33286fe337
11:58:55:WU01:FS00:Connecting to 128.143.231.202:8080
11:58:55:WU01:FS00:Server responded WORK_QUIT (404)
11:58:55:WARNING:WU01:FS00:Server did not like results, dumping
11:58:55:WU01:FS00:Cleaning up
11:58:55:WU00:FS00:Connecting to assign3.stanford.edu:8080
11:58:55:WU00:FS00:News: Welcome to Folding@Home
11:58:55:WU00:FS00:Assigned to work server 128.143.231.202
11:58:55:WU00:FS00:Requesting new work unit for slot 00: READY cpu:2 from 128.143.231.202
11:58:55:WU00:FS00:Connecting to 128.143.231.202:8080
11:58:57:WU00:FS00:Downloading 3.64MiB
11:59:01:WU00:FS00:Download complete
11:59:01:WU00:FS00:Received Unit: id:00 state:DOWNLOAD error:NO_ERROR project:6099 run:5 clone:6 gen:427 core:0xa3 unit:0x0000023a0a3b1e594e88a01f5a0fb850
11:59:01:WU00:FS00:Starting
11:59:01:WU00:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 00 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:59:01:WU00:FS00:Started FahCore on PID 9093
11:59:01:WU00:FS00:Core PID:9097
11:59:01:WU00:FS00:FahCore 0xa3 started
11:59:01:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)
11:59:02:WU00:FS00:Starting
11:59:02:WU00:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 00 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
11:59:02:WU00:FS00:Started FahCore on PID 9100
11:59:02:WU00:FS00:Core PID:9104
11:59:02:WU00:FS00:FahCore 0xa3 started
11:59:02:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)
12:00:02:WU00:FS00:Starting
12:00:02:WU00:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 00 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
12:00:02:WU00:FS00:Started FahCore on PID 9161
12:00:02:WU00:FS00:Core PID:9165
12:00:02:WU00:FS00:FahCore 0xa3 started
12:00:02:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)
12:01:02:WU00:FS00:Starting
12:01:02:WU00:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 00 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
12:01:02:WU00:FS00:Started FahCore on PID 9202
12:01:02:WU00:FS00:Core PID:9206
12:01:02:WU00:FS00:FahCore 0xa3 started
12:01:02:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)
12:02:02:WU00:FS00:Starting
12:02:02:WU00:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a3.fah/FahCore_a3 -dir 00 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
12:02:02:WU00:FS00:Started FahCore on PID 9243
12:02:02:WU00:FS00:Core PID:9247
12:02:02:WU00:FS00:FahCore 0xa3 started
12:02:03:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)
12:02:03:WARNING:WU00:FS00:Too many errors, failing
12:02:03:WU00:FS00:Sending unit results: id:00 state:SEND error:FAILED project:6099 run:5 clone:6 gen:427 core:0xa3 unit:0x0000023a0a3b1e594e88a01f5a0fb850
12:02:03:WU00:FS00:Connecting to 128.143.231.202:8080
12:02:03:WU00:FS00:Server responded WORK_QUIT (404)
12:02:03:WARNING:WU00:FS00:Server did not like results, dumping
12:02:03:WU00:FS00:Cleaning up
12:02:03:WU01:FS00:Connecting to assign3.stanford.edu:8080
12:02:04:WU01:FS00:News: Welcome to Folding@Home
12:02:04:WU01:FS00:Assigned to work server 155.247.166.219
12:02:04:WU01:FS00:Requesting new work unit for slot 00: READY cpu:2 from 155.247.166.219
12:02:04:WU01:FS00:Connecting to 155.247.166.219:8080
12:02:04:WU01:FS00:Downloading 80.99KiB
12:02:04:WU01:FS00:Download complete
12:02:04:WU01:FS00:Received Unit: id:01 state:DOWNLOAD error:NO_ERROR project:6357 run:4 clone:8 gen:125 core:0xa4 unit:0x000000a70002894b5147f625be3cc627
12:02:04:WU01:FS00:Starting
12:02:04:WU01:FS00:Running FahCore: /usr/bin/FAHCoreWrapper /var/lib/fahclient/cores/www.stanford.edu/~pande/Linux/x86/Core_a4.fah/FahCore_a4 -dir 01 -suffix 01 -version 703 -lifeline 1165 -checkpoint 15 -np 2
12:02:04:WU01:FS00:Started FahCore on PID 9250
12:02:04:WU01:FS00:Core PID:9254
12:02:04:WU01:FS00:FahCore 0xa4 started
12:02:05:WU01:FS00:0xa4:
12:02:05:WU01:FS00:0xa4:*------------------------------*
12:02:05:WU01:FS00:0xa4:Folding@Home Gromacs GB Core
12:02:05:WU01:FS00:0xa4:Version 2.27 (Dec. 15, 2010)
12:02:05:WU01:FS00:0xa4:
12:02:05:WU01:FS00:0xa4:Preparing to commence simulation
12:02:05:WU01:FS00:0xa4:- Looking at optimizations...
12:02:05:WU01:FS00:0xa4:- Created dyn
12:02:05:WU01:FS00:0xa4:- Files status OK
12:02:05:WU01:FS00:0xa4:- Expanded 82418 -> 175532 (decompressed 212.9 percent)
12:02:05:WU01:FS00:0xa4:Called DecompressByteArray: compressed_data_size=82418 data_size=175532, decompressed_data_size=175532 diff=0
12:02:05:WU01:FS00:0xa4:- Digital signature verified
12:02:05:WU01:FS00:0xa4:
12:02:05:WU01:FS00:0xa4:Project: 6357 (Run 4, Clone 8, Gen 125)
12:02:05:WU01:FS00:0xa4:
12:02:05:WU01:FS00:0xa4:Assembly optimizations on if available.
12:02:05:WU01:FS00:0xa4:Entering M.D.
12:02:11:WU01:FS00:0xa4:Mapping NT from 2 to 2
12:02:11:WU01:FS00:0xa4:Completed 0 out of 5000000 steps  (0%)
12:07:24:WU01:FS00:0xa4:Completed 50000 out of 5000000 steps  (1%)
12:12:40:WU01:FS00:0xa4:Completed 100000 out of 5000000 steps  (2%)
Anyone have a clue as to why?

Re: v7 default linux failing on A3 cores?

PostPosted: Mon Jul 08, 2013 4:23 pm
by bruce
You said you loaded the 32-bit FAHClient. Is your Linux 32- or 64-bit? If it's reasonable, I recommend you run both 64-bit codes. If not, we can explore the possibilities with you.

Re: v7 default linux failing on A3 cores?

PostPosted: Mon Jul 08, 2013 4:56 pm
by k1wi
It's a 32bit version of Ubuntu.

Re: v7 default linux failing on A3 cores?

PostPosted: Mon Jul 08, 2013 8:47 pm
by folding_hoomer
May be your Core A3 is corrupted?

Try to delete Core A3 and let the client dowload it from scratch.

Re: v7 default linux failing on A3 cores?

PostPosted: Sun Jul 14, 2013 11:01 pm
by PantherX
I was under the impression that FahCore_a3 needs 64-bit but FahCore_a4 can work on 32/64 bit. The reason being that there are UNI WUs for FahCore_a4 but none for FahCore_a3.

Re: v7 default linux failing on A3 cores?

PostPosted: Mon Jul 15, 2013 3:10 am
by bruce
You may be right, PantherX, but if that's true, the AS should be smart enough to assign either A3 or A4 projects to 64-bit Linux and only A4 projects to 32-bit Linux. Do we have definitive reports of that policy working as stated or failing to work as stated?

If it works as intended, the 32-bit Linux folks might get no WUs when a critical server is down or it runs out of WUs while the folks with 64-bit carry on since from their perspective, nothing has happened.

Re: v7 default linux failing on A3 cores?

PostPosted: Mon Jul 15, 2013 3:26 am
by 7im
Contact the researcher and find out about the bitness...

Re: v7 default linux failing on A3 cores?

PostPosted: Mon Jul 15, 2013 12:45 pm
by PantherX
I don't think that the AS even knows about the "existence" of 32-bit Linux. The reason being is that with v6, there was only 64-bit Linux clients and AFAIK, this wasn't changed when V7 was released. It was discovered that 32-bit Linux can fold with FahCore_a4 since that FahCore can run on 32/64 bit systems. Similar to the fact that OSX can be configured for UNI WUs since FahCore_a4 can run on a single CPU.

Re: v7 default linux failing on A3 cores?

PostPosted: Thu Jul 18, 2013 6:44 pm
by k1wi
Is P8578 a 64bit project?

If so, that would explain why its failing...

Re: v7 default linux failing on A3 cores?

PostPosted: Fri Jul 19, 2013 4:42 am
by bruce
P8578 uses FahCore_a3 which is a 64-bit core, so yes. As PantherX said, traditionally, only 64-bit linux was supported until somebody discovered that a few projects would run on 32-bit linux. The design goals for FAH targeted 64-bit linux plus 32-bit libraries.

Re: v7 default linux failing on A3 cores?

PostPosted: Fri Jul 19, 2013 10:23 am
by k1wi
Then they should either pull the bit about '32bit Linux (.deb + .rpm)' from the download page or stop those clients from being able to download A3s?

Re: v7 default linux failing on A3 cores?

PostPosted: Fri Jul 19, 2013 12:32 pm
by Joe_H
I checked the download directories on the Stanford F@H site. There are two different cores available for the client to download for Linux, one in the x86 path and the other in the AMD64.
Code: Select all
http://www.stanford.edu/~pande/Linux/x86/

Core_a3.fah   11-Jun-2010 16:04   2.2M   


http://www.stanford.edu/~pande/Linux/AMD64/

Core_a3.fah   02-Mar-2011 09:38   2.6M   

I don't have a Linux system active currently, so can't examine the cores after downloading to see what bitness the executable from each path actually is. But the difference in creation dates could also point to version differences. Usually the core version is the first thing entered into the log after starting, but from the log posted there is an immediate crash reported instead.

So the A3 core available in the x86 path may or may not actually be 32-bit and is possibly an older version than needed by the projects assigned. It at least looks like an attempt was made at one time to support folding on 32-bit Linux with the A3 core, further digging through posts from 3 years ago looks to be needed to see if there is any reports on problems.

Re: v7 default linux failing on A3 cores?

PostPosted: Fri Jul 19, 2013 1:37 pm
by kasson
My recollection is that the problem is as follows:
older client versions always look in the x86 directory, whether 32-bit or 64-bit. The v7 client looks in x86 for 32-bit and AMD64 for 64-bit.
The Linux A3 client has IIRC always been 64-bit only. Older clients didn't support 32-bit attempts here.
(In general, unless you have a specific reason you need 32-bit linux, we'd really recommend 64-bit linux, but that's another issue...)