FahCore_a7 doesn't work on Unsupported Linux distros (CentOS)

[RamonetB]

ERROR:Guru Meditation: not checked 'config.xml' is caused by running the core in the client's directory. The core sees and reads config.xml from the current directory but that config is for the client not the core. If you run the core manually in a directory that does not contain config.xml this error should go away.

Ah. This explains that output then.

Bruce, I was mistaken. I was running the A7 core manually and generating this error. However, after starting everything using the FAHClient service script or binary, it is working fine on CentOS 7.2 and producing great results!

As for CentOS 6.8, I've gotten all the missing libraries installed (glibc-2.17 and libstdc++.so.0.6 that supports GLIBCXX 3.14.5). When running the A7 core manually with config.xml present:

Code: Select all

/opt/glibc-2.17/lib/ld-linux-x86-64.so.2 --library-path /opt/glibc-2.17/lib:/opt/glibc-2.17/lib64/:/lib:/lib64:/usr/lib64 ./FahCore_a7 

it produces the same result above "implying" that the core is launching correctly, or as far as it can go when done manually. Properly, with no config.xml present, no errors are generated on manually running the core.

But, when calling the service script or binary, I get the following error: FAILED_3 (255 = 0xff)

Full log:

Code: Select all

*********************** Log Started 2016-11-09T14:59:01Z ***********************
14:59:01:************************* Folding@home Client *************************
14:59:01:    Website: http://folding.stanford.edu/
14:59:01:  Copyright: (c) 2009-2016 Stanford University
14:59:01:     Author: Joseph Coffland <[email protected]>
14:59:01:       Args:
14:59:01:     Config: /root/Folding/config.xml
14:59:01:******************************** Build ********************************
14:59:01:    Version: 7.4.15
14:59:01:       Date: Aug 17 2016
14:59:01:       Time: 04:32:12
14:59:01: Repository: Git
14:59:01:   Revision: 4f3e0e25571a9f691719f0c273739294bde517dd
14:59:01:     Branch: master
14:59:01:   Compiler: GNU 4.4.7 20120313 (Red Hat 4.4.7-17)
14:59:01:    Options: -std=gnu++98 -O3 -funroll-loops -ffast-math -mfpmath=sse
14:59:01:             -fno-unsafe-math-optimizations -msse2
14:59:01:   Platform: linux2 4.6.0-1-amd64
14:59:01:       Bits: 64
14:59:01:       Mode: Release
14:59:01:******************************* System ********************************
14:59:01:        CPU: AMD Opteron(tm) Processor 6176 SE
14:59:01:     CPU ID: AuthenticAMD Family 16 Model 9 Stepping 1
14:59:01:       CPUs: 48
14:59:01:     Memory: 126.03GiB
14:59:01:Free Memory: 120.48GiB
14:59:01:    Threads: POSIX_THREADS
14:59:01: OS Version: 2.6
14:59:01:Has Battery: false
14:59:01: On Battery: false
14:59:01: UTC Offset: 0
14:59:01:        PID: 17058
14:59:01:        CWD: /root/Folding
14:59:01:         OS: Linux 2.6.32-642.6.2.el6.x86_64 x86_64
14:59:01:    OS Arch: AMD64
14:59:01:       GPUs: 0
14:59:01:       CUDA: Not detected
14:59:01:     OpenCL: Not detected
14:59:01:***********************************************************************
14:59:01:<config>
14:59:01:  <!-- Folding Slot Configuration -->
14:59:01:  <gpu v='false'/>
14:59:01:
14:59:01:  <!-- Network -->
14:59:01:  <proxy v=':8080'/>
14:59:01:
14:59:01:  <!-- Slot Control -->
14:59:01:  <power v='light'/>
14:59:01:
14:59:01:  <!-- User Information -->
14:59:01:  <passkey v='********************************'/>
14:59:01:  <team v='32'/>
14:59:01:  <user v='RamonetB'/>
14:59:01:
14:59:01:  <!-- Folding Slots -->
14:59:01:  <slot id='0' type='CPU'>
14:59:01:    <cpus v='48'/>
14:59:01:  </slot>
14:59:01:</config>
14:59:01:Trying to access database...
14:59:01:Successfully acquired database lock
14:59:01:Enabled folding slot 00: READY cpu:48
^[[93m14:59:01:WARNING:WU00:FS00:Past final deadline 2016-11-03T15:17:58Z, dumping^[[0m
14:59:01:WU00:FS00:Cleaning up
14:59:01:WU00:FS00:Connecting to 171.67.108.45:8080
14:59:02:WU00:FS00:Assigned to work server 171.64.65.41
14:59:02:WU00:FS00:Requesting new work unit for slot 00: READY cpu:48 from 171.64.65.41
14:59:02:WU00:FS00:Connecting to 171.64.65.41:8080
14:59:06:WU00:FS00:Downloading 21.03MiB
14:59:19:WU00:FS00:Download 0.59%
14:59:26:WU00:FS00:Download 1.19%
14:59:36:WU00:FS00:Download 1.78%
14:59:44:WU00:FS00:Download 2.38%
14:59:57:WU00:FS00:Download 2.68%
15:00:09:WU00:FS00:Download 2.97%
15:00:16:WU00:FS00:Download 3.86%
15:00:26:WU00:FS00:Download 5.65%
15:00:34:WU00:FS00:Download 5.95%
15:00:40:WU00:FS00:Download 8.03%
15:00:46:WU00:FS00:Download 9.51%
15:00:59:WU00:FS00:Download 11.00%
15:01:20:WU00:FS00:Download 11.30%
15:01:26:WU00:FS00:Download 13.67%
15:01:32:WU00:FS00:Download 16.35%
15:01:51:WU00:FS00:Download 17.84%
15:02:15:WU00:FS00:Download 18.13%
15:03:04:WU00:FS00:Download 18.43%
15:03:33:WU00:FS00:Download 18.73%
15:03:45:WU00:FS00:Download 19.02%
15:03:52:WU00:FS00:Download 22.89%
15:03:58:WU00:FS00:Download 26.75%
15:04:11:WU00:FS00:Download 27.05%
15:04:21:WU00:FS00:Download 30.62%
15:04:30:WU00:FS00:Download 33.89%
15:04:41:WU00:FS00:Download 35.67%
15:04:51:WU00:FS00:Download 37.45%
15:04:57:WU00:FS00:Download 39.83%
15:05:13:WU00:FS00:Download 44.89%
15:05:19:WU00:FS00:Download 46.37%
15:05:29:WU00:FS00:Download 47.86%
15:05:38:WU00:FS00:Download 52.91%
15:05:56:WU00:FS00:Download 56.48%
15:06:13:WU00:FS00:Download 56.78%
15:06:22:WU00:FS00:Download 57.37%
15:06:28:WU00:FS00:Download 58.56%
15:06:34:WU00:FS00:Download 64.21%
15:06:46:WU00:FS00:Download 70.75%
15:07:00:WU00:FS00:Download 71.04%
15:07:25:WU00:FS00:Download 71.64%
15:07:31:WU00:FS00:Download 72.83%
15:07:37:WU00:FS00:Download 77.88%
15:07:44:WU00:FS00:Download 78.77%
15:07:54:WU00:FS00:Download 80.56%
15:08:05:WU00:FS00:Download 82.64%
15:08:11:WU00:FS00:Download 83.23%
15:08:19:WU00:FS00:Download 87.39%
15:08:31:WU00:FS00:Download 89.47%
15:08:42:WU00:FS00:Download 91.55%
15:08:48:WU00:FS00:Download 95.72%
15:08:52:WU00:FS00:Download complete
15:08:52:WU00:FS00:Received Unit: id:00 state:DOWNLOAD error:NO_ERROR project:11920 run:987 clone:0 gen:90 core:0xa7 unit:0x00000087ab40412957bff59ef84afd10
15:08:52:WU00:FS00:Starting
15:08:52:WU00:FS00:Running FahCore: /opt/glibc-2.17/lib/FAHCoreWrapper /root/Folding/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a7.fah/FahCore_a7 -dir 00 -suffix 01 -version 704 -lifeline 17058 -checkpoint 15 -np 48
15:08:52:WU00:FS00:Started FahCore on PID 17101
15:08:52:WU00:FS00:FahCore 0xa7 started
^[[93m15:08:52:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)^[[0m
15:08:53:WU00:FS00:Starting
15:08:53:WU00:FS00:Running FahCore: /opt/glibc-2.17/lib/FAHCoreWrapper /root/Folding/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a7.fah/FahCore_a7 -dir 00 -suffix 01 -version 704 -lifeline 17058 -checkpoint 15 -np 48
15:08:53:WU00:FS00:Started FahCore on PID 17103
15:08:53:WU00:FS00:FahCore 0xa7 started
^[[93m15:08:53:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)^[[0m
15:09:53:WU00:FS00:Starting
15:09:53:WU00:FS00:Running FahCore: /opt/glibc-2.17/lib/FAHCoreWrapper /root/Folding/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a7.fah/FahCore_a7 -dir 00 -suffix 01 -version 704 -lifeline 17058 -checkpoint 15 -np 48
15:09:53:WU00:FS00:Started FahCore on PID 17106
15:09:53:WU00:FS00:FahCore 0xa7 started
^[[93m15:09:53:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)^[[0m
15:10:53:WU00:FS00:Starting
15:10:53:WU00:FS00:Running FahCore: /opt/glibc-2.17/lib/FAHCoreWrapper /root/Folding/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a7.fah/FahCore_a7 -dir 00 -suffix 01 -version 704 -lifeline 17058 -checkpoint 15 -np 48
15:10:53:WU00:FS00:Started FahCore on PID 17114
15:10:53:WU00:FS00:FahCore 0xa7 started
^[[93m15:10:53:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)^[[0m
15:11:53:WU00:FS00:Starting
15:11:53:WU00:FS00:Running FahCore: /opt/glibc-2.17/lib/FAHCoreWrapper /root/Folding/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a7.fah/FahCore_a7 -dir 00 -suffix 01 -version 704 -lifeline 17058 -checkpoint 15 -np 48
15:11:53:WU00:FS00:Started FahCore on PID 17118
15:11:53:WU00:FS00:FahCore 0xa7 started
^[[93m15:11:53:WARNING:WU00:FS00:FahCore returned: FAILED_3 (255 = 0xff)^[[0m
^[[93m15:11:53:WARNING:WU00:FS00:Too many errors, failing^[[0m
15:11:53:WU00:FS00:Sending unit results: id:00 state:SEND error:FAILED project:11920 run:987 clone:0 gen:90 core:0xa7 unit:0x00000087ab40412957bff59ef84afd10

Preliminary searches on this error suggest it's a 32-bit vs 64-bit thing. But I've not yet had time to poke at this much.
It could also be because of the location of the FAHCoreWrapper, which is certainly NOT "/opt/glibc-2.17/lib/FAHCoreWrapper".

However, when creating a script there that calls the wrapper:

Code: Select all

/opt/glibc-2.17/lib/ld-linux-x86-64.so.2 --library-path /opt/glibc-2.17/lib:/opt/glibc-2.17/lib64/:/lib:/lib64:/usr/lib64 /usr/bin/FAHCoreWrapper

the same Failed 3 error occurs.

If creating just a symbolic link, the following error happens: 00:17:40:WARNING:WU00:FS00:FahCore returned: UNKNOWN_ENUM (127 = 0x7f) But this is most likely because of not calling it through the new ld.

So this is where I'm at and I'm not sure how to proceed. I'm hoping for some insights. I'd love to be able to get the A7 cores running on CentOS 6.8 (a system built on glibc-2.14) as many of my systems run 6.8. Upgrading to 7.2 is not impossible but not trivial either, given the other tasks they are serving. And I know many other users would love to find a solution as well.

Thanks for all your guy's help on this.

-Kirk

[bruce]

I'm not sure if it's a 32-/64-bit issue, but if it is, perhaps that's because of your OS.

FAH assumes that all Linux installations are 64-bit Linux with the added 32-bit libraries to support backwards capabilities. Running a 32-bit Linux may be possible with a lot of work. Running a 64-bit Linux without the 32-bit libraries is asking for trouble, too.

[[Ars] For Caitlin]

I'm also unable to unravel the a7 library requirements. The only thing I can think to do at this point is blackhole the a7 work servers.

[toTOW]

If you try to start the core manually from command line, it will print which libraries are required and which ones are found on your system ...

[[Ars] For Caitlin]

Yes, I've done that, its just too difficult to munge the libraries on a 6.9 system that I plan on upgrading to 7 anyway. In the meantime, its easy to look at the project summary page and blackhole the a7 work servers.

[Joe_H]

That may work in the short term, but it will also blackhole some of the A4 work servers. A4 and A7 projects are assigned by the same servers in several cases. In the long run as A4 projects complete, new ones will be coming out using the A7 core as it is more efficient. At this point the only new A4 projects are continuations of older ones that were started using the A4 core.

Another side effect if you are also doing GPU folding would be blocking servers that also have A7 work being assigned as well.