The underwelming response to my question leaves me wondering:
Am I to gather that there are no CentOSFAH users here?
or
Am I to understand that none of the CentOS FAH users have a GPU that works with FAH?
5 thoughts on - Folding At Home OT
Yes, I’m a CentOS Folder. I’m one of 3 or 4 currently active on Team CentOS
However, I’ve been unable to get GPU folding working, so I just dedicate all six cores to it.
You may want to ask your original question on the forums at the FAH site, also.
Fred
+1
Another long time folder, I don’t however do GPU folding just run it on my C7 24 core monster :)
Hey Fred,
Thanks for the tip. I googled the FAH Hardware Forum. They recommend the NVIDIA 970 or 960 as their first choice.
True. but it “should” work on Fermi chipsets or later, and it doesn’t for me. I may well have a configuration problem, but I drew the line at butchering my system configuration to find out, this is my personal workstation, at home, and I don’t wanna break it.
(it tries, and tries, and tries to fold but takes some error or other, backs off, tries again, and again, and again.)
Hello all,
I tried also to run a GPU queue with FAH and it seems that the client program needs a newer libc library than available in C6. The logs were not helpfull:
5 thoughts on - Folding At Home OT
Yes, I’m a CentOS Folder. I’m one of 3 or 4 currently active on Team CentOS
However, I’ve been unable to get GPU folding working, so I just dedicate all six cores to it.
You may want to ask your original question on the forums at the FAH site, also.
Fred
+1
Another long time folder, I don’t however do GPU folding just run it on my C7 24 core monster :)
Hey Fred,
Thanks for the tip. I googled the FAH Hardware Forum. They recommend the NVIDIA 970 or 960 as their first choice.
True. but it “should” work on Fermi chipsets or later, and it doesn’t for me. I may well have a configuration problem, but I drew the line at butchering my system configuration to find out, this is my personal workstation, at home, and I don’t wanna break it.
(it tries, and tries, and tries to fold but takes some error or other, backs off, tries again, and again, and again.)
Hello all,
I tried also to run a GPU queue with FAH and it seems that the client program needs a newer libc library than available in C6. The logs were not helpfull:
09:33:52:WU00:FS01:Running FahCore: /usr/bin/FAHCoreWrapper
/var/lib/fahclient/cores/web.stanford.edu/~pande/Linux/AMD64
/NVIDIA/Fermi/Core_17.fah/FahCore_17 -dir 00 -suffix 01 -version 704 -lifeline 13249
-checkpoint 15 -gpu 0 -gpu-vendor nvidia
09:33:52:WU00:FS01:Started FahCore on PID 13558
09:33:52:WU00:FS01:Core PID:13562
09:33:52:WU00:FS01:FahCore 0x17 started
09:33:53:WARNING:WU00:FS01:FahCore returned: FAILED_2 (1 = 0x1)
I have discovered the real error only after manually running the command listed in logs with strace:
[pid 13645] writev(2, [{“/var/lib/fahclient/cores/web.sta”…, 96}, {“: “, 2},
{“/lib64/libm.so.6”, 16}, {“: “, 2}, {“version `GLIBC_2.15′ not found (“…, 141}, {“\n”,
1}], 6) = 258
[pid 13645] writev(2, [{“/var/lib/fahclient/cores/web.sta”…, 96}, {“: “, 2},
{“/usr/lib64/libstdc++.so.6”, 25}, {“: “, 2}, {“version `GLIBCXX_3.4.15′ not fou”…,
145}, {“\n”, 1}], 6) = 271
[pid 13645] writev(2, [{“/var/lib/fahclient/cores/web.sta”…, 96}, {“: “, 2},
{“/lib64/libc.so.6”, 16}, {“: “, 2}, {“version `GLIBC_2.15′ not found (“…, 141}, {“\n”,
1}], 6) = 258
[pid 13645] writev(2, [{“/var/lib/fahclient/cores/web.sta”…, 96}, {“: “, 2},
{“/lib64/libc.so.6”, 16}, {“: “, 2}, {“version `GLIBC_2.14′ not found (“…, 141}, {“\n”,
1}], 6) = 258
I have to investigate if there is a alternate fahclient (older version maybe) which works in C6
Lec
PS: I have seen that the FAH executable for cpu queue is statically linked:
file /var/lib/fahclient/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a4.fah/FahCore_a4
/var/lib/fahclient/cores/web.stanford.edu/~pande/Linux/AMD64/Core_a4.fah/FahCore_a4: ELF
64-bit LSB executable, x86-64, version 1 (SYSV), statically linked, for GNU/Linux 2.6.15, stripped