FAH 硬體及客戶端綜合討論 (part 1)

[minihi001]

不會麻煩拉 ~ :D

下一次就輪到您囉, salunt 大 :MMM:

現在把cpu*3個都停掉了…
期待smp教學:PPP:
看來明天該ghost還原一下了…

 

[ssss8kimo]

現在跑CPU + GPU 運算，GPU部分是沒什麼影響
CPU設下載小於10MB的來運算，大約1d10h完成一個，只加了225point ;oq;
之前跑smp，3~4天就+了19xx ;face12;
差好多:PPP:

 

[84036980]

請問monkp兄有沒有遇過SMP跑完卻回傳失敗~換了6.22之後跑了2台都一樣~
真的是冏到不行~

 

[monkp]

SMP 實用的小工具 - qfix

qfix 補充

執行過後, 應該可以繼續跑 SMP 運算, 由上一次的良好的備份. :)

 

[monkp]

請問monkp兄有沒有遇過SMP跑完卻回傳失敗~換了6.22之後跑了2台都一樣~
真的是冏到不行~

是回傳失敗還是工作單位損毀 ? 如果失敗的話就要繼續等次服器恢復了, 這是沒辦法的, 不過我很少遇過 SMP 的 result server 會罷工, GPU 的倒是常常發生

如果是損毀的話, 請用 qfix 來繼續完成它

我也有好幾次遇上工作單位損壞的經驗, 那時真的很心痛說.

MPICH 這版的 SMP 比較有這樣的狀況, Deino 相對的比較少, 有比較穩定, 由官方的說明未來將一律改由 Deino 版本, 同時也簡化設定上的手續 ... :)

 

[84036980]

是回傳失敗還是工作單位損毀 ? 如果失敗的話就要繼續等次服器恢復了, 這是沒辦法的, 不過我很少遇過 SMP 的 result server 會罷工, GPU 的倒是常常發生

如果是損毀的話, 請用 qfix 來繼續完成它

我也有好幾次遇上工作單位損壞的經驗, 那時真的很心痛說.

MPICH 這版的 SMP 比較有這樣的狀況, Deino 相對的比較少, 有比較穩定, 由官方的說明未來將一律改由 Deino 版本, 同時也簡化設定上的手續 ... :)...

因該是回傳失敗~我看她已顯示完成~但最後出錯~
再來我把它關閉~加了-send all參數~他一秒就關了~不知道有沒有傳....:PPP:
現在把-send all參數拿掉~又繼續抓新的跑了~

# Windows SMP Console Edition #################################################
###############################################################################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: D:\Folding@Home Windows SMP Client V1.01
Executable: D:\Folding@Home Windows SMP Client V1.01\Folding@home-Win32-x86.exe
Arguments: -smp

[16:50:11] - Ask before connecting: No
[16:50:11] - User name: 84036980 (Team 133977)
[16:50:11] - User ID: 68CBEA1E1725841E
[16:50:11] - Machine ID: 1
[16:50:11]
[16:50:12] Loaded queue successfully.
[16:50:12]
[16:50:12] + Processing work unit
[16:50:12] Work type a1 not eligible for variable processors
[16:50:12] Core required: FahCore_a1.exe
[16:50:12] Core found.
[16:50:12] Using generic mpiexec calls
[16:50:12] Working on queue slot 01 [August 6 16:50:12 UTC]
[16:50:12] + Working ...
[16:50:12]
[16:50:12] *------------------------------*
[16:50:12] Folding@Home Gromacs SMP Core
[16:50:12] Version 1.74 (March 10, 2007)
[16:50:12]
[16:50:12] Preparing to commence simulation
[16:50:12] - Ensuring status. Please wait.
[16:50:12] - Looking at optimizations...
[16:50:12] - Working with standard loops on this execution.
[16:50:12] Examination of work files indicates 8 consecutive improper terminations of core.
[16:50:12] - Expanded 4798864 -> 24810145 (decompressed 517.0 percent)
[16:50:12]
[16:50:12] Project: 2665 (Run 2, Clone 630, Gen 38)
[16:50:12]
[16:50:12] Entering M.D.
[16:50:12] Protein: HGG with glycosylations
[16:50:12] Writing local files
[16:50:12] Completed 173650 out of 250000 steps (69 percent)
[16:50:12] Extra SSE boost OK.
[16:50:12] Past main M.D. loop
[16:50:12]
[16:50:12] Finished Work Unit:
[16:50:12] Leaving Run
[16:50:12] - Shutting down core
[16:50:12]
[16:50:12] Folding@home Core Shutdown: FINISHED_UNIT
[16:50:12] Finalizing output
[16:50:12] percent)
[16:50:12] Writing local files
[16:50:12] Completed 177500 out of 250000 steps (71 percent)
[16:50:12] Writing local files
[16:50:12] Completed 180000 out of 250000 steps (72 percent)
[16:50:12] Writing local files
[16:50:12] Completed 182500 out of 250000 steps (73 percent)
[16:50:12] Writing local files
[16:50:12] Completed 185000 out of 250000 steps (74 percent)
[16:50:12] Writing local files
[16:50:12] Completed 187500 out of 250000 steps (75 percent)
[16:50:12] Writing local files
[16:50:12] Completed 190000 out of 250000 steps (76 percent)
[16:50:12] Writing local files
[16:50:12] Completed 192500 out of 250000 steps (77 percent)
[16:50:12] Writing local files
[16:50:12] Completed 195000 out of 250000 steps (78 percent)
[16:50:12] Writing local files
[16:50:12] Completed 197500 out of 250000 steps (79 percent)
[16:50:12] Writing local files
[16:50:12] Completed 200000 out of 250000 steps (80 percent)
[16:50:12] Writing local files
[16:50:12] Completed 202500 out of 250000 steps (81 percent)
[16:50:12] Writing local files
[16:50:12] Completed 205000 out of 250000 steps (82 percent)
[16:50:12] Writing local files
[16:50:12] Completed 207500 out of 250000 steps (83 percent)
[16:50:12] Writing local files
[16:50:12] Completed 210000 out of 250000 steps (84 percent)
[16:50:12] Writing local files
[16:50:12] Completed 212500 out of 250000 steps (85 percent)
[16:50:12] Writing local files
[16:50:12] Completed 215000 out of 250000 steps (86 percent)
[16:50:12] Writing local files
[16:50:12] Completed 217500 out of 250000 steps (87 percent)
[16:50:12] Writing local files
[16:50:12] Completed 220000 out of 250000 steps (88 percent)
[16:50:12] Writing local files
[16:50:12] Completed 222500 out of 250000 steps (89 percent)
[16:50:12] Writing local files
[16:50:12] Completed 225000 out of 250000 steps (90 percent)
[16:50:12] Writing local files
[16:50:12] Completed 227500 out of 250000 steps (91 percent)
[16:50:12] Writing local files
[16:50:12] Completed 230000 out of 250000 steps (92 percent)
[16:50:12] Writing local files
[16:50:12] Completed 232500 out of 250000 steps (93 percent)
[16:50:12] Writing local files
[16:50:12] Completed 235000 out of 250000 steps (94 percent)
[16:50:12] Writing local files
[16:50:12] Completed 237500 out of 250000 steps (95 percent)
[16:50:12] Writing local files
[16:50:12] Completed 240000 out of 250000 steps (96 percent)
[16:50:12] Writing local files
[16:50:12] Completed 242500 out of 250000 steps (97 percent)
[16:50:12] Writing local files
[16:50:12] Completed 245000 out of 250000 steps (98 percent)
[16:50:12] Writing local files
[16:50:12] Completed 247500 out of 250000 steps (99 percent)
[16:50:12] Writing local files
[16:50:12] Completed 250000 out of 250000 steps (100 percent)
[16:50:12] Writing final coordinates.
[16:50:12] Past main M.D. loop
[16:50:12] Will end MPI now
[16:50:12]
[16:50:12] Finished Work Unit:
[16:50:12] - Reading up to 21421872 from "work/wudata_01.arc": Read 21421872
[16:50:12] - Reading up to 594256 from "work/wudata_01.xtc": Read 594256
[16:50:12] goefile size: 0
[16:50:12] logfile size: 221579
[16:50:12] Leaving Run
[16:50:12] - Writing 22244079 bytes of core data to disk...
[16:50:12] ... Done.
[16:50:12] - Failed to delete work/wudata_01.sas
[16:50:12] - Failed to delete work/wudata_01.goe
[16:50:12] Warning: check for stray files
[16:50:12] - Shutting down core
[16:50:12]
[16:50:12] Folding@home Core Shutdown: FINISHED_UNIT
[16:50:12]
[16:50:12] Folding@home Core Shutdown: FINISHED_UNIT
[16:52:32] CoreStatus = 1 (1)
[16:52:32] Client-core communications error: ERROR 0x1
[16:52:32] This is a sign of more serious problems, shutting down.


--- Opening Log file [August 6 16:52:56 UTC]


# Windows SMP Console Edition #################################################
###############################################################################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: D:\Folding@Home Windows SMP Client V1.01
Executable: D:\Folding@Home Windows SMP Client V1.01\Folding@home-Win32-x86.exe
Arguments: -smp

[16:52:56] - Ask before connecting: No
[16:52:56] - User name: 84036980 (Team 133977)
[16:52:56] - User ID: 68CBEA1E1725841E
[16:52:56] - Machine ID: 1
[16:52:56]
[16:52:56] Loaded queue successfully.
[16:52:56]
[16:52:56] + Processing work unit
[16:52:56] Work type a1 not eligible for variable processors
[16:52:56] Core required: FahCore_a1.exe
[16:52:56] Core found.
[16:52:56] Using generic mpiexec calls
[16:52:56] Working on queue slot 01 [August 6 16:52:56 UTC]
[16:52:56] + Working ...
[16:52:56]
[16:52:56] *------------------------------*
[16:52:56] Folding@Home Gromacs SMP Core
[16:52:56] Version 1.74 (March 10, 2007)
[16:52:56]
[16:52:56] Preparing to commence simulation
[16:52:56] - Ensuring status. Please wait.
[16:52:56] - Looking at optimizations...
[16:52:56] - Working with standard loops on this execution.
[16:52:56] - Previous termination of core was improper.
[16:52:56] - Going to use sta
[16:52:56] Folding@home Core Shutdown: MI
[16:52:56] Folding@home Core Shutdown: MISSING_WORK_FILES
[16:52:56] Finalizing output
[16:55:16] CoreStatus = 1 (1)
[16:55:16] Client-core communications error: ERROR 0x1
[16:55:16] This is a sign of more serious problems, shutting down.


--- Opening Log file [August 6 16:56:10 UTC]


# Windows SMP Console Edition #################################################


###############################################################################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: D:\Folding@Home Windows SMP Client V1.01
Executable: D:\Folding@Home Windows SMP Client V1.01\Folding@home-Win32-x86.exe
Arguments: -smp

[16:56:10] - Ask before connecting: No
[16:56:10] - User name: 84036980 (Team 133977)
[16:56:10] - User ID: 68CBEA1E1725841E
[16:56:10] - Machine ID: 1
[16:56:10]
[16:56:10] Could not open work queue, generating new queue...
[16:56:10] - Preparing to get new work unit...
[16:56:10] + Attempting to get work packet
[16:56:10] - Connecting to assignment server
[16:56:11] - Successful: assigned to (171.64.65.64).
[16:56:11] + News From Folding@Home: Welcome to Folding@Home
[16:56:11] Loaded queue successfully.
[16:56:23] + Closed connections
[16:56:24]
[16:56:24] + Processing work unit
[16:56:24] Work type a1 not eligible for variable processors
[16:56:24] Core required: FahCore_a1.exe
[16:56:24] Core found.
[16:56:24] Using generic mpiexec calls
[16:56:24] Working on queue slot 01 [August 6 16:56:24 UTC]
[16:56:24] + Working ...
[16:56:24]
[16:56:24] *------------------------------*
[16:56:24] Folding@Home Gromacs SMP Core
[16:56:24] Version 1.74 (March 10, 2007)
[16:56:24]
[16:56:24] Preparing to commence simulation
[16:56:24] - Ensuring status. Please wait.
[16:56:24] - Looking at optimizations...
[16:56:24] - Working with standard loops on this execution.
[16:56:24] - Previous termination of core was improper.
[16:56:24] - Going to use standard loops.
[16:56:24] - Files status OK
[16:56:24]
[16:56:24] Folding@home Core Shutdown: MISSING_WORK_FILES
[16:56:24] Finalizing output

Folding@Home Client Shutdown at user request.

Folding@Home Client Shutdown.


--- Opening Log file [August 7 00:40:38 UTC]


# Windows SMP Console Edition #################################################
###############################################################################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: D:\Folding@Home Windows SMP Client V1.01
Executable: D:\Folding@Home Windows SMP Client V1.01\Folding@home-Win32-x86.exe
Arguments: -send all -smp

[00:40:38] - Ask before connecting: No
[00:40:38] - User name: 84036980 (Team 133977)
[00:40:38] - User ID: 68CBEA1E1725841E
[00:40:38] - Machine ID: 1
[00:40:38]
[00:40:38] Loaded queue successfully.
[00:40:38] Attempting to return result(s) to server...

Folding@Home Client Shutdown.


--- Opening Log file [August 7 00:42:56 UTC]


# Windows SMP Console Edition #################################################
###############################################################################

Folding@Home Client Version 6.22 SMP Beta2

http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: D:\Folding@Home Windows SMP Client V1.01
Executable: D:\Folding@Home Windows SMP Client V1.01\Folding@home-Win32-x86.exe
Arguments: -smp

[00:42:56] - Ask before connecting: No
[00:42:56] - User name: 84036980 (Team 133977)
[00:42:56] - User ID: 68CBEA1E1725841E
[00:42:56] - Machine ID: 1
[00:42:56]
[00:42:57] Loaded queue successfully.
[00:42:57]
[00:42:57] + Processing work unit
[00:42:57] Work type a1 not eligible for variable processors
[00:42:57] Core required: FahCore_a1.exe
[00:42:57] Core found.
[00:42:57] Using generic mpiexec calls
[00:42:57] Working on queue slot 01 [August 7 00:42:57 UTC]
[00:42:57] + Working ...
[00:42:57]
[00:42:57] *------------------------------*
[00:42:57] Folding@Home Gromacs SMP Core
[00:42:57] Version 1.74 (March 10, 2007)
[00:42:57]
[00:42:57] Preparing to commence simulation
[00:42:57] - Ensuring status. Please wait.
[00:43:14] - Looking at optimizations...
[00:43:14] - Working with standard loops on this execution.
[00:43:14] Examination of work files indicates 8 consecutive improper terminations of core.
[00:43:33] - Expanded 4746784 -> 24426905 (decompressed 514.5 percent)
[00:43:34] - Starting from initial work packet
[00:43:34]
[00:43:34] Project: 2665 (Run 3, Clone 227, Gen 39)
[00:43:34]
[00:43:39] Entering M.D.
[00:43:47] Rejecting checkpoint
[00:43:50] Protein: HGG in water
[00:43:50] Writing local files
[00:44:05] Extra SSE boost OK.
[00:44:07] Writing local files
[00:44:07] Completed 0 out of 250000 steps (0 percent)

 

[84036980]

話說回來~有人試過一個平台同時插ATI 與nvidia的卡嗎?
昨天試了半天搞不出來雙GPU......又是冏rz....

 

[monkp]

維基裡的解釋 - Error 0x0 and 0x1

通常和系統的穩定性有關聯, 尤其是記憶體, 假如有超頻的話

大大的例子需要 qfix

log 裡面沒有 send successful 的訊息耶 ... 好像沒有成功的樣子

對了 ... 加上 -verbosity 9 參數回報的訊息會比較完整點 ...

 

[ssss8kimo]

話說回來~有人試過一個平台同時插ATI 與nvidia的卡嗎?
昨天試了半天搞不出來雙GPU......又是冏rz....

應該不可以這樣裝吧;rr;
光驅動就不知道怎麼裝了

會衝突

 

[CrazyDiamond]

話說回來~有人試過一個平台同時插ATI 與nvidia的卡嗎?
昨天試了半天搞不出來雙GPU......又是冏rz....

XP OK(看過老外的post但我沒試過), Vista就不行(系統限單顯卡趨動) :lkl: