TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Ok so I'm running native Linux on one of my rigs folding BigAdv WU's. It gets 100% done & looks to hang at this spot, I waited a long time before Ctrl+C. I did a quick back up before restarting, once I restart it starts out back at 0% on the same WU. Project: 8101 (Run 18, Clone 1, Gen 71) have I lost 49.6 hours of work? qfix didn't do anything to help. Help me out here? [01:11:37] Completed 250000 out of 250000 steps (100%)
[01:11:49] DynamicWrapper: Finished Work Unit: sleep=10000
[01:11:59]
[01:11:59] Finished Work Unit:
[01:11:59] - Reading up to 64340496 from "work/wudata_02.trr": Read 64340496
[01:11:59] trr file hash check passed.
[01:11:59] - Reading up to 31677056 from "work/wudata_02.xtc": Read 31677056
[01:11:59] xtc file hash check passed.
[01:11:59] edr file hash check passed.
[01:11:59] logfile size: 221087
[01:11:59] Leaving Run
[01:12:03] - Writing 96399515 bytes of core data to disk...
[01:12:22] Done: 96399003 -> 91639782 (compressed to 5.9 percent)
[01:12:22] ... Done.
post edited by TheWolf - 2013/01/22 22:18:29
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/11/30 18:44:18
(permalink)
Here is the complete log, pay no attention to the -betatem flag used later on, as I just wanted to try a different type Wu next to see how it would work on this machine.
--- Opening Log file [November 28 22:06:26 UTC]
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /home/thewolf/fah
Executable: ./fah6
Arguments: -bigadv -smp 24 -verbosity 9
[22:06:26] - Ask before connecting: No
[22:06:26] - User name: TheWolf (Team 111065)
[22:06:26] - User ID: 7161C4375E45BADA
[22:06:26] - Machine ID: 1
[22:06:26]
[22:06:26] Loaded queue successfully.
[22:06:26]
[22:06:26] + Processing work unit
[22:06:26] - Autosending finished units... [November 28 22:06:26 UTC]
[22:06:26] Trying to send all finished work units
[22:06:26] Core required: FahCore_a5.exe
[22:06:26] + No unsent completed units remaining.
[22:06:26] Core found.
[22:06:26] - Autosend completed
[22:06:26] Working on queue slot 02 [November 28 22:06:26 UTC]
[22:06:26] + Working ...
[22:06:26] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 02 -np 24 -checkpoint 6 -verbose -lifeline 3557 -version 634'
[22:06:26]
[22:06:26] *------------------------------*
[22:06:26] Folding@Home Gromacs SMP Core
[22:06:26] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[22:06:26]
[22:06:26] Preparing to commence simulation
[22:06:26] - Looking at optimizations...
[22:06:26] - Files status OK
[22:06:29] - Expanded 30307691 -> 33158020 (decompressed 109.4 percent)
[22:06:29] Called DecompressByteArray: compressed_data_size=30307691 data_size=33158020, decompressed_data_size=33158020 diff=0
[22:06:29] - Digital signature verified
[22:06:29]
[22:06:29] Project: 8101 (Run 18, Clone 1, Gen 71)
[22:06:29]
[22:06:29] Assembly optimizations on if available.
[22:06:29] Entering M.D.
[22:06:36] Mapping NT from 24 to 24
[22:06:47] Completed 0 out of 250000 steps (0%)
[22:37:50] Completed 2500 out of 250000 steps (1%)
[23:06:56] Completed 5000 out of 250000 steps (2%)
[23:35:59] Completed 7500 out of 250000 steps (3%)
[00:10:03] Completed 10000 out of 250000 steps (4%)
[00:44:26] ***** Got an Activate signal (2)
[00:44:26] Killing all core threads
Folding@Home Client Shutdown.
--- Opening Log file [November 29 00:53:21 UTC]
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /home/thewolf/fah
Executable: ./fah6
Arguments: -betateam -smp 24 -verbosity 9
[00:53:21] - Ask before connecting: No
[00:53:21] - User name: TheWolf (Team 111065)
[00:53:21] - User ID: 7161C4375E45BADA
[00:53:21] - Machine ID: 1
[00:53:21]
[00:53:21] Loaded queue successfully.
[00:53:21]
[00:53:21] + Processing work unit
[00:53:21] Core required: FahCore_a5.exe
[00:53:21] Core found.
[00:53:21] - Autosending finished units... [00:53:21]
[00:53:21] Trying to send all finished work units
[00:53:21] + No unsent completed units remaining.
[00:53:21] - Autosend completed
[00:53:21] Working on queue slot 02 [November 29 00:53:21 UTC]
[00:53:21] + Working ...
[00:53:21] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 02 -np 24 -checkpoint 6 -verbose -lifeline 3011 -version 634'
[00:53:21]
[00:53:21] *------------------------------*
[00:53:21] Folding@Home Gromacs SMP Core
[00:53:21] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[00:53:21]
[00:53:21] Preparing to commence simulation
[00:53:21] - Looking at optimizations...
[00:53:21] - Files status OK
[00:53:24] - Expanded 30307691 -> 33158020 (decompressed 109.4 percent)
[00:53:24] Called DecompressByteArray: compressed_data_size=30307691 data_size=33158020, decompressed_data_size=33158020 diff=0
[00:53:24] - Digital signature verified
[00:53:24]
[00:53:24] Project: 8101 (Run 18, Clone 1, Gen 71)
[00:53:24]
[00:53:25] Assembly optimizations on if available.
[00:53:25] Entering M.D.
[00:53:31] Using Gromacs checkpoints
[00:53:33] Mapping NT from 24 to 24
[00:59:29] Resuming from checkpoint
[00:59:30] Verified work/wudata_02.log
[00:59:38] Verified work/wudata_02.trr
[00:59:39] Verified work/wudata_02.xtc
[00:59:39] Verified work/wudata_02.edr
[00:59:40] Completed 12085 out of 250000 steps (4%)
[01:08:23] Completed 12500 out of 250000 steps (5%)
[01:45:00] Completed 15000 out of 250000 steps (6%)
[02:15:40] Completed 17500 out of 250000 steps (7%)
[02:46:20] Completed 20000 out of 250000 steps (8%)
[03:17:00] Completed 22500 out of 250000 steps (9%)
[03:47:41] Completed 25000 out of 250000 steps (10%)
[03:53:47] ***** Got an Activate signal (2)
[03:53:47] Killing all core threads
Folding@Home Client Shutdown.
--- Opening Log file [November 29 03:57:42 UTC]
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /home/thewolf/fah
Executable: ./fah6
Arguments: -betateam -smp 24 -verbosity 9
[03:57:42] - Ask before connecting: No
[03:57:42] - User name: TheWolf (Team 111065)
[03:57:42] - User ID: 7161C4375E45BADA
[03:57:42] - Machine ID: 1
[03:57:42]
[03:57:42] Loaded queue successfully.
[03:57:42]
[03:57:42] - Autosending finished units... [November 29 03:57:42 UTC]
[03:57:42] + Processing work unit
[03:57:42] Trying to send all finished work units
[03:57:42] Core required: FahCore_a5.exe
[03:57:42] + No unsent completed units remaining.
[03:57:42] Core found.
[03:57:42] - Autosend completed
[03:57:42] Working on queue slot 02 [November 29 03:57:42 UTC]
[03:57:42] + Working ...
[03:57:42] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 02 -np 24 -checkpoint 6 -verbose -lifeline 2687 -version 634'
[03:57:42]
[03:57:42] *------------------------------*
[03:57:42] Folding@Home Gromacs SMP Core
[03:57:42] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[03:57:42]
[03:57:42] Preparing to commence simulation
[03:57:42] - Looking at optimizations...
[03:57:42] - Files status OK
[03:57:45] - Expanded 30307691 -> 33158020 (decompressed 109.4 percent)
[03:57:45] Called DecompressByteArray: compressed_data_size=30307691 data_size=33158020, decompressed_data_size=33158020 diff=0
[03:57:46] - Digital signature verified
[03:57:46]
[03:57:46] Project: 8101 (Run 18, Clone 1, Gen 71)
[03:57:46]
[03:57:46] Assembly optimizations on if available.
[03:57:46] Entering M.D.
[03:57:52] Using Gromacs checkpoints
[03:57:54] Mapping NT from 24 to 24
[03:58:27] Resuming from checkpoint
[03:58:28] Verified work/wudata_02.log
[03:58:31] Verified work/wudata_02.trr
[03:58:32] Verified work/wudata_02.xtc
[03:58:32] Verified work/wudata_02.edr
[03:58:32] Completed 25480 out of 250000 steps (10%)
[04:32:56] Completed 27500 out of 250000 steps (11%)
[04:51:43] ***** Got an Activate signal (2)
[04:51:43] Killing all core threads
Folding@Home Client Shutdown.
--- Opening Log file [November 29 05:11:37 UTC]
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /home/thewolf/fah
Executable: ./fah6
Arguments: -betateam -smp 24 -verbosity 9
[05:11:37] - Ask before connecting: No
[05:11:37] - User name: TheWolf (Team 111065)
[05:11:37] - User ID: 7161C4375E45BADA
[05:11:37] - Machine ID: 1
[05:11:37]
[05:11:37] Loaded queue successfully.
[05:11:37]
[05:11:37] + Processing work unit
[05:11:37] Core required: FahCore_a5.exe
[05:11:37] - Autosending finished units... [05:11:37]
[05:11:37] Core found.
[05:11:37] Trying to send all finished work units
[05:11:37] + No unsent completed units remaining.
[05:11:37] - Autosend completed
[05:11:38] Working on queue slot 02 [November 29 05:11:38 UTC]
[05:11:38] + Working ...
[05:11:38] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 02 -np 24 -checkpoint 6 -verbose -lifeline 2694 -version 634'
[05:11:38]
[05:11:38] *------------------------------*
[05:11:38] Folding@Home Gromacs SMP Core
[05:11:38] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[05:11:38]
[05:11:38] Preparing to commence simulation
[05:11:38] - Looking at optimizations...
[05:11:38] - Files status OK
[05:11:41] - Expanded 30307691 -> 33158020 (decompressed 109.4 percent)
[05:11:41] Called DecompressByteArray: compressed_data_size=30307691 data_size=33158020, decompressed_data_size=33158020 diff=0
[05:11:41] - Digital signature verified
[05:11:41]
[05:11:41] Project: 8101 (Run 18, Clone 1, Gen 71)
[05:11:41]
[05:11:41] Assembly optimizations on if available.
[05:11:41] Entering M.D.
[05:11:47] Using Gromacs checkpoints
[05:11:50] Mapping NT from 24 to 24
[05:12:23] Resuming from checkpoint
[05:12:26] Verified work/wudata_02.log
[05:12:28] Verified work/wudata_02.trr
[05:12:29] Verified work/wudata_02.xtc
[05:12:29] Verified work/wudata_02.edr
[05:12:30] Completed 28290 out of 250000 steps (11%)
[05:38:22] Completed 30000 out of 250000 steps (12%)
[06:10:37] Completed 32500 out of 250000 steps (13%)
[06:40:05] Completed 35000 out of 250000 steps (14%)
[07:09:32] Completed 37500 out of 250000 steps (15%)
[07:39:00] Completed 40000 out of 250000 steps (16%)
[08:08:28] Completed 42500 out of 250000 steps (17%)
[08:37:55] Completed 45000 out of 250000 steps (18%)
[09:07:22] Completed 47500 out of 250000 steps (19%)
[09:36:49] Completed 50000 out of 250000 steps (20%)
[10:06:12] Completed 52500 out of 250000 steps (21%)
[10:35:39] Completed 55000 out of 250000 steps (22%)
[11:05:05] Completed 57500 out of 250000 steps (23%)
[11:11:37] - Autosending finished units... [November 29 11:11:37 UTC]
[11:11:37] Trying to send all finished work units
[11:11:37] + No unsent completed units remaining.
[11:11:37] - Autosend completed
[11:34:30] Completed 60000 out of 250000 steps (24%)
[12:03:57] Completed 62500 out of 250000 steps (25%)
[12:33:23] Completed 65000 out of 250000 steps (26%)
[13:02:48] Completed 67500 out of 250000 steps (27%)
[13:32:16] Completed 70000 out of 250000 steps (28%)
[14:01:51] Completed 72500 out of 250000 steps (29%)
[14:31:17] Completed 75000 out of 250000 steps (30%)
[15:00:45] Completed 77500 out of 250000 steps (31%)
[15:30:10] Completed 80000 out of 250000 steps (32%)
[15:59:38] Completed 82500 out of 250000 steps (33%)
[16:29:07] Completed 85000 out of 250000 steps (34%)
[16:58:36] Completed 87500 out of 250000 steps (35%)
[17:11:37] - Autosending finished units... [November 29 17:11:37 UTC]
[17:11:37] Trying to send all finished work units
[17:11:37] + No unsent completed units remaining.
[17:11:37] - Autosend completed
[17:28:05] Completed 90000 out of 250000 steps (36%)
[18:03:34] Completed 92500 out of 250000 steps (37%)
[18:35:08] Completed 95000 out of 250000 steps (38%)
[19:05:33] Completed 97500 out of 250000 steps (39%)
[19:35:02] Completed 100000 out of 250000 steps (40%)
[20:04:31] Completed 102500 out of 250000 steps (41%)
[20:33:59] Completed 105000 out of 250000 steps (42%)
[21:03:29] Completed 107500 out of 250000 steps (43%)
[21:32:58] Completed 110000 out of 250000 steps (44%)
[22:02:28] Completed 112500 out of 250000 steps (45%)
[22:31:57] Completed 115000 out of 250000 steps (46%)
[23:01:25] Completed 117500 out of 250000 steps (47%)
[23:11:37] - Autosending finished units... [November 29 23:11:37 UTC]
[23:11:37] Trying to send all finished work units
[23:11:37] + No unsent completed units remaining.
[23:11:37] - Autosend completed
[23:30:54] Completed 120000 out of 250000 steps (48%)
[00:00:21] Completed 122500 out of 250000 steps (49%)
[00:29:49] Completed 125000 out of 250000 steps (50%)
[00:59:17] Completed 127500 out of 250000 steps (51%)
[01:28:47] Completed 130000 out of 250000 steps (52%)
[01:58:17] Completed 132500 out of 250000 steps (53%)
[02:27:48] Completed 135000 out of 250000 steps (54%)
[02:57:18] Completed 137500 out of 250000 steps (55%)
[03:26:47] Completed 140000 out of 250000 steps (56%)
[03:56:16] Completed 142500 out of 250000 steps (57%)
[04:25:44] Completed 145000 out of 250000 steps (58%)
[04:55:12] Completed 147500 out of 250000 steps (59%)
[05:11:37] - Autosending finished units... [November 30 05:11:37 UTC]
[05:11:37] Trying to send all finished work units
[05:11:37] + No unsent completed units remaining.
[05:11:37] - Autosend completed
[05:24:41] Completed 150000 out of 250000 steps (60%)
[05:54:07] Completed 152500 out of 250000 steps (61%)
[06:23:37] Completed 155000 out of 250000 steps (62%)
[06:53:05] Completed 157500 out of 250000 steps (63%)
[07:22:34] Completed 160000 out of 250000 steps (64%)
[07:52:03] Completed 162500 out of 250000 steps (65%)
[08:21:31] Completed 165000 out of 250000 steps (66%)
[08:50:58] Completed 167500 out of 250000 steps (67%)
[09:20:27] Completed 170000 out of 250000 steps (68%)
[09:49:56] Completed 172500 out of 250000 steps (69%)
[10:19:24] Completed 175000 out of 250000 steps (70%)
[10:48:54] Completed 177500 out of 250000 steps (71%)
[11:11:37] - Autosending finished units... [November 30 11:11:37 UTC]
[11:11:37] Trying to send all finished work units
[11:11:37] + No unsent completed units remaining.
[11:11:37] - Autosend completed
[11:18:20] Completed 180000 out of 250000 steps (72%)
[11:47:48] Completed 182500 out of 250000 steps (73%)
[12:17:18] Completed 185000 out of 250000 steps (74%)
[12:46:48] Completed 187500 out of 250000 steps (75%)
[13:16:20] Completed 190000 out of 250000 steps (76%)
[13:45:57] Completed 192500 out of 250000 steps (77%)
[14:15:26] Completed 195000 out of 250000 steps (78%)
[14:44:55] Completed 197500 out of 250000 steps (79%)
[15:14:27] Completed 200000 out of 250000 steps (80%)
[15:44:00] Completed 202500 out of 250000 steps (81%)
[16:13:35] Completed 205000 out of 250000 steps (82%)
[16:43:12] Completed 207500 out of 250000 steps (83%)
[17:11:37] - Autosending finished units... [November 30 17:11:37 UTC]
[17:11:37] Trying to send all finished work units
[17:11:37] + No unsent completed units remaining.
[17:11:37] - Autosend completed
[17:12:49] Completed 210000 out of 250000 steps (84%)
[17:42:24] Completed 212500 out of 250000 steps (85%)
[18:12:03] Completed 215000 out of 250000 steps (86%)
[18:41:43] Completed 217500 out of 250000 steps (87%)
[19:11:24] Completed 220000 out of 250000 steps (88%)
[19:41:05] Completed 222500 out of 250000 steps (89%)
[20:10:47] Completed 225000 out of 250000 steps (90%)
[20:40:28] Completed 227500 out of 250000 steps (91%)
[21:10:11] Completed 230000 out of 250000 steps (92%)
[21:39:54] Completed 232500 out of 250000 steps (93%)
[22:09:35] Completed 235000 out of 250000 steps (94%)
[22:39:15] Completed 237500 out of 250000 steps (95%)
[23:10:45] Completed 240000 out of 250000 steps (96%)
[23:11:37] - Autosending finished units... [November 30 23:11:37 UTC]
[23:11:37] Trying to send all finished work units
[23:11:37] + No unsent completed units remaining.
[23:11:37] - Autosend completed
[23:40:25] Completed 242500 out of 250000 steps (97%)
[00:09:58] Completed 245000 out of 250000 steps (98%)
[00:39:30] Completed 247500 out of 250000 steps (99%)
[01:11:37] Completed 250000 out of 250000 steps (100%)
[01:11:49] DynamicWrapper: Finished Work Unit: sleep=10000
[01:11:59]
[01:11:59] Finished Work Unit:
[01:11:59] - Reading up to 64340496 from "work/wudata_02.trr": Read 64340496
[01:11:59] trr file hash check passed.
[01:11:59] - Reading up to 31677056 from "work/wudata_02.xtc": Read 31677056
[01:11:59] xtc file hash check passed.
[01:11:59] edr file hash check passed.
[01:11:59] logfile size: 221087
[01:11:59] Leaving Run
[01:12:03] - Writing 96399515 bytes of core data to disk...
[01:12:22] Done: 96399003 -> 91639782 (compressed to 5.9 percent)
[01:12:22] ... Done.
--- Opening Log file [December 1 02:17:42 UTC]
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /home/thewolf/fah
Executable: ./fah6
Arguments: -send all -smp 24 -verbosity 9
[02:17:42] - Ask before connecting: No
[02:17:42] - User name: TheWolf (Team 111065)
[02:17:42] - User ID: 7161C4375E45BADA
[02:17:42] - Machine ID: 1
[02:17:42]
[02:17:42] Loaded queue successfully.
[02:17:42] Attempting to return result(s) to server...
[02:17:42] Trying to send all finished work units
[02:17:42] + No unsent completed units remaining.
[02:17:42] ***** Got a SIGTERM signal (15)
[02:17:42] Killing all core threads
Folding@Home Client Shutdown.
--- Opening Log file [December 1 02:18:56 UTC]
# Linux SMP Console Edition ###################################################
###############################################################################
Folding@Home Client Version 6.34
http://folding.stanford.edu
###############################################################################
###############################################################################
Launch directory: /home/thewolf/fah
Executable: ./fah6
Arguments: -betateam -smp 24 -verbosity 9
[02:18:56] - Ask before connecting: No
[02:18:56] - User name: TheWolf (Team 111065)
[02:18:56] - User ID: 7161C4375E45BADA
[02:18:56] - Machine ID: 1
[02:18:56]
[02:18:56] Loaded queue successfully.
[02:18:56]
[02:18:56] + Processing work unit
[02:18:56] Core required: FahCore_a5.exe
[02:18:56] Core found.
[02:18:56] - Autosending finished units... [02:18:56]
[02:18:56] Trying to send all finished work units
[02:18:56] + No unsent completed units remaining.
[02:18:56] - Autosend completed
[02:18:56] Working on queue slot 02 [December 1 02:18:56 UTC]
[02:18:56] + Working ...
[02:18:56] - Calling './FahCore_a5.exe -dir work/ -nice 19 -suffix 02 -np 24 -checkpoint 6 -verbose -lifeline 7715 -version 634'
[02:18:56]
[02:18:56] *------------------------------*
[02:18:56] Folding@Home Gromacs SMP Core
[02:18:56] Version 2.27 (Thu Feb 10 09:46:40 PST 2011)
[02:18:56]
[02:18:56] Preparing to commence simulation
[02:18:56] - Ensuring status. Please wait.
[02:19:06] - Looking at optimizations...
[02:19:06] - Working with standard loops on this execution.
[02:19:06] - Previous termination of core was improper.
[02:19:06] - Files status OK
[02:19:08] - Expanded 30307691 -> 33158020 (decompressed 109.4 percent)
[02:19:08] Called DecompressByteArray: compressed_data_size=30307691 data_size=33158020, decompressed_data_size=33158020 diff=0
[02:19:09] - Digital signature verified
[02:19:09]
[02:19:09] Project: 8101 (Run 18, Clone 1, Gen 71)
[02:19:09]
[02:19:09] Entering M.D.
[02:19:15] Mapping NT from 24 to 24
[02:19:27] Completed 0 out of 250000 steps (0%)
|
Re:Native Linux SMP questions
2012/11/30 18:49:19
(permalink)
How long did it hang at that point? The next line should be "Shutting down core". Then, the unit should upload to the server. So, something is happening right before the core shuts down. I don't know what it could be.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/11/30 18:59:17
(permalink)
I'm not sure maybe 10 mins. This is a fresh install of Linux & 1st run of fah on the new setup. I had this happen once before but in vbox Linux folding, but just let it slide. But I can't keep letting these rigs run for 2+ day and not be able to return the work. The wuresults_02.dat file in the backup folder is 91.6 MB in size. Does this sound about right & could this some how be uploaded to get credit for?
post edited by TheWolf - 2012/11/30 19:02:19
|
Re:Native Linux SMP questions
2012/11/30 19:16:17
(permalink)
Can you try starting the client with the additional flag of -send 02 or -send all through the command line? Not sure if it will do it, but it's worth a try.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/11/30 19:49:50
(permalink)
I tried the -send all but there is not a "I guess" finish log tided to the file, so it doesn't send. Man I was looking forward to that little boost of 220k of points for this WU. I'd go to the ff to seek help, but seems every time I do someone there get a smart a$$ attitude.
post edited by TheWolf - 2012/11/30 19:51:17
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/01 00:35:12
(permalink)
Found the answer here: http://foldingforum.org/v...ic.php?f=44&t=3889 Got the WU uploaded, see newly added notes if your in deep like I was with this one.  From my thread at FF: "TheWolf"Just wanted to let everyone know that the above link I posted did & does still work.
I think it would be nice if a note was added to that link if in a case like mine where
using a back copy to use "./fah6 -send all" on the re-start of folding. This will let it close
after the upload has finished & will not download another work-unit on that start of F@H.
Since I thought it might be lost I did a dump & started clean. Had already started a new WU
I then backup a second time before starting this, of the new WU that was at 25% by now.
So not thinking I just started the old backup with just ./fah6 this in turn started the upload but also downloaded a new WU.
So it would be great if the help topic was edit to reflect to use "./fah6 -send all" on a re-start, just in case you are working in a situation like mine.
Now how to finish two WU's in time with out having to dump one or the other. lol
I'll get this worked out too ;-)
post edited by TheWolf - 2012/12/01 00:50:13
|
Re:Native Linux SMP questions
2012/12/01 04:39:14
(permalink)
Glad to see you got it worked out. And bruce added your comments to the end of that thread like you asked.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/01 06:08:41
(permalink)
There is a newer version of Qfix out that will be needed to fix this problem. You can get more info here and things that might help if you run into this problem.
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Native Linux SMP questions
2012/12/01 06:12:24
(permalink)
Whooo, glad to see that you were able to recover that Bigadv work.  Wonder what caused it in the first place and hope it was just a silicon glitch.
|
Grandpa_01
New Member
- Total Posts : 92
- Reward points : 0
- Joined: 2012/04/28 20:59:00
- Status: offline
- Ribbons : 0
Re:Native Linux SMP questions
2012/12/01 06:54:51
(permalink)
Wolf
I was just curious as to which version of Linux you are running, And did you figure out the reason for the long write time and get it fixed. I did not see where you posted the version of Linux or if that problem had been solved and fixed over at the FF.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/01 18:42:11
(permalink)
I think it was more or less me not waiting long enough for the write to finish.
I'm using the latest Linux Mint 14. Can't say if the long wait is fixed since
I haven't completed another Bigadv since this happened.
I did work a couple WU between these bigadv & they seem to write and send in a timely manner.
File system is ext4.
|
Punchy
CLASSIFIED Member
- Total Posts : 2872
- Reward points : 0
- Joined: 2010/02/06 09:33:05
- Status: offline
- Ribbons : 14
Re:Native Linux SMP questions
2012/12/01 19:10:03
(permalink)
ooh, ext4 very very bad for bigadv... there is a change you will need to make (to /etc/fstab I think) to get the write problem fixed.
|
Grandpa_01
New Member
- Total Posts : 92
- Reward points : 0
- Joined: 2012/04/28 20:59:00
- Status: offline
- Ribbons : 0
Re:Native Linux SMP questions
2012/12/01 21:13:56
(permalink)
No the long wait will not be fixed ext4 and bigadv do not get along. I do not know about Linux Mint 14 but on the latest versions of Ubuntu barriers are enabled on all file systems not just ext4. In Ubuntu you can disable barriers using
( sudo mount / -o rw,remount,barrier=0 ) but you will need to do that every time you reboot, I am not sure if that command will work for Mint but it should.
The permanent fix is posted over at the [H] somewhere but I do not remember where.
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Native Linux SMP questions
2012/12/02 03:44:00
(permalink)
Depends upon how you installed Linux (discrete stand-alone installation) or from within Windows using Wubi or Mint4Win. If Linux is installed within Windows, from what I've read, it has to use the EXT4 file system, and I have never found a problem with long uploads write-times on Bigadvs. I've used several forms of Linux installed within Windows (Ubuntu via Wubi and Mint via Mint4Win). My post-Folding processing (including upload time) is 6-mins at most for a Bigadv vs the 30-mins that has been reported with EXT4 in a standalone Linux install. Many people have reported the same thing about EXT4 not being an issue as long as the Linux OS installation is managed within Windows.
post edited by texinga - 2012/12/02 14:31:10
|
Grandpa_01
New Member
- Total Posts : 92
- Reward points : 0
- Joined: 2012/04/28 20:59:00
- Status: offline
- Ribbons : 0
Re:Native Linux SMP questions
2012/12/02 07:23:01
(permalink)
texigna that is correct when Wubi is used and you do a along side install on a Windows machine, but it does not use a true ext4 file system when it is done that way on a Windows machine. If you do a stand alone install using Wuibi the problem will still be there and it can take up to 1 1/2 hrs to write to disk after the WU completes before it sends. It appears that Wolf has the problem due to the long write time he mentioned above and he will need to disable barriers. The un noticeable write delay on a smp unit does not mean the problem is not there it just means the smp WU's are very small compared to bigadv thus not a noticeable delay in write times.
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Native Linux SMP questions
2012/12/02 08:56:20
(permalink)
Yeah, I can't imagine doing a Wubi-based install of Linux without being in Windows. I mean, what's the point? Wubi, in my mind, is mainly useful for those people that need to run a dual boot operation (Windows and Linux). I've used Wubi for that very reason on several rigs, but hafta agree that I prefer (for a dedicated Folding rig) to do a straight install of Linux where I can choose the EXT3 file system and few other custom parameters. Even today, I have some rigs on 10.10 (with EXT3) and other rigs with Ubuntu/Mint installed from within Windows using Wubi. I was just trying to clarify that not all Linux installations have to use the EXT3 file system to avoid the long upload write-time issue that is oft reported. Someone can correct me if I have this wrong, but from what I've read about installing Linux using Wubi (within Windows), the Linux file system is stored as a single Windows file and to Linux is an EXT4 format. I think that I read somewhere that Windows uses some file manager (or something) that allows it to work with that EXT4 system. I'm no Linux expert, but just saying that people don't have to avoid all EXT4 based installations (or especially a Wubi installation of Linux from within Windows). I'm not saying that Wubi is optimal by any stretch either, but can be a solution for those that need Linux to be installed in that kind of configuration. For me, a dedicated Linux 10.10 installation is still King for Folding.
post edited by texinga - 2012/12/02 14:29:40
|
carldxpc
New Member
- Total Posts : 66
- Reward points : 0
- Joined: 2012/02/04 13:04:37
- Location: New Jersey
- Status: offline
- Ribbons : 0
Re:Native Linux SMP questions
2012/12/02 10:27:33
(permalink)
I only seen the download/upload hang one time running Ubuntu 12.04 LTS with the Phenom 2 X4 920. I think after 15 minutes? it finally would upload. I had the WU set to finish, and it was a long running 2 day+ WU too. This was with 7.2.9, and I'm sure Ubuntu was on the "default" installation of swap/boot ext2/root ext4. I didn't see it happen with faster, well less than "2" day, WUs. Native installation, of course.
post edited by carldxpc - 2012/12/02 10:28:53
ASUS P9X79 Pro/i7 3930k Antec H2O 920/Corsair Vengence 8GB/EVGA GTX680/ WD Black 640GB/Sound Blaster X-Fi Ti/PC Power 900w/Antec 300 modded DFI UT LP X58/i7 920 cooled Thermalright Ultra 120/Corsair 6GB/HD 7870 for dnet/WD Black 640GB/Sound Blaster X-Fi Ti/PC Power 750w/Antec 300 modded ASUS M3A78/Phenom II x4 920 cooled stock/2GB Crucial/GT 440/WD Blue IDE 200GB/Sound Blaster X-FI/PC Power 750w/Antec 300 modded
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/02 13:40:26
(permalink)
The problem I was seeing wasn't so much with the upload but with the write time of the WU as it was finished at 100%.
There seemed to be a long hang or write time before the actual attempt send. Witch didn't happen in the time I waited.
Here is a question. If I'd like to drop to a dos shell from the GUI is this possible once in the GUI?
I tested not booting right into the GUI & logging in at the console text shell, but the command "sudo service gdm start" doesn't allow me to enter the GUI from there if I wish to.
Perhaps I'm misunderstanding the command and what it suppose to do? It doesn't work for Linux Mint 14.
Thanks
|
texinga
CLASSIFIED ULTRA Member
- Total Posts : 5121
- Reward points : 0
- Joined: 2010/07/03 14:30:32
- Status: offline
- Ribbons : 20
Re:Native Linux SMP questions
2012/12/02 14:28:37
(permalink)
Wolf, I should have said "write time" issue and not upload (have corrected my post above). The problem mentioned with the EXT4 file system does have to do with write time IIRC.
When you say "drop to a Dos Shell", do you mean run under a Terminal window only without the GUI? I was thinking that is what you meant. I haven't tried doing that, but I know we have some Linux people 'round here that do know that answer. I'll be watching too for that answer...
|
rklapp
CLASSIFIED Member
- Total Posts : 2495
- Reward points : 0
- Joined: 2009/10/20 12:42:37
- Location: Las Vegas, NV
- Status: offline
- Ribbons : 8
Re:Native Linux SMP questions
2012/12/02 15:09:43
(permalink)
Native linux was slightly faster than virtual linux but the main advantage I found with native linux is that the overclock is much more stable so I could run at a higher clock speed for days on end without having to worry about a blue screen. The disadvantage is that I could never get gpu clients to work in native linux.
|
Punchy
CLASSIFIED Member
- Total Posts : 2872
- Reward points : 0
- Joined: 2010/02/06 09:33:05
- Status: offline
- Ribbons : 14
Re:Native Linux SMP questions
2012/12/02 15:19:10
(permalink)
If you started in single-user mode at the command shell (or even multiuser without graphics), the easiest way I know of to get to graphical mode is "sudo /sbin/init 5" (at least that's my guess for Ubuntu). "sudo /sbin/init 3" will probably get you from graphical mode to terminal mode.
post edited by Punchy - 2012/12/02 15:20:59
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/02 16:25:15
(permalink)
texinga
Wolf, I should have said "write time" issue and not upload (have corrected my post above). The problem mentioned with the EXT4 file system does have to do with write time IIRC.
When you say "drop to a Dos Shell", do you mean run under a Terminal window only without the GUI? I was thinking that is what you meant. I haven't tried doing that, but I know we have some Linux people 'round here that do know that answer. I'll be watching too for that answer...
Just thought I let you know with the Wubi type setup/installation in Linux Mint 14 you can now pick what EXT type you would like to use. Found this out on yet another fresh install of Mint14 last night. You will have to pick custom over into the install and at the HDD size screen there is a place to change the type EXT you would like the install to have. I am now running EXT3 with the Linux Mint 14 install. So I guess I'll see how that works out. Seems like there was something else I found that I wanted to tell you about, but its not coming to me right now. lol Punchy
If you started in single-user mode at the command shell (or even multiuser without graphics), the easiest way I know of to get to graphical mode is "sudo /sbin/init 5" (at least that's my guess for Ubuntu). "sudo /sbin/init 3" will probably get you from graphical mode to terminal mode.
Thanks, I'll have to test these later. In the middle of a bigadv at the moment. But I find it "for me" to be much easier to setup & install packages in graphical mode but I think my folding time could be some what better by dropping to terminal mode once I'm done in graphical mode. But still would like to pop back into graphical mode from time to time to check on temps and other things I might want to look at or use the Web from that rig. I used this guide, but it didn't do what I wanted, almost got stuck in terminal mode, but I worked my way back.
Once you have everything set up, there is no need to boot into the GUI.
This will let you machine boot to a command line.
1. Open a terminal and type: Code: sudo nano /etc/default/grub
Change this line: Code: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
to this: Code: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash text"
Now, update grub: Code: sudo update-grub
The next time you reboot, you will boot to a command line.
To start the GUI, type: Code: sudo service gdm start
To stop the GUI, type:Code: sudo service gdm stop
post edited by TheWolf - 2012/12/02 16:46:51
|
_IanJ
ACX Member
- Total Posts : 394
- Reward points : 0
- Joined: 2011/04/30 14:22:41
- Status: offline
- Ribbons : 0
Re:Native Linux SMP questions
2012/12/02 17:05:57
(permalink)
TheWolf
I used this guide, but it didn't do what I wanted, almost got stuck in terminal mode, but I worked my way back.
Once you have everything set up, there is no need to boot into the GUI.
This will let you machine boot to a command line.
1. Open a terminal and type: Code: sudo nano /etc/default/grub
Change this line: Code: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash"
to this: Code: GRUB_CMDLINE_LINUX_DEFAULT="quiet splash text"
Now, update grub: Code: sudo update-grub
The next time you reboot, you will boot to a command line.
To start the GUI, type: Code: sudo service gdm start
To stop the GUI, type:Code: sudo service gdm stop
Another way to do this is power on the system and hold down the left shift key just as the motherboard finishes POSTing before the OS starts to load. You will then get the GRUB menu. Highlight the option below "default" and press the c key on your keyboard. At this point you will be editing the boot command. Look for "quiet splash" in the command line, and replace it with "text", without the quotation marks. Then press CTRL+E or press F10. This is good for booting into text occasionally, if you normally still use the GUI. The command will revert back once you reboot the system, and the OS will load the GUI normally on the next boot.
Contact us by phone at 1-888-881-3842 or email at support@evga.com. We are available 24/7 to assist you.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/02 17:17:54
(permalink)
Nice tip, thanks, I'll give it a try. But what I really want to do is drop from the GUI once there, into text command line, then be able to get back into GUI if needed. Another words once dropped down to text terminal mode from GUI start folding in the pure text terminal mode, still be folding & enter back in the GUI mode if need be. Not sure this can be done, but seems I remember doing it with some old Linux "Knoopx 3.7?"
post edited by TheWolf - 2012/12/02 17:21:05
|
_IanJ
ACX Member
- Total Posts : 394
- Reward points : 0
- Joined: 2011/04/30 14:22:41
- Status: offline
- Ribbons : 0
Re:Native Linux SMP questions
2012/12/02 17:24:04
(permalink)
TheWolf
Nice tip, thanks, I'll give it a try.
But what I really want to do is drop from the GUI one there, into text command line,
then be able to get back into GUI if needed.
Another words once dropped down to text terminal mode from GUI start folding in the pure text terminal mode,
still be folding & enter back in the GUI mode if need be.
Not sure this can be done, but seems I remember doing it with some old Linux "Knoopx 3.7?"
It used to be sudo telinit 3 to enter text mode, and then sudo telinit 5 to launch the GUI again, but that does not seem to work with some Linux distributions lately. It does not work for me in my heavily modified Linuxmint install, I just boot into text mode as needed.
Contact us by phone at 1-888-881-3842 or email at support@evga.com. We are available 24/7 to assist you.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/02 17:35:51
(permalink)
EVGATech_IanJ
TheWolf
Nice tip, thanks, I'll give it a try.
But what I really want to do is drop from the GUI one there, into text command line,
then be able to get back into GUI if needed.
Another words once dropped down to text terminal mode from GUI start folding in the pure text terminal mode,
still be folding & enter back in the GUI mode if need be.
Not sure this can be done, but seems I remember doing it with some old Linux "Knoopx 3.7?"
It used to be sudo telinit 3 to enter text mode, and then sudo telinit 5 to launch the GUI again, but that does not seem to work with some Linux distributions lately. It does not work for me in my heavily modified Linuxmint install, I just boot into text mode as needed.
Ya that what I'm trying to avoid, all the stopping & rebooting to get where I want to be. Stopping bigadv WU is a big risk of loss of work & starting back at 0%. Thanks for the extra tips I'll give those a try & post back my findings.
|
Punchy
CLASSIFIED Member
- Total Posts : 2872
- Reward points : 0
- Joined: 2010/02/06 09:33:05
- Status: offline
- Ribbons : 14
Re:Native Linux SMP questions
2012/12/02 17:59:13
(permalink)
/sbin is not always in the path for a standard user; that's why you may need to use /sbin/init (which is the same as /sbin/telinit) instead of just the command name alone. In the good old days, when men were men and you could log in as root, /sbin wasn't needed. Oh, and I've done plenty of testing, and running in text mode does not make a significant performance difference versus running with the GUI.
post edited by Punchy - 2012/12/02 18:04:11
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/02 18:06:28
(permalink)
Punchy
/sbin is not always in the path for a standard user; that's why you may need to use /sbin/init (which is the same as /sbin/telinit) instead of just the command name alone.
In the good old days, when men were men and you could log in as root, /sbin wasn't needed.
Lots of great info I'm getting here! I would like to thank each & everyone for there time in posting to try and help me. Thanks everyone! I'm sure these are not the last questions I will have, so bear with me here.
|
TheWolf
CLASSIFIED Member
- Total Posts : 3841
- Reward points : 0
- Joined: 2007/11/14 16:05:23
- Location: Pascagoula, Ms
- Status: offline
- Ribbons : 9
Re:Native Linux SMP questions
2012/12/02 18:11:54
(permalink)
Punchy
In the good old days, when men were men and you could log in as root, /sbin wasn't needed.
Oh, and I've done plenty of testing, and running in text mode does not make a significant performance difference versus running with the GUI.
Only thing about the GUI is I'm tempted to do other things.  I've found just surfing the web can cut into the folding times pretty good. I still like the old days, as a old guy would I guess. 
|
