Folding@Home Weekly Team News - 4th September 2008

Either Ubuntu is slightly unstable or it's not entirely happy with VMware Workstation 6.

I have the -SMP running as a service

If I leave it on the login screen for any length of time it will fold away quite happily, but if I try and log in, it accepts the username and password and the password box goes grey but then it stays like that. It's still folding I just cant do anything with it except reboot the VM.

I have no problems with the client although it randomly got a SIGTERM signal and stopped folding about an hour ago :confused:

Also as a side note just though I'd illustrate the difference in CPU utilisation between these two WU's:

1st image is VM1 SMP Gromacs 2605
2nd image is VM2 SMP Gromacs CVS 2669





 
Last edited:
Try disabling compiz, it's rather prone on marking windows 'grey' as crashed or whatever even when they aren't.
 
Yeah, CVS WUs are great. Really hoping they come to Windows soon, should see a serious boost for lots of people :D

tm_100.png


Makes me feel all warm inside... or maybe that's just 'cause it's hot in here :D
 
The Linux client sure does know how to be a pain in the backside. Every couple of hours the entire machine hangs, then whilst I was sleeping, I come back to be greeted by this.

Code: 
[22:58:27] Completed 100000 out of 250000 steps  (40%)
[23:09:20] Completed 102500 out of 250000 steps  (41%)
[23:20:11] Completed 105000 out of 250000 steps  (42%)
[02:44:42] - Autosending finished units...
[02:44:43] Trying to send all finished work units
[02:44:43] + No unsent completed units remaining.
[02:44:43] - Autosend completed
[02:44:44] 
[02:44:44] Folding@home Core Shutdown: INTERRUPTED
[02:44:49] CoreStatus = 1 (1)
[02:44:49] Client-core communications error: ERROR 0x1
[02:44:49] Deleting current work unit & continuing...
[03:27:10] ***** Got an Activate signal (2)
[03:27:10] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [September 8 01:48:52]

:confused::mad:
 
Senture said:
The Linux client sure does know how to be a pain in the backside. Every couple of hours the entire machine hangs, then whilst I was sleeping, I come back to be greeted by this.

.......

[/CODE]

:confused::mad:
Click to expand...

This is the one thing holding me back on this, I'm tempted to try running Linux in a VM to get the extra PPD but atm the stability of the SMP is quite good and runs nicely even when using the comp for normal use. Hopefully the a2 core won't be long off and the windows crew will get some PPD lovin.
 
The Linux SMP client is more stable than the Windows SMP client, and gives better PPD most of the time :p I've not had any VM hang on me yet, you might want to check your overclock.

Also Senture, you're missing out on an extra 400 PPD by not using 2 SMP clients on those VMs :p
 
Yeggstry said:
Hmmmm... quite possibly :p even so, you still get approx 10% more PPD with using two SMP clients per VM rather than one :)
Click to expand...

You can see why looking at the CPU usage on the non CVS WU in the graphs I posted.

If I can get these two clients running smoothly I will also install another on each VM.

One thing I love is that I can set VMware Workstation to run in low priority mode, when ungrabbed, and in the background so that I can still use Windows with no detrimental affects. :D
 
It's been a good while since I resorted to this but....


News will have to come tomorrow I'm afraid guys :o


I've worked the last 11 days straight and didn't get home till late - have saved the data for the news so it'll just be a day late in the writing. All this working has done wonders for my PPD though :D
 
Ok linux is so wierd:

It seems to have gone down at 23:52:25 and randomly restarted all by itself at 6:56:57 :confused:

What's up with that? And what's up with the random INTERRUPTED bits, the computer has had no user interaction between 23:00 and now, so what does this mean?

Code: 
[23:50:42] Completed 5000 out of 500000 steps  (1 percent)
[23:52:20] - Autosending finished units...
[23:52:20] Trying to send all finished work units
[23:52:20] + No unsent completed units remaining.
[23:52:20] - Autosend completed
[23:52:20] 
[23:52:20] Folding@home Core Shutdown: INTERRUPTED
[23:52:25] CoreStatus = 66 (102)
[23:52:25] + Shutdown requested by user. Exiting.***** Got a SIGTERM signal (15)
[23:52:25] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [June 3 06:56:57] 


# SMP Client ##################################################################
###############################################################################

                       Folding@Home Client Version 6.02

                          http://folding.stanford.edu

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

Launch directory: /home/jack/folding2
Executable: ./fah6
Arguments: -smp -verbosity 9 

[06:56:57] - Ask before connecting: No
[06:56:57] - User name: oceaness (Team 10)
[06:56:57] - User ID: 74D8D78A1C65B615
[06:56:57] - Machine ID: 3
[06:56:57] 
[06:56:57] Loaded queue successfully.
[06:56:57] - Autosending finished units...
[06:56:57] Trying to send all finished work units
[06:56:57] + No unsent completed units remaining.
[06:56:57] - Autosend completed
[06:56:57] 
[06:56:57] + Processing work unit
[06:56:57] Core required: FahCore_a1.exe
[06:56:57] Core found.
[06:56:57] Working on Unit 07 [June 3 06:56:57]
[06:56:57] + Working ...
[06:56:57] - Calling './mpiexec -np 4 -host 127.0.0.1 ./FahCore_a1.exe -dir work/ -suffix 07 -checkpoint 30 -verbose -lifeline 6424 -version 602'

[06:56:57] 
[06:56:57] *------------------------------*
[06:56:57] Folding@Home Gromacs SMP Core
[06:56:57] Version 1.74 (November 27, 2006)
[06:56:57] 
[06:56:57] Preparing to commence simulation
[06:56:57] - Ensuring status. Please wait.
[06:56:58] 
[06:56:58] Project: 2605 (Run 15, Clone 219, Gen 86)
[06:56:58] 
[06:56:58] Assembly optimizations on if available.
[06:56:58] Entering M.D.

The other VM hung as well last night:

Code: 
[00:21:03] Completed 92500 out of 250000 steps  (37%)
[00:29:45] Completed 95000 out of 250000 steps  (38%)
[00:38:28] Completed 97500 out of 250000 steps  (39%)
[07:06:35] - Autosending finished units...
[07:06:35] Trying to send all finished work units
[07:06:35] + No unsent completed units remaining.
[07:06:35] - Autosend completed
[07:06:36] 
[07:06:36] Folding@home Core Shutdown: INTERRUPTED
[13:06:35] - Autosending finished units...
[13:06:35] Trying to send all finished work units
[13:06:35] + No unsent completed units remaining.
[13:06:35] - Autosend completed


--- Opening Log file [June 3 13:28:41] 


# SMP Client ##################################################################
###############################################################################

                       Folding@Home Client Version 6.02

                          http://folding.stanford.edu

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

Launch directory: /home/jack/folding1
Executable: ./fah6
Arguments: -smp -verbosity 9 

[13:28:41] - Ask before connecting: No
[13:28:41] - User name: oceaness (Team 10)
[13:28:41] - User ID: 264E75286FEF4DBC
[13:28:41] - Machine ID: 1
[13:28:41] 

A potential conflict was detected:

Process 4793 is currently running and may also be a client with Mach. ID 1.
Program will now exit. Upon restart, this check will not be done -- 
you may wish to check that no client is currently running in
/home/jack/folding1 before restarting.

Please press any key to exit.
[13:28:44] ***** Got a SIGTERM signal (15)
[13:28:44] Killing all core threads

Folding@Home Client Shutdown.


--- Opening Log file [September 12 08:42:23] 


# SMP Client ##################################################################
###############################################################################

                       Folding@Home Client Version 6.02

                          http://folding.stanford.edu

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

Launch directory: /home/jack/folding1
Executable: ./fah6
Arguments: -smp -verbosity 9 

[08:42:23] - Ask before connecting: No
[08:42:23] - User name: oceaness (Team 10)
[08:42:23] - User ID: 264E75286FEF4DBC
[08:42:23] - Machine ID: 1
[08:42:23] 
[08:42:24] Loaded queue successfully.
[08:42:24] - Autosending finished units...
[08:42:24] Trying to send all finished work units
[08:42:24] + No unsent completed units remaining.
[08:42:24] - Autosend completed
[08:42:24] 
[08:42:24] + Processing work unit
[08:42:24] Core required: FahCore_a2.exe
[08:42:24] Core found.
[08:42:24] Working on Unit 07 [September 12 08:42:24]
[08:42:24] + Working ...
[08:42:24] - Calling './mpiexec -np 4 -host 127.0.0.1 ./FahCore_a2.exe -dir work/ -suffix 07 -checkpoint 30 -verbose -lifeline 4744 -version 602'

[08:42:24] 
[08:42:24] *------------------------------*
[08:42:24] Folding@Home Gromacs SMP Core
[08:42:24] Version 2.01 (Wed Aug 13 13:11:25 PDT 2008)
[08:42:24] 
[08:42:24] Preparing to commence simulation
[08:42:24] - Looking at optimizations...
[08:42:24] - Files status OK
[08:42:26] - Expanded 4835771 -> 23983821 (decompressed 495.9 percent)
[08:42:27] Called DecompressByteArray: compressed_data_size=4835771 data_size=23983821, decompressed_data_size=23983821 diff=0
[08:42:27] - Digital signature verified
[08:42:27] 
[08:42:27] Project: 2669 (Run 1, Clone 125, Gen 9)
[08:42:27] 
[08:42:27] Assembly optimizations on if available.
[08:42:27] Entering M.D.
[08:42:33] Will resume from checkpoint file
[08:42:37] Resuming from checkpoint
[08:42:38] Verified work/wudata_07.log
[08:42:38] Verified work/wudata_07.trr
[08:42:38] Verified work/wudata_07.xtc
[08:42:38] Verified work/wudata_07.edr
[08:42:38] Completed 97510 out of 250000 steps  (39%)

Anyone got any idea what's going on :confused:
 
No idea mate, I've been having some severe problems on the windows SMP client recently with the 2665 WU's. It will work fine for a few hours averaging 35-50min per % and then overnight it will just start slowing right down until it basically hangs.

If I restart the client, sometimes it sorts itself out and then the other night, i got a whole load of EUE's and it bombed out. When I restarted, for some reason it dropped from 88% complete it was before down to 53% and started over. Something not right atm so I've disabled it until I can have a proper look tonight and over the weekend.
 
oceaness said:
I should have said I know nothing about Linux :p

What's compiz? What does it do, and how do I disable it?
Click to expand...

Compiz is what gives you the 'fancy' graphics. Wobbly windows and so on.

I'm not booted into linux right at this moment, but if you go to the system menu, top option from there, you should see an item called Appearance. Select that and on the new window go to the final tab (something like 'effects' or similar), and you should see three options. By default the middle option will be selected. Select the top option "None" and hit ok.
Voila, compiz now disabled (hopefully)
 
Visual Effects tab defaulted to None (the top option).

Seems to have sorted itself out now it doesnt seem to be freezing when i try and logon at least.

However either Linux or the Linux client is still complete pants. The Windows SMP client was 10x more reliable and gave me 10x the points because it would actually fold rather than stop randomly in the middle of the night or every few % :rolleyes:
 
You must log in or register to reply here.
Back
Top Bottom