Ubuntu fun

Associate
Joined
7 Dec 2005
Posts
1,667
Location
Hartley
My VM's like throwing this error at me

Code:
# SMP Client ##################################################################
###############################################################################

                       Folding@Home Client Version 6.02

                          http://folding.stanford.edu

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

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

[18:35:47] - Ask before connecting: No
[18:35:47] - User name: oceaness (Team 10)
[18:35:47] - User ID: 7E85C8613BFF583D
[18:35:47] - Machine ID: 1
[18:35:47] 
[18:35:47] Loaded queue successfully.
[18:35:47] - Autosending finished units...
[18:35:47] Trying to send all finished work units
[18:35:47] + No unsent completed units remaining.
[18:35:47] - Autosend completed
[18:35:47] 
[18:35:47] + Processing work unit
[18:35:47] Core required: FahCore_a1.exe
[18:35:47] Core found.
[18:35:47] Working on Unit 09 [October 1 18:35:47]
[18:35:47] + Working ...
[18:35:47] - Calling './mpiexec -np 4 -host 127.0.0.1 ./FahCore_a1.exe -dir work/ -suffix 09 -checkpoint 30 -verbose -lifeline 4317 -version 602'

[18:35:47] 
[18:35:47] *------------------------------*
[18:35:47] Folding@Home Gromacs SMP Core
[18:35:47] Version 1.74 (November 27, 2006)
[18:35:47] 
[18:35:47] Preparing to commence simulation
[18:35:47] - Ensuring status. Please wait.
[18:36:04] - Looking at optimizations...
[18:36:04] - Working with standard loops on this execution.
[18:36:04] - Previous termination of core was improper.
[18:36:04] - Going to use standard loops.
[18:36:04] - Files status OK
[18:36:04] Finalizing output
[18:36:04] - Starting from initial work packet
[18:36:04] 
[18:36:04] Project: 0 (Run 0, Clone 0, Gen 0)
[18:36:04] 
[18:36:04] Error: Could not write local file.  Exiting.
[18:36:04] - Shutting down core

I've tried manually sending the WU but that wont work either.

This is what my first Ubuntu desktop install was doing and now my Ubuntu Server install is doing the same thing, they are set up with 1.5GB of hard disk space of which 65% is being used and 512mb of ram.

So far the only fix I have found is to reinstall Ubuntu but there's no point in having to reinstall Ubuntu every few days or so to make it work :confused:

Prob gonna have to go back to Windows client at this rate unless anyone knows why my clients are doing this :(
 
Maybe worth giving them more disk space and see what happens? My desktop installs are using 54% of 5GB each and the one with 2 clients running is using 635MB of 768MB memory and 175MB of 275MB swap.

I did have one install go bad but since then they've been fine, especially since I leave them on the log-in screen now though that shouldn't be an issue with your server version :o
 
My VM's like throwing this error at me

Code:
# SMP Client ##################################################################
###############################################################################

                       Folding@Home Client Version 6.02

                          http://folding.stanford.edu

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

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

[18:35:47] - Ask before connecting: No
[18:35:47] - User name: oceaness (Team 10)
[18:35:47] - User ID: 7E85C8613BFF583D
[18:35:47] - Machine ID: 1
[18:35:47] 
[18:35:47] Loaded queue successfully.
[18:35:47] - Autosending finished units...
[18:35:47] Trying to send all finished work units
[18:35:47] + No unsent completed units remaining.
[18:35:47] - Autosend completed
[18:35:47] 
[18:35:47] + Processing work unit
[18:35:47] Core required: FahCore_a1.exe
[18:35:47] Core found.
[18:35:47] Working on Unit 09 [October 1 18:35:47]
[18:35:47] + Working ...
[18:35:47] - Calling './mpiexec -np 4 -host 127.0.0.1 ./FahCore_a1.exe -dir work/ -suffix 09 -checkpoint 30 -verbose -lifeline 4317 -version 602'

[18:35:47] 
[18:35:47] *------------------------------*
[18:35:47] Folding@Home Gromacs SMP Core
[18:35:47] Version 1.74 (November 27, 2006)
[18:35:47] 
[18:35:47] Preparing to commence simulation
[18:35:47] - Ensuring status. Please wait.
[18:36:04] - Looking at optimizations...
[18:36:04] - Working with standard loops on this execution.
[18:36:04] - Previous termination of core was improper.
[18:36:04] - Going to use standard loops.
[18:36:04] - Files status OK
[18:36:04] Finalizing output
[18:36:04] - Starting from initial work packet
[18:36:04] 
[18:36:04] Project: 0 (Run 0, Clone 0, Gen 0)
[18:36:04] 
[18:36:04] Error: Could not write local file.  Exiting.
[18:36:04] - Shutting down core

I've tried manually sending the WU but that wont work either.

This is what my first Ubuntu desktop install was doing and now my Ubuntu Server install is doing the same thing, they are set up with 1.5GB of hard disk space of which 65% is being used and 512mb of ram.

So far the only fix I have found is to reinstall Ubuntu but there's no point in having to reinstall Ubuntu every few days or so to make it work :confused:

Prob gonna have to go back to Windows client at this rate unless anyone knows why my clients are doing this :(

Follow the directions here http://foldingforum.org/viewtopic.php?f=44&t=3889

I get this problem daily.
 
That fixed it alright Senture, thanks, but what a lot of hassle having to do that constantly :confused:

Have added a little more memory, up to 768mb now. Will see if that helps at all Rich.
 
[18:36:04] Error: Could not write local file. Exiting.

That suggests it's purely a permissions issue. What user are you running folding as and what user actually owns the directory it's stored in, and with what rights?

If absolutely desperate (and I strongly don't advise this although it'll work) run "chmod -R 777" against the directory
 
I'm afraid I dont know enough about Linux to answer that.

I'm running it as whatever user it asks you to create automatically during setup.
I dont mind finding out if you can tell me how.

My first thought when I saw that error was that it sounded like a permissions problem.
 
Just restarted my smp fun with ubuntu server/VMware 1.04 will let you know how it goes.

I know VMw is on 2.0 not but never had any problems with 1.04 so when back to using it.
 
I'm afraid I dont know enough about Linux to answer that.

I'm running it as whatever user it asks you to create automatically during setup.
I dont mind finding out if you can tell me how.

My first thought when I saw that error was that it sounded like a permissions problem.

I've not played with the Linux SMP client yet, so might get this confused. Which instructions did you follow, these ones? http://folding.stanford.edu/English/LinSMPGuide

If so, bring up a terminal and type in
Code:
cd ~/folding
ls -stlra
and paste the output here.
Then, assuming you've got the client running, do:
Code:
ps waxu | grep fah6
and post the output here as well.
 
Ok I think I've done it right :)

Sorry it took a while, not been home.

lsstlrawn0.jpg


pswaxugrepfah6uo3.jpg
 
Yup, that's exactly what I was wanting, and now baffled by. Looks like it's all owned correctly, and the process is running as the user that owns the files (jack).

Quick introduction to Unix file permissions / ownership. Files can be owned by a single user and a single group, in this case both are "jack". There are three types of permissions you can give, read, write and execute, which are pretty self explanatory names :)
You can set up separate permissions to both, so typically you'd give full read and write to the user, and only read to their group, so members of that group couldn't modify it but could read its content.
There is a third group you can control permissions for with every file or directory, called world or public, and every account is part of world.

Because the user that's running the fah6 process is "jack" (first column of that ps waxu output) We needed to confirm that the user jack had read, write and execute permissions where appropriate. Which they seem to be. Damn :D

One small off chance, could you shown me the output from
Code:
ls -ld ~/folding
 
From what you say it seems to be correct.

lsldfoldingde9.jpg


Oh well it was worth checking.

It seems to be crunching ok at the moment, got a couple of CVS WU's so churning out a cool 10k ppd.

Thanks for taking the time to explain all that Garp. I'm gradually learning a little more about Linux commands :)
 
That does look good too, I was specifically hoping to see the execute permission on that folder, lack of that causes problems for reasons I'm not entirely sure about.

Permissions are the bit that goes (in that last example) drwxr-xr-x. The first bit identifies what kind of file it is, if there is a specific appropriate type, e.g. directory (d), symlink (l).
then you've got Read (r), Write (w), and Execute (x) for the jack user, followed by read and execute for jack group, finally followed by read and execute for world / public.

Oh well, see what happens at the end of this.
 
I've just noticed that whenever the SMP client seems to mess up the timestamps in the log jump miles out. The time in ubuntu is correct it's just FAH for some reason seems to think it's 6:36 when its 20:49. It's normally 1 hour behind my system time because believe its UTC and we're on DST at the moment.

Could these clock issues be arriving through some setting in vmware perhaps?
 
Back
Top Bottom