You can try another one and see if you like it temp wise. I don't think you'd have a problem but Idunno that much about your proc. I do know that mine makes no difference whether I run one or two (temp wise).
Do you know how to set up the second client?
Very simple. Do exactly what you did before. Make a folder and put a copy of the client in it. Run it and configure it and you are good to go. The only thing different is that the machineID=2 instead of 1. Then you will be doing it!
Made new folder - downloaded FAH exe - it ran automatically I answered the ?'s as per quick & dirty guide - it then proceeded to download a wu.
There was no ? about machine id!
Used Prof's info for config only - went through the questions til it got to machine id changed this from 1 to 2 continued to end.
Opened em3 see pic
No info showing on Daisy
When I look at the Queue dat info for either of the console's it shows machine id = 1 (even though I have changed it as above)
The -local flag prevent's the client from looking for the registry enrtry for the unique user id, and creates one in the local folder.
without this in place the client sometimes get's confused and tries to fold the same wu on each virtual core, this lead's to problems and many trashed wu's. Are you running the console or console as a service?
And both showing as machine id 1 is not good either
EMIII may be pointing to the same folder ...is that a possibility?
Box path is different for each boxen.
The EM screen is only showing the details ie.time per frame - wu time etc for one of the boxen the other doesn't show the time factors at all just the two bars one showing 4/100
All the time factors show 00:00:00 etc
The EM screen is only showing the details ie.time per frame - wu time etc for one of the boxen the other doesn't show the time factors at all just the two bars one showing 4/100
All the time factors show 00:00:00 etc
Well I'm not sure 100% if they are working properly cause on each boxen it is taking 1hr 24 per frame.
FAH services are running at CPU 49% 5000+mem CPU 50% 55,000+mem
Don't know why the big difference in mem
Spike what does EUE mean?
CSimon The logs are now showing different machine no's - although one of the screens in EM is still showing both as 1 perhaps this will change after the first wu is finished on the new boxen.
Thank you both for being so patient with me, will keep you posted as to whether they upload to Stanford okay (1st on or about the 09/12/06)
Well just to be sure choose the button I have circled in green. Then when the screen comes up tab over to this page. Then highlight your first box like I have circled in red. Then browse with the button I have circled in magenta.
If you follow those steps it should take you to the correct box. If not then bring it to the correct box and select the correct client.
Repeat the same for the other box and that should verify that you made no mistake setting up EMIII in that respect.
Looking good EUE's are early unit end's, they occur sometimes. It can be a bad WU but the bad one's are usually weeded out by the Beta tester's. Data corruption is the usual cause for an eue you get thing's like NAN which is not a number e.g. trying to divide by zero. Overclocking, bad ram, overheating are all contributory factors to causing eue's.
Looks like you need the -local flag especially important on a HT machine. Which is 2 virtual cores not 2 physical cores
At the moment both of my FAH Console exe just start as is.
Having discovered that I have hyper threading - Do I need to add the -local flag to the console exe program start line
ie) console exe -local
or not
If yes, can I do it now whilst the wu are already started by stopping the services adding the parameter then re-starting the services -
or will this break the wu's.
Also went onto the stanford site to check my stats and it says that only one processor used in last 7 days would this be correct as hopefully I now have two wu's working away.
My bad you are running as a service and that automatically applies the -local flag, you are good to go you will not see 2 processor's until you return a WU from both cores on the HT machine.
Are the 1066/67's only obtainable with the -advmethods flag Leo? I didn't follow that whole comment at the community.
0
LeonardoWake up and smell the glaciersEagle River, AlaskaIcrontian
edited December 2006
You mean 1166 and 1167. I wouldn't know if -advmethods off would draw one of those units or not. All my clients are version 5.04, which has advmethods on by default. In the advanced settings for the client, default for -advmethods is "Yes." If you run 5.04 it is redundant to set that flag in the client startup properties.
Comments
Do you know how to set up the second client?
please advise in simple terms feeling fragile today
There was no ? about machine id!
Used Prof's info for config only - went through the questions til it got to machine id changed this from 1 to 2 continued to end.
Opened em3 see pic
No info showing on Daisy
When I look at the Queue dat info for either of the console's it shows machine id = 1 (even though I have changed it as above)
:banghead: What have I done wrong? Advice needed
Closed and re-opend Em3
No apparent change
Went into the em config from icon for each boxen each shows a different machine id
Then went into the queue dat from icon - each boxen is showing machine id
= 1
pics available
Hope these help
the em3 screen is now showing daisy as having 94 hours to complete BUT
the above pics are still the same.
Spike just saw your reply as I was writing this
please explain
without this in place the client sometimes get's confused and tries to fold the same wu on each virtual core, this lead's to problems and many trashed wu's. Are you running the console or console as a service?
And both showing as machine id 1 is not good either
Pic of Services
sorry about this guys & girls, this is a twilight zone for me, but am trying to learn as fast as possible. One step at a time.
Box path is different for each boxen.
The EM screen is only showing the details ie.time per frame - wu time etc for one of the boxen the other doesn't show the time factors at all just the two bars one showing 4/100
All the time factors show 00:00:00 etc
below is the end part of the log file
Launch directory: F:\foldinghome2
Service: F:\foldinghome2\FAH504-Console.exe
Arguments: -svcstart
Launched as a service.
Entered F:\foldinghome2 to do work.
[21:29:13] - Ask before connecting: No
[21:29:13] - User name: Kentigern (Team 93)
[21:29:13] - User ID: 31A213BD175D90CA
[21:29:13] - Machine ID: 2
[21:29:13]
[21:29:13] Loaded queue successfully.
[21:29:13] + Benchmarking ...
[21:29:16]
[21:29:16] + Processing work unit
[21:29:16] Core required: FahCore_78.exe
[21:29:16] Core found.
[21:29:17] Working on Unit 01 [December 6 21:29:17]
[21:29:17] + Working ...
[21:29:17]
[21:29:17] *
*
[21:29:17] Folding@Home Gromacs Core
[21:29:17] Version 1.90 (March 8, 2006)
[21:29:17]
[21:29:17] Preparing to commence simulation
[21:29:17] - Looking at optimizations...
[21:29:17] - Files status OK
[21:29:21] - Expanded 1569419 -> 8082985 (decompressed 515.0 percent)
[21:29:21]
[21:29:21] Project: 1862 (Run 6, Clone 47, Gen 8)
[21:29:21]
[21:29:22] Assembly optimizations on if available.
[21:29:22] Entering M.D.
[21:29:43] (Starting from checkpoint)
[21:29:43] Protein: p1862_Myosin6_PT_US_TIP3P_bbox
[21:29:43]
[21:29:43] Writing local files
[21:29:43] Completed 22717 out of 500000 steps (5)
[21:29:46] Extra SSE boost OK.
There you have it ...you did it right!
Don't know right off what kind of issues you could be having w/ EMIII ...maybe we can look into that further?
FAH services are running at CPU 49% 5000+mem CPU 50% 55,000+mem
Don't know why the big difference in mem
Spike what does EUE mean?
CSimon The logs are now showing different machine no's - although one of the screens in EM is still showing both as 1 perhaps this will change after the first wu is finished on the new boxen.
Thank you both for being so patient with me, will keep you posted as to whether they upload to Stanford okay (1st on or about the 09/12/06)
cheers Kentigern
If you follow those steps it should take you to the correct box. If not then bring it to the correct box and select the correct client.
Repeat the same for the other box and that should verify that you made no mistake setting up EMIII in that respect.
Let me know how it goes.
This mornings pics are
Pippa
EM3
At the moment both of my FAH Console exe just start as is.
Having discovered that I have hyper threading - Do I need to add the -local flag to the console exe program start line
ie) console exe -local
or not
If yes, can I do it now whilst the wu are already started by stopping the services adding the parameter then re-starting the services -
or will this break the wu's.
Also went onto the stanford site to check my stats and it says that only one processor used in last 7 days would this be correct as hopefully I now have two wu's working away.
YGPM
Have now got two FAH running slow (1.15hr per frame) but steady both doing either a 2124 or a 2125
This is how I have them configured (for machine id read 1 and 2)
[settings]
username=Kentigern
team=93
asknet=no
bigpackets=no
machineid=2
local=2
[http]
active=no
host=localhost
port=8080
usereg=no
[core]
priority=96
cpuusage=96
disableassembly=no
checkpoint=15
ignoredeadlines=yes
[power]
battery=no
[clienttype]
memory=1000
type=0
A big thank you to all for helping me