Just wanted to post and thank the EVGA community for turning me on to GPU Tracker.  Makes things SO much easier vs. running the old school clients for both CPU and GPU.

No pressure from me, for now the current tracker works great. I just think that would be a cool feature so that if you go to bed it can be set it to say 12 wu's then If you forget to check it in the morning and don't want it to fold all day for heat reasons or otherwise it will stop folding after the allotted wu's. Good luck with the rest of your classes.

Hi
Having just popped in a P67 mobo and a 2600K, my next step was to put my cpu under 1 unit and then go for big adv :)
 
I have done this but still pulled a smp WU, I understand this is still possible, but on checking in - View client stats, I'm not showing Core A5 just Core A3.
 
Now is this correct? I have read the last few pages here and saw that the update should be automatic.
 
So what I'm saying is: Help please :)
Log:
Arguments: -oneunit -forceasm -smp -bigadv -verbosity 9
[19:20:04] - Ask before connecting: No
[19:20:04] - User name: johnerz (Team 111065)
[19:20:04] - User ID: ...........................
[19:20:04] - Machine ID: 2
[19:20:04]
[19:20:04] Work directory not found. Creating...
[19:20:04] Could not open work queue, generating new queue...
[19:20:04] - Preparing to get new work unit...
[19:20:04] - Autosending finished units... [April 5 19:20:04 UTC]
[19:20:04] Cleaning up work directory
[19:20:04] Trying to send all finished work units
[19:20:04] + No unsent completed units remaining.
[19:20:04] - Autosend completed
[19:20:04] + Attempting to get work packet
[19:20:04] Passkey found
[19:20:04] - Will indicate memory of 4072 MB
[19:20:04] - Detect CPU. Vendor: GenuineIntel, Family: 6, Model: 10, Stepping: 7
[19:20:04] - Connecting to assignment server
[19:20:04] Connecting to http://assign.stanford.edu:8080/
[19:20:05] Posted data.
[19:20:05] Initial: 8F80; - Successful: assigned to (128.143.199.96).
[19:20:05] + News From [link=mailto:Folding@Home]Folding@Home[/link]: Welcome to [link=mailto:Folding@Home]Folding@Home[/link]
[19:20:05] Loaded queue successfully.
[19:20:05] Sent data
[19:20:05] Connecting to http://128.143.199.96:8080/
[19:20:07] Posted data.
[19:20:07] Initial: 0000; - Receiving payload (expected size: 1772304)
[19:20:10] - Downloaded at ~576 kB/s
[19:20:10] - Averaged speed for that direction ~576 kB/s
[19:20:10] + Received work.
[19:20:10] + Closed connections
[19:20:10]
[19:20:10] + Processing work unit
[19:20:10] Core required: FahCore_a3.exe
[19:20:10] Core found.
[19:20:10] Working on queue slot 01 [April 5 19:20:10 UTC]
[19:20:10] + Working ...
[19:20:10] - Calling '.\FahCore_a3.exe -dir work/ -nice 19 -suffix 01 -np 8 -nocpulock -checkpoint 3 -forceasm -verbose -lifeline 6108 -version 630'
[19:20:10]
[19:20:10] *------------------------------*
[19:20:10] [link=mailto:Folding@Home]Folding@Home[/link] Gromacs SMP Core
[19:20:10] Version 2.27 (Dec. 15, 2010)
[19:20:10]
[19:20:10] Preparing to commence simulation
[19:20:10] - Assembly optimizations manually forced on.
[19:20:10] - Not checking prior termination.
[19:20:10] - Expanded 1771792 -> 1967124 (decompressed 111.0 percent)
[19:20:10] Called DecompressByteArray: compressed_data_size=1771792 data_size=1967124, decompressed_data_size=1967124 diff=0
[19:20:10] - Digital signature verified
[19:20:10]
[19:20:10] Project: 7137 (Run 0, Clone 49, Gen 26)
[19:20:10]
[19:20:10] Assembly optimizations on if available.
[19:20:10] Entering M.D.
[19:20:16] Mapping NT from 8 to 8
[19:20:16] Completed 0 out of 500000 steps  (0%)

 


mtm78

A scheduler sounds more like what you're wanting, so you can set it to fold at preset times/stop folding at preset times. What if you set an smp wu on 5 wu's, and you get a strain of wu's which have a TPF one minute higher then 'average', that's allot of minutes difference.. Heat controlled pausing should also be possible ( not on all platforms but on windows it is possible ).

I guess a scheduler is a better way to put it, but if it were set to fold say midnight to 6 am I wouldn't want it to stop folding in the middle of a wu because it was set to stop folding at 3. I think it would be better to have a set number of wu's to fold that way you didn't have anything just hanging waiting to get folded if 6 am came around and you weren't done with the wu.

As Jedi says, it should be 'smart' enough to judge the average runtime of the previous x work unit's where x needs to be large enough to indicate the current spread of work units but not large enough to have data which is invalidated by age ( for instance with tpf references from projects which are no longer active ).
 
 

OK V7 it is then :(

Hey Jedi! I recently re-booted one of my rigs after stopping all clients [smp (-advmethods) and 3 GPU's] to install window updates. When I went to start tracker again, the SMP client does nothing. The logfile just shows that "client was stopped by the user". The GPU clients pick up right where they left off. This happened once on another one of my rigs too. Ended up re-installing tracker to fix the problem. Am I missing something or is this just a glitch? Any way to kick start the client or something?
I still use process lasso from my old manual flag days, just as an FYI.
 

I dumped everything except the client config and client info for the SMP client.. sorry! Tried to get it going before leaving for work and was in a hurry. If it helps, the log file for the SMP simply went through the %complete steps then just stated that I stopped the client, which I did to install the windows updates. No other info there. No error messages or anything like that. It just would not "run". The log file showed no further progression when I restarted it and the client stayed "yellow' all night.

Thanks for your help but I'm stuck at work anyways until 5PM est.
This is only the second time its happened in the months I've been running tracker, so I'm not too worried. I can always -oneunit the GPU's and just re-install tracker afterwards. Just wondering if it's a common issue, or just a random glitch.
The only thing I have left is the two config files in the SMP folder anyway, so I don't know how much help they would be, wait.. they just go to the recycle bin I think.. so I may be able to retrieve the logfiles. I'll check when I get home. Thanks for the help.

OK.. got home.. finished up the GPU wu's. Completely deleted Tracker and re-installed. Clients downloaded fine. GPU's are running. SMP says "running" but is NOT. No core or wu downloaded. Let it sit for a half hour. Nothing in the log file because nothing is working as far as SMP goes.
Stumped here... halp!

cudenver

Is there is way to save your config setting. so each time windows reboots it remembers your settings.

 
Um, if your settings are being wiped out after rebooting you did something seriously wrong. Did you run the Tracker directly from the .zip file without extracting it? If so, you will need to redownload it and set it up properly.
 
How to:
1. Extract zip archive
2. Move extracted folder to someplace where you can keep it (the Tracker sets itself up in whatever folder it's placed in)
3. Run the Tracker
 
You can also move the Tracker's folder without losing config settings because it keeps everything in its main folder.

I was wondering if anyone was using the XML FTP upload option and how it was working for them.
 
I recently installed the F@HMonitor app on my phone and saw that they had the option to do check status via XML uploads. Before I go through the trouble of setting a FTP server, I wanted to see if people had success with this app/upload. Maybe even ask if anyone would be willing to share a small amount of space on their FTP site for the sole purpose of this app/file. I would just like to keep an eye on my machine when I'm at work to ensure everything is running like it should.
 
Thanks all.

nick1551

 Yup, I'm running the FAHMonitor app and gpu tracker.  Works great.  Every now and again you get a ftp upload failure from the tracker, but it doesn't affect it, because it will get the next one.  Allows you to see tpf, wu#, ppd, and completed %, all on your phone.

Setting up a free ftp site is pretty easy, there are a bunch of tutorials just for folding monitoring.

 
Thanks for the reply. I'm about at my wits end with trying to set it up to be honest. I made an account over at drivehq.com and can't seem to get it play nice. Maybe I just need some sleep since not matter what I do seems to working at this point. 
 
At least I know it works and its my fault so I'm not left to wonder if it would never work no matter what I do. 

kody7839

nick1551

Yup, I'm running the FAHMonitor app and gpu tracker.  Works great.  Every now and again you get a ftp upload failure from the tracker, but it doesn't affect it, because it will get the next one.  Allows you to see tpf, wu#, ppd, and completed %, all on your phone.

Setting up a free ftp site is pretty easy, there are a bunch of tutorials just for folding monitoring.

Thanks for the reply. I'm about at my wits end with trying to set it up to be honest. I made an account over at drivehq.com and can't seem to get it play nice. Maybe I just need some sleep since not matter what I do seems to working at this point. 

At least I know it works and its my fault so I'm not left to wonder if it would never work no matter what I do. 

 
I thought I already offered to host you stats? :) Offer still stands anyhow.

Its harder on those sort of sites as theyre uploading to a specific directory but not telling you where. I found another free hoster (made a forum topic in here somewhere) but likely it would be the same issue.
 
Problem is you dont have a direct link to YOUR space.
 
Like mine, which would be, http://sleinous.dyndns.in...ingkody7839/status.xml
 
Also whats not helping is you're not putting "status.xml" at the end of the URL, you NEED to do this or it WONT save the xml is what I found.