milzmaster

I am pretty new to fah, but i am totally interested whats it all about.

Maybe you can answer me a few question to sweep out my newbie  knowledge about fah.

What is -bigadv?
How can i raise the PPD by using ?-bigadv?
At the moment im folding with my gtx260 console client and a cpu console client with an a3 core and my good old E6600 Conroe.

Is there anything i should improve (exept of my hardware. got No money ;( )

Bigadv work units are much larger work units that take on average 2-3 days to complete and have a preferred deadline of 4 days. If you can make the preferred deadline and also meet the other bonus point requirements, you will get bonus points based on how fast the work units are completed. The faster you complete these units, the better.

To run bigadv, the client need to be able to detect at least 8-cores. There are a number of different Intel processor options that support 8-threads with HT. I have a list of them in a spreadsheet linked below. Also, AMD now has an 8-core and 12-core processor that appears to be performing well.

linuxrouter

Governator

I'm at a  34 min TPF on my 4 GHz 920 smp 8 setup. I'm wondering if perhaps I should scale it down to smp 7 since right now I have a GTX 260 folding on it as well. I do have a $20 Geforce 6600 I picked up off of B-stock to provide a basic GUI which will be going in the 920 rig, but until then not sure how to play this, smp 7 or 8, thanks all.

BTW, would PhysX set either on or off help much with gpu clients, thanks.

Scaling down to smp 7 may help with GPU performance. 34-minutes sounds about right for a 4 GHz 920 if you are GPU folding at the same time. Without GPU folding, you may see a 2-3 minute improvement for TPF. I use 6600 and 6800GT cards for my bigadv rigs currently to help keep the cost down since these systems are not GPU folding.

Excellent stuff LR, got your PM btw thanks. I'll put the 6600 in for GUI then and probably stay with smp 8 core and meanwhile tweak further before CC begins.

folder

Is my vmware folding or not? 
  It threw a wobbly  (something was hanging) so I restarted  it with the intention of running from a back up but when I started it it said process 981 ( I think) is already running and maybe a machine id 1 or something similar.  Nothing is being written in the vmware window, but FAHmon shows it started at 58% and 10mins later moved on to 59%

It sounds like the vm is folding. To be sure you can run this command to see if the process is running:

ps xa |grep fah

Governator

You know, this is driving me nuts  Every now and then the IP address of my bigadv client drops in my vmware machine, forcing me to suspend and reset but more often than not I have to reboot my rig and reset my router, why is this happening?? I just lost a 5% completed client because of this bs, usually though I can restart where I left off. I'm at my wits end on this, suffering this problem since I started bigadv folding, is there any real solution for this such as forcing an IP perhaps or something editable in the config, anything??

If the DHCP lease time on the the router is set to a short amount, this could cause the IP to drop prematurely or there could be some other network issue going on.

If you follow the screenshots in the guide in this thread, there is an option for static IP. You can unselect DHCP and enter your static IP information. If you prefer to start the client manually, you can leave auto-startup unchecked.

DeRico

I know this is a noob question but do I have to have Linux to run BigAdv?

Yes. At this time, bigadv only runs via FahCore_a2.exe which is only available for Linux. At some point, bigadv is supposed to be ported over to FahCore_a3.exe (SMP2) which a Windows client is available for, but I do not know when that will happen.

How do i view the log so i know how long its takin to finish 1%.

linuxrouter

re4l warrior

How do i view the log so i know how long its takin to finish 1%.

You can do it by web or via a program like HFM.net. To access the web page go to this address:


http://IP/log.php


Change IP to the IP assigned to the virtual machine.

thank you i think i'll use HFM.

General Iroh

How you determine your tpf?

TPF is time per frame. One frame is one percent. I would suggest downloading HFM.net. Then either use the instructions in the virtual machine or the guide here to setup the monitoring.

General Iroh

O ok. I was using fahmon. I'll try hfm.net too,

FahMon will tell you TPF also if you go to Tools -> Benchmarks. However, FahMon does not calculate bonus points so it will calculate PPD based on the base points of 25403.

slugbug

Weirdest thing happened last night. I rebooted after applying some updates and when I restarted the -bigadv WU crapped out and deleted itself :(.
It downloaded a new WU and it's the exact same one.

If you have the auto-backup enabled, you can try and restore from an older backup. Stop the client and run the restore:

fahrestore 0

0 is the latest and 4 is the oldest. The backup runs hourly so at most there will be a backlog of the last five hours.

If you configured the client by console the first time around then the web server does not have access to the client.cfg file. To fix that run this command:

chown apache:apache client.cfg

Yoste

First of all, I'd like to thank you for all the hard work you put into this!
Then again, as being quite a noob, I'm having a problem I don't know how to solve myself.
I configured the client through VMWare itself, by following the guide offered by SimC33, but I changed some options via the web config (such as ramdisk and autostartup)
The client wasn't able to get online however, so I changed the network adapter setting to NAT.
I got it up and folding, but somehow I'm not able to find the log file anywhere. Not through the web config, not through the network...
I'm sure it is folding however, cpu is always running at 100%. Any help?

If you hit ALT+F2 and type in the command info, the VM will tell you its IP info. You can view the log by IP:

http://IP/log.php

You can also use a program like HFM.net to monitor your clients. The info command also tells you how to set that up. You can also type in ping -c 5 google.com in the console to confirm good connectivity to the Internet.

ALT+F1 will bring you back to the main console.

Yoste

Thanks for the help!

I'm still not able to get the log showing up, it says "Firefox can't find server at 169.254.89.16".
HFM.net can't find the ip on the network, nor can FahMon.
I tried to ping google, but no luck: "ping: unknown host google.com".
As it isn't apparently able to connect to the network, how can it be folding? There are no files, no work-folder, no log,...?
Thanks again!

The 169 IP indicated that the VM was unable to get an IP from a DHCP server. Do you have a DHCP system setup on your network? If so, a firewall could also be blocking access to the virtual machine. I am not sure how it is able to fold unless the system was able to get a work unit while you were setup in NAT mode. The other possibility is that network access was working temporarily but then went offline for some reason.

Yoste

Hmm, I have a DHCP system setup, yes, but I can't see any firewall blocking VMWare. Windows Firewall is off and the firewall I'm using (Kaspersky) is configured properly. Is there any way to get network access back?

Try changing back to bridged mode. Then run these commands in the console:

cleanup
reboot

After the VM reboots, check what IP it was assigned.

okay, I able to run folding on 7 cores, but I unable to get HFM to recognize it to monitor. Everytime I enter the url, its says does not exist. I can go into my browser and enter url and setup shows up. How can I fix this?

Edit:Same with Fahmon.