• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

MIA

P

Patriot

[H]ard|DCer of the Month - March 2011/June 2013/De
Joined
Dec 15, 2010
Messages
2,502
Hey guys...
stanford shows 10 active clients in the past 7 days...
http://fah-web.stanford.edu/cgi-bin/main.py?qtype=userpage&username=Eagle07

I did find 2 systems down... but I had 16 running... and they are all running again...

any Ideas... I checked them all and they were successfully dropping Stanford units...
I will check them all again on thursday to make sure they somehow didn't drop the client settings... I don't see how they could... I am behind a proxy so they wouldn't be getting connected without my client files being right...

I am missing about 2-300k... one of the systems down did 100k but...

dunno...
 
Any Ideas... I have looked through the logs is there something I am missing that I should be looking for...

what is a normal 100% core status.... since I don't know which of the the 16 the 6 are its kinda hard to know what is right and what isn't.....
 
I take it they are on all different networks......?
 
i figure you've already looked, but double check to make sure team # and user ID are both correct.. I've fired up more than a few clients with a simple spelling mistake!
 
Activate: AMD said:
i figure you've already looked, but double check to make sure team # and user ID are both correct.. I've fired up more than a few clients with a simple spelling mistake!
Click to expand...

I'd also make sure the passkey stuck. Sometimes it doesn't
 
Kendrak said:
I take it they are on all different networks......?
Click to expand...

Same network for 16 of the 17...forgot about the other one at home lol...

they received their client file from my usbkey...shhh...I know... but I always run -configonly before I run them and doublecheck...

I have been hit on pts since the a5 core...

I did blow away a few installs... I will have to check the client info... but... if it didn't have the client info... it couldn't get out of the proxy...

I had ruled out the possibility of the client file because of the proxy...

then again... I can type in that info blindfolded... I could have done it without thinking... whew... hmm... well I have 16 client files to check lol...
because if I missed the passkey...
 
should each have a unique user ID?
 
they all have my passkey in ... but I noticed a few get a different user id from that... team and name are correct...

I know I have some machine id overlap >16 clients... but only 1 client per machine....
 
no, the user ID and passkey should be exactly the same. the other possibility is the machine ID's which are in the config file in the client folder might be identical to another client which could be causing the error. if any are the same just go into the config file and change 1 or 2 numbers/letters and save the config.
 
sirmonkey1985 said:
no, the user ID and passkey should be exactly the same. the other possibility is the machine ID's which are in the config file in the client folder might be identical to another client which could be causing the error. if any are the same just go into the config file and change 1 or 2 numbers/letters and save the config.
Click to expand...

well I have 5 mia... so I may lose some wu trying to fix this... but they were not going to me anyways...
 
Machine ids need to be unique within each individual machine, (eg if you have multiple uniprocessor clients or CPU + gpu) but you can use the same machine id on multiple separate machines.

User name, passkey and team name all need to be identical.

H.
 
passkey is in there I don't know why or how they got a diff userid from same passkey... grr

I run all bigadv no gpu here cept at home... then just gpu no smp lol
 
User id and passkey are linked and have to match. Also make sure you have the correct case. Patriot is different to patriot .....

H.
 
um...
change machine id and the user id stays the same.... I have 2 ids across 16 boxen... somehow...
 
Patriot said:
um...
change machine id and the user id stays the same.... I have 2 ids across 16 boxen... somehow...
Click to expand...
User ID is unique to each system and is assigned by Stanford servers. Don't modify that.

Machine IDs need to be different for every client on each system. Change this only if you have the same Machine ID for multiple clients on one system. Actually, if there are the same Machine IDs configured on one system, usually an error will come up notifying you of this. I believe you are running only one client per system, correct? If so, your config should be rather simple.

Don't fret about the User ID because the process is automated and does not require fiddling about.
 
you can't mod user id... the post u linked said it would change if you chaned machine id... it doesnt...

I am fretting because I have 17 clients in total and only 10 that stanford is crediting me for...

with the wu and tpf I have on those clients I should be pulling 800k ppd .... not 450k
 
Patriot said:
I am fretting because I have 17 clients in total and only 10 that stanford is crediting me for...

with the wu and tpf I have on those clients I should be pulling 800k ppd .... not 450k
Click to expand...
I understand and I am also confused providing your setup is OK. By far the easiest way to determine if everything is configured properly at a glance is to have a client monitoring app. You're using Linux and I don't know how you can go about this. I guess you need to check each system individually...

If you have only one client per system, totally ignore MID because it is irrelevant. Check that the User ID is unique for each system. Next, make certain you have your Pass Key configured for each client on every system and that it is the same because that alphanumeric designation identifies you. The only thing that needs to be the same is your Pass Key - that's it.
 
Patriot said:
you can't mod user id... the post u linked said it would change if you chaned machine id... it doesnt...

I am fretting because I have 17 clients in total and only 10 that stanford is crediting me for...

with the wu and tpf I have on those clients I should be pulling 800k ppd .... not 450k
Click to expand...


I would suggest post the initial part of the logs on here so that we can have a look. Whilst i am not doubting your skills lots of extra pars of eyes might help spot the problem.
 
Nathan_P said:
I would suggest post the initial part of the logs on here so that we can have a look. Whilst i am not doubting your skills lots of extra pars of eyes might help spot the problem.
Click to expand...

most are running beta units...
just made a post in the beta section of folding forum...
 
It is quite possible something else is the problem. Although such conflicts are not unheard of even for veteran folders, there is too much work unaccounted for in your case to dismiss the problem as an ID conflict of some sort. I'm thinking it's likely otherwise. I guess we'll find out soon enough whatever the case may be.
 
In my experience when I just copy the folding directory between Linux machines Stanford doesn't always pick up on the fact that I have another client. What I tend to do is rsync the directory over and then empty out the work directory. Fire it UP!

So under that process I have had five clients running and I see the points increase based on wu but Stanford only thinks I have three. What you could be seeing is that scenario combined with the bigadv shortage on A5 to cause the loss of points.
 
Patriot said:
about 5 bigadv wu a day...
Click to expand...
To tell you the truth, I was wondering what happened to your production this week. I hope Stanford can rectify this for you - that's an enormous amount. :(
 
nwrtarget said:
In my experience when I just copy the folding directory between Linux machines Stanford doesn't always pick up on the fact that I have another client. What I tend to do is rsync the directory over and then empty out the work directory. Fire it UP!
Click to expand...
That's because, on Linux, the unique UserID is stored in a file called machinedependent.dat. You need to delete this file whenever you copy a Linux install to another machine.
 
nwrtarget said:
In my experience when I just copy the folding directory between Linux machines Stanford doesn't always pick up on the fact that I have another client. What I tend to do is rsync the directory over and then empty out the work directory. Fire it UP!

So under that process I have had five clients running and I see the points increase based on wu but Stanford only thinks I have three. What you could be seeing is that scenario combined with the bigadv shortage on A5 to cause the loss of points.
Click to expand...

wu count is down by 5 or more... was up to 15-18 a day... now 10-12 under EOC... and 800k is a major loss of points... I could have been breaching 1m a day...


APOLLO said:
To tell you the truth, I was wondering what happened to your production this week. I hope Stanford can rectify this for you - that's an enormous amount. :(
Click to expand...

yeah its big.. but I will recover IF it gets fixed...

Tobit said:
That's because, on Linux, the unique UserID is stored in a file called machinedependent.dat. You need to delete this file whenever you copy a Linux install to another machine.
Click to expand...

bless you... I don't know why this hasn't bitten me before... I may not have had that file being copied before.. I sneakernet with my jumpdrive for rapid deployment...

stopping blasting adjusting client id and restarting... I hope this is it...
 
Usually when i clone them, I just copy the client.cfg and the FAH executable. The other stuff will be created on the fly. I have never had an issue like this, and I rarely set up clients from scratch anymore.
 
the units went in... but I basically got no bonus for most of them...
 
musky said:
Usually when i clone them, I just copy the client.cfg and the FAH executable. The other stuff will be created on the fly. I have never had an issue like this, and I rarely set up clients from scratch anymore.
Click to expand...

I do the same thing...I have a template folder with a preset config and the executable.

I run though the config though to install it as a service. I suppose I could figure out how to create the service manually with the sc create command....but I've been too lazy.
 
ouch.... I shouldn't have done that.... lost 3.15m over 9 days...
 
Good stuff to know in here. I probably never hit this issue because the ones I copy tend to only be around for a short period of time then they are deleted and the server ships. In the future though I will clear that stuff out too.

Thanks to all!
Patriot good luck on getting bonus applied to those units.

At least it wasn't 9 mil over 3.15 days.
 
nwrtarget said:
Good stuff to know in here. I probably never hit this issue because the ones I copy tend to only be around for a short period of time then they are deleted and the server ships. In the future though I will clear that stuff out too.

Thanks to all!
Patriot good luck on getting bonus applied to those units.

At least it wasn't 9 mil over 3.15 days.
Click to expand...

lol... I doubt they will fix it... it was kinda my fault...

and at 800k I can do 3m in short order
 
Patriot said:
lol... I doubt they will fix it... it was kinda my fault...

and at 800k I can do 3m in short order
Click to expand...


if you know how many WU's there are they will just give you the max base value which is 45k points per WU. but like you said i doubt they will do it since this is an user error :p
 
your just teasing cause I lost more points per day than u make in a week :p
 
Patriot said:
your just teasing cause I lost more points per day than u make in a week :p
Click to expand...

Whew I am glad that comment doesn't apply to me anymore! The Linux Bigadv has certainly helped out in that department. Last week I was kicking out 300K PPD then we shipped that server :(
 
by 7im » Thu Mar 10, 2011 6:24 pm

With the same ID, the systems were all working on the exact same work unit, as they all appear to Stanford as the same computer, and they only give credit for a returned work unit one time.

No science wasted, just electricity.
Click to expand...

they didn't all have the same wu.... but w/e
expect my ppd to get out of the dumps and start chasing musky round again...
 
Patriot, even if you don't see anything being credited for your loss, at least you'll be up in a day or so. The team can sure use it.

nwrtarget said:
Whew I am glad that comment doesn't apply to me anymore! The Linux Bigadv has certainly helped out in that department. Last week I was kicking out 300K PPD then we shipped that server :(
Click to expand...
That's unfortunate. Any possibility for new servers in the near future?
 
I have 15 servers reporting to stanford as of this morning... still missing 2 but they could have just not turned in units yet...
 
You must log in or register to reply here.
Back
Top