• 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.

Cornell is folding -- please help?

Status
Not open for further replies.
Tried out the new program and.....

Your gonna need to spice this up a bit.... back to that sharing deal.... :)

Lays into CPU time much more, very bad from a F@H point of view.

Bandwidth crazy- frankly I've got more CPU power at the shop than bandwidth. With the old version the download was pretty much maxed out. This would crank the hell out of that. Got plenty-o- upload so things might even out somewhat.

The plus side is faster machine will be more useful. If you only have a few on a cable/DSL line, go for it. This wil cut into you F@H cpu time in a big way. As far as running a larger number of these behind one pipe, well one way to find out.

So, how ya gonna spice this up? :D :D


edit: sure does cough up a lot of errors
 
Now you got me pissed off. You got a bunch on GenMay boys on the wagon too.

Just wait till my account is active over there....

Alright, I'm off to the shop. Only reason now is cuz I want to bitch slap that other admin.
 
marty9876 said:
Now you got me pissed off. You got a bunch on GenMay boys on the wagon too.

Just wait till my account is active over there....

Alright, I'm off to the shop. Only reason now is cuz I want to bitch slap that other admin.
Click to expand...
What do you mean by splice it up?

And how is it bad that genmay is involved? Psyduck's program is better than the original, and those guys are helping a lot too...so what's the downside? :confused:

Also, people concerned about bandwidth...Psyduck's latest creation:
Psyduck said:
I had this semi good idea to make a proxy webservice that sits on a fast host relaying WU requests. But in this case the client->proxy stage would compress the WUs. This seems to work quite well, it will generally max out your CPU, but the best thing is that is doesn't max out the net connection as the compression is about 8:1 :)

However this version is not for everybody, there's no change apart from the proxy and if you're on a very fast connection then I doubt you'll see much increase. But worse than that, the server is only on 100MBit connection, so at 500KB per WU I can see it getting starved very quickly.

http://www.7internet.co.uk/public/ctc_contest/CTC_Contest__CV_Bin.rar

Source

http://www.7internet.co.uk/public/ctc_contest/CTC_Contest_CV_Source.rar

The proxy only has one method which is below, so no point posting that.
Code: 
[WebMethod]
		public bool GetNextWorkUnit(ref byte[] pdbFile1, ref byte[] pdbFile2, ref string pdbFile1Name, ref string pdbFile2Name,
			ref string pdbId1, ref string pdbId2, ref string pdb1ChainId, ref string pdb2ChainId, long maxFileLen)
		{
			CEService service = new CEService("http://ctcdn.tc.cornell.edu/BioContest/CEWebService.asmx");
			bool retVal = service.GetNextWorkUnit(ref pdbFile1, ref pdbFile2, ref pdbFile1Name, ref pdbFile2Name, ref pdbId1, ref pdbId2, ref pdb1ChainId, ref pdb2ChainId, maxFileLen);

			MemoryStream stream;
			GZipOutputStream cstream;
			if (pdbFile1 != null)
			{
				stream = new MemoryStream();
				cstream = new GZipOutputStream(stream);
				cstream.Write(pdbFile1, 0, pdbFile1.Length);
				cstream.Finish();
				pdbFile1 = stream.ToArray();
				cstream.Close();
			}
			if (pdbFile2 != null)
			{
				stream = new MemoryStream();
				cstream = new GZipOutputStream(stream);
				cstream.Write(pdbFile2, 0, pdbFile2.Length);
				cstream.Finish();
				pdbFile2 = stream.ToArray();
				cstream.Close();
			}
			return retVal;
		}



Reverse engineering, hence why all the local variables were named flag1 and text1 etc. I don't even know whether the original was written in C# or not. From all the global variables I would guess that it came from a VB coders hands.
Click to expand...
 
Just got tot he shop, let me see what I can cook up.

Proxy- ain't got one of those... Guess I could always figure out how to set one up.

GenMay- Just giving you a hard time- mostly.

So if you finish #1 how about giving 1/2 the winings to the next CPU contest here?
 
just don't give it through paypal. that'll take a cut of it. use a money order or something like that.
 
rogue_jedi said:
just don't give it through paypal. that'll take a cut of it. use a money order or something like that.
Click to expand...

You end up paying more for a money order too though. Either way someone gets a cut :(

 
hey where's your downstairs partner in crime? I'm gonna need some linux help, sent him a PM
 
1 140484 brh27
2 139648 ell28
3 82653 cc323

Damn, managed to pull into first place. Not bad.

 
Yea 1st place. I'm thinking about folding for the last place guy to see what kinda of numbers I'm putting up.
 
You know every time i check in you guys have broken these poor servers.
 
rogue_jedi said:
just don't give it through paypal. that'll take a cut of it. use a money order or something like that.
Click to expand...
use non CC paypal - no fees that way
so we're in first? woohoo, I'll keep my 2.8GHz machine running, maybe borg my brother's and mom's rigs too - that'd be another 5GHz
 
Edit: found link and now running this prog. Hope this helps man. Please, if you are feeling kindly do donate some cash to Team 33 lol.
 
I think the stats server crashed again. Or either they haven't fixed it from the last crash and I haven't been paying attn.
 
Added two of my boxen to this program... Any bets on how long it will be before we complete the whole program ourselves? ;)

 
Heh I also just noticed that the stats are down.

Anyone able to get in to those?

Anyway, I got the rewritten proggy from Genmay up and runnin on an xp 2200 machine.

Not gonna do much more than that, cuz the threaded proggy workin several WU's does chew throug a bit more bandwidth it seems.


 
This program has made me realize just how good F@H really is. F@H is way more stable, customizable, and is so much easier to use. The original CTC only worked the first night brh27 asked us to help then it just went downhill. There servers are grossly underpowered for this programs bandwidth/communicating requirements. There just seems like there was poor planning for the program. The rewritten program from Genmay is better but the CE.exe program still freezes up after a little bit under 3 hrs. If the Cornell people are reading this please consider; compressing your work packets, making your program not freeze when it can't send or recieve work units, not throw out errors all the time, not dependent on direct x and images, either use dual procs or allow multiple program instances to run, and more user friendly in general. I should be able to let this program run for days and not have to tend to it. Thanks for listenning.

-MN Scout

 
MN Scout said:
This program has made me realize just how good F@H really is. F@H is way more stable, customizable, and is so much easier to use. The original CTC only worked the first night brh27 asked us to help then it just went downhill. There servers are grossly underpowered for this programs bandwidth/communicating requirements. There just seems like there was poor planning for the program. The rewritten program from Genmay is better but the CE.exe program still freezes up after a little bit under 3 hrs. If the Cornell people are reading this please consider; compressing your work packets, making your program not freeze when it can't send or recieve work units, not throw out errors all the time, not dependent on direct x and images, either use dual procs or allow multiple program instances to run, and more user friendly in general. I should be able to let this program run for days and not have to tend to it. Thanks for listenning.

-MN Scout

Click to expand...

Yea, I was thinking the same thing. The bandwidth usage is off the charts. Really does make you think about the quality of the Stanford program.
 
Rank Job Total Email Delta
1 347996 brh27
2 194150 ell28 -153846
3 162208 cc323 -185788
4 92304 dcc34 -255692
5 72914 hl336 -275082
6 41705 mts37 -306291
7 34339 yz84 -313657
8 25064 jlb264 -322932
9 13682 awr8 -334314
10 12870 rv47 -335126

Weeeeeee!


edit: Removed emails (sorry)
 
well, the stats server was actually up for a few minutes there... I calculated average points/minute across 5 updates:

brh27 27.55
ell28 11.85
cc323 18.4
dcc34 7.5
hl336 6.6
mts37 2.55
yz84 2.25
jlb264 4.85
awr8 0.45
rv47 0


edit: emails again :p
 
DaSmurf said:
well, the stats server was actually up for a few minutes there... I calculated average points/minute across 5 updates:

brh27 27.55
ell28 11.85
cc323 18.4
dcc34 7.5
hl336 6.6
mts37 2.55
yz84 2.25
jlb264 4.85
awr8 0.45
rv47 0


edit: emails again :p
Click to expand...

huh, that was after I took all mine off for HL, I was doing about 25 by my self.
 
lol, we're absolutely clogging their servers.

The graph of these results looks just like a weak DNS attack.

Fine results.... bandwidth clobbered, few results.... back, gone, back...

rofl
 
These guys are going to have to figure out a better way to run this project if.........big IF there........it ever gets out of beta. It may work fine for just a couple of students running it on campus but I think they seriously underestimated what would happen if people really put some effort into this.
 
Yea, they've both been down about 24 hours now, I dont have high hopes of them returning.
 
Looks like we've opened up quite the lead over everyone else. Stats current as or 8:30cst when i pulled them

1 400471 brh27
2 228283 ell28
3 202155 cc323
4 95193 dcc34
5 83377 hl336
6 45574 mts37
7 37527 yz84
8 29281 jlb264
9 14683 sag44
10 14413 awr8
 
MustangWorld said:
Looks like we've opened up quite the lead over everyone else. Stats current as or 8:30cst when i pulled them

1 400471 brh27
2 228283 ell28
3 202155 cc323
4 95193 dcc34
5 83377 hl336
6 45574 mts37
7 37527 yz84
8 29281 jlb264
9 14683 sag44
10 14413 awr8
Click to expand...
Yeah, with all the help the competition is getting blown away. Thanks everyone. :)
 
Status
Not open for further replies.
Back
Top