Posted on 08/02/2006 5:16:19 PM PDT by texas booster
I just got back from visiting my son, who is a linux software engineer with Novell. He configured his home linux server fast as lightning to report on Egon's FAH Monitor. I told him about using Windows XP scheduler on the Windows machines and he said, "Oh, yeah. I can set up *mumble* *mumble* in linux to the same thing. Piece of cake." I didn't have the slightest idea what he was talking about, but he said he'd email me the details. I'll share them here when he does.
Very clever.
One of my other boxes got another one of those. They will take sometime so my average ppd will probably drop while it crunches. I think at 1161 points it is their largest unit.
I'm curious about this WU.
How big is the file? (That's got to be a monster WU.)
What settings do you use for your config file on the system that downloaded it?
How much RAM do you have on that system?
CPU speed?
Intel or AMD?
Windows or Linux?
Would you let us know how long it taks to crunch?
Thanks for any info you can provide.
RT
The current size of the work folder is 25MB. The wu____.dat that was downloaded is 3304956 Bytes. The file wu____.xyz is 13718 KB.
client.cfg has bigpackets=yes
Allow receipt of work assignments and return of work results greater than 5MB in size (such work units may have large memory demands)? yes
Set -advmethods flag always, requesting new advanced scientific cores and/or work units if available? yes
One system is:
2.0GHz AMD Athlon 64 OC to 2.1GHz
1 GB Corsair CL2 DDR400
Debian and Ubuntu Linux
19% after 35 hours.
The other is:
2.4GHz AMD Athlon 64 X2 4600+ Socket 939; haven't overclocked it yet
2 GB OCZ Platinum CL 2-3-2-5 DDR400
Slackware and SUSE Linux
9% after 12 hours.
For this dual core: memory=1000 for each core.
Well, I didn't add all the fields, but I added some. More in the next post...
Well, that got me to looking at the project sheet. Sure enough, it is. I started looking at the next highest, and found one called "p2390_Coiled Coil" that is slightly smaller. But its name caused me some panic, and will lead to a script update.
Dang it!
I've updated the FAHUpload script for both Windows and Linux.
There are two modifications/additions.
First, a patch to fix a potential problem with incoming projects (this one is for the Windows script only). Apparently, FAH has departed from its naming convention for projects. Traditionally, they've been one word, using underscores as word separators. Project "p2390_Coiled Coil", however, departs from this.
As a result, version 1.2 of the Windows script will only return the "p2390_Coiled" portion of the name, and won't match correctly with the remainder for reporting purposes.
This is fixed in version 1.3, available on the Help page. I would recommend updating to that script as soon as possible. (Sorry Klutz!). Freepmail me if you'd rather just have the replacement lines to make the change yourself.
The second modification is the addition of three fields: RAM, Processor, and Notes. Take a look at my Europa machine for an idea on how to utilize these fields. These fields are optional-- and are also available in the 1.3 version of both the Windows and Linux scripts.
I've got 2 set up, and reran Scheduler. I assune I need to wait for the server to update for it to show?
I think Stanford has become the premier university in the US...they have given us Google, employed Condi Rice and recently acquired the Ampex collection - and now this
You don't need to do anything to the scheduler, as long as you overwrite the script with the new one.
You can wait, or you run manually by double-clicking the script.
There's a PASSWORD field under the user section. If you leave it blank, it's not required.
Once you fill in a value, however, you'll need to make sure that ALL your machines use the same value thereafter. Otherwise, your updates won't be accepted.
dfwddr, you may have grabbed the script before I got the new one in place. If so, I can walk you through adding the change.
I grabbed the one without the password twice. I have one working, the other I dumped and grabbed a new one. Still not working, but I think it's a FAH vs. F@H problem (two instances on one machine). I'll LYK if I can't figure it out.
Your config file confirms my suspicion that big WU and Advanced methods must both be set. I won't be getting that WU - not enough horsepower under the hood for that configuration, LOL.
Man, you've got some sweet systems there. I'll be happy to trade you some old P-IIs and P-IIIs for them. I just know you can't resist the challenge to tweak them up to max performance level. After all, your current systems are most likely maxed out - nothing left to tweak - no fun. You really need something more challenging. (Am I getting anywhere, yet? LOL)
;^D
BTW, I suspected you might be running AMD processors and Linux OS. I have this feeling that they outperform Intel and Windows when matched heads up. It will be interesting to see if Egon's stats bear that out over time.
RT
Don't know what it was, but I dumped everything on viper1. Script and scheduled task. Started from scratch, and got it going. Viper2 is still running on the new script, pre-password version. No plans to dump it unless you think there's a conflict in my future.
Y'all check out these two folders of mine if you want another look at the info you can enter. This is pretty cool !!! The only reason I had a problem is because I have two instances on one machine ( and I probally screwed something up by readind the directions AFTER the fact !!! LOL).
No conflict expected. I'm just offering the security of password-protecting the uploads, if and when it becomes necessary.
It's optional until you use it. Once you've implemented it (non-blank) for one machine, it's required across all your machines.
I noticed that SISD has not restarted the labs folding this year. They just have a few systems banging away, probably personal systems or servers.
The only "new" team that has outdone us is TSC!Russia, and their growth has leveled off. Still, they are averaging 110,000 ppd, so that is pretty good leveling.
I am happiest that we have several new folders that have broken into the top 20 of daily producers. Our team is having new members fuel the growth.
Our next goal is to produce 100,000 ppd, and we can hit that by popping up to 200 active members, and having a few current members add a few faster systems.
Next stop, Top 10 daily producers!
I certainly root for Stanford over Ivy league schools. Much of its' faculty is liberal; so is nearly every other major university.
I haven't seen you on the FReeper team folding charts yet. Do you need any help getting started?
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.