PDA

View Full Version : "Storming" back up in the G@H listings



Jorden
14th August 2001, 08:10
Ah finally I got to the place I set to reach 6 months ago. Place 30. My only problem is that I set that goal to be in 2 months ;)

Oh well, my new goal is place 25. I think in another 3 months:)
My GenoScan tells me that it takes another 24 genes to reach that place. Errm yes, but every listing wants the point counts, not the gene counts, or I would be higher up in the list already ;)

A question here as well. Has anyone tried to rerun a couple of genes continuously already? Did you ever find that at some point it would just stop in the middle of crunching, with either a memory or 'outside of bounds' error?

I had the memory error on a 16aa I ran for 2 weeks continuously with Genome V0.98. I had the 'outside of bounds' error on a 58aa I ran for one week with Genome V0.99

I'm now going into week two with a 62aa that runs at 8.5 to 9 hours. Although I might want to check with a 82aa or 93aa how long that will run :) (Who comes up with these strange numbered genes btw? I thought it would be a 3 week test back in June...)

Jord.

TomNuttall
14th August 2001, 10:57
I was away on holiday for a couple of weeks, returning 2 weeks ago. I left my 2 machines running two instances of genome on each. When I got back my Athlon was still crunching both units and my p3 had 'errored' on one but was still crunching the other.

The reason I run 2 instances is for such an occurance. If one blows up the other still runs. That way I don't lose crunching time:p

Tom

Guru
14th August 2001, 11:04
TomNuttall why don't you use Genome Crash Monitor?

Jorden
14th August 2001, 12:08
Hey Rattledagger, I see you're following me up in the list again :)
Passing the defunct CA-Cowboy. Woohoo :)

Looks like v0.99 is working okay for you, although with a lower time per sequence than I do ;) I guess you're running a 65aa?

Jord.

TomNuttall
14th August 2001, 14:01
Guru

If I had used crash monitor whilst running one instance of G@H and I was away from the computer for 2 weeks, no net access, would it have been able to keep crunching??

I must admit I haven't even looked at crash monitor so I really don't know if it could.

Tom

Guru
14th August 2001, 14:27
You can set it up to restart the client(s [max 2]) x number of times before getting a new unit it also has the ability to run the client(s) with the -nonet option and run the client(s) hidden!

d/load here! (http://www.sms-testing.fi/distributedatmurc/files/GenomeMon.exe)

More info here! (http://www.hughesclan.com/downloads/GenomeMon.html)

Rattledagger
14th August 2001, 15:25
I've only run the 3 test-wu as I said I should do. I also tested if the new client worked with a bad wu, and it did. I've unfortunately deleted all the other bad wu
As for the test-wu, the first was 56aa, then a 61aa and lastly a 95aa.

Looking at seti, I've got around 400 wu to 4k, so it will take some weeks before I've continue genome-producing.

As for THINK, I've nearly done my target to 100 wu. The new RAS-protein is really fast. A normal seti-wu uses more than double the cpu-time.
Maybe Jorden and the 3 other MURC'ers in the club without any returned result could produce one result now... :)

Guru
15th August 2001, 02:17
Rattledagger and/or Jorden have you thought about adding some think info to the distributed at murc site?

Any links or files you would like me to ad/upload to the files (http://www.setiatmurc.f2s.com/files/index.htm) or links (http://www.setiatmurc.f2s.com/main/links.htm) section?

Or maybe put some stuff in the articles (http://www.setiatmurc.f2s.com/articles/index.htm) section?

Jorden
15th August 2001, 08:02
Errrm, Think? Oh that Think. I tried it again, but I have to reboot my PC at least once a day for I have an intermittent Explorer error somewhere that will crash my Windows. With Think not saving anything to the harddrive, I will wait for a while until I have a rock stable PC again :)

TomNuttall
15th August 2001, 10:32
Thanks Guru


Tom

Rattledagger
15th August 2001, 15:05
Guru, let me THINK a little bit on this...
BTW, the forum-links on the articles-page is broken.
Mmm, and the new setiqueue-guide has still not shown up... Someone isn't keeping up the pace... :)

Jorden, did you continue to run your old wu, or did you get a new one? If you got one of the new RAS-wu, I'll guess you can run through the wu in 4 hours if you pause genome in the meantime, or set the ud_344089-process to real-time... Of course, if you don't pause genome or set the priority to real-time, it can take very long time to do a whole wu then you reboot once a day.
Oh, yes, the current UD-version is 1.2 build 2519, and the current think is 1.03g build 2522. If you've got an older version, I'll recommend stopping ud, deleting all files except ud.exe and cs.ud, and restart...
If you didn't re-install windows then you upgraded the cpu, I'll recommend doing it now. If you're running win9x, maybe changing to a more stable OS will fix the problems...

Jorden
15th August 2001, 15:46
Originally posted by Rattledagger
If you didn't re-install windows then you upgraded the cpu, I'll recommend doing it now. If you're running win9x, maybe changing to a more stable OS will fix the problems...
LOL, thanks for the advice. My perrils started after Internet Explorer wanted an update for macromedia flash. I'll be re-installing an old explorer and Internet Explorer this weekend when I find time for it. The rest is working fine :)]
My next major overhaul will come after I got my adsl, which I figure will give me extra problems :)

And about Think? I'll think about it. First I want to run a 200+ genome week :)

Rattledagger
15th August 2001, 16:31
Originally posted by Jorden

And about Think? I'll think about it. First I want to run a 200+ genome week :)

That's easy... It's just for me to put in a number larger than 200 and if questioned say Excel screwed up again... :D

Jorden
17th August 2001, 07:50
I guess I can stop crunching G@H for tonight. Or I must try to reach 300+ by Sunday. Which would be easy as far as I see ;)

Rattledagger
17th August 2001, 16:31
Well, 16,31 wu more, and I don't need to do some "creative calculations" :)

Jorden
17th August 2001, 18:58
Naaaah RD, that would be cheating !!

I uploaded another 13.27 WU about an hour ago. Strange that a 62aa should give a 12.71, yet I upload higher points... I haven't got a clue why that is :)

Jord.

Sisyfos
17th August 2001, 19:58
Ah, you noticed that too.
A 99aa should give you 29.4525, but you get 30.46.

Kim

Rattledagger
18th August 2001, 14:49
13,27 :confused:
Then I throws the numbers into excel I get 13,13...
Either it's slow stats-updating, or either you or excel can't calculate...

Jorden
18th August 2001, 16:47
I blame it on the battery of my solar powered calculator :)

You're right RD. It was 13.13 per sequence. My bad :)
Still different than what I should upload though

Jord.