It's taken a while....

[phil]

....but I finally think that I have this client sort of settled down a little. It still has some issues but with a little babysitting, it works OK. I finally have the client installed on all my home machines....time to see what they can do :)

[BennyRop]

If you run offline, you have to make sure to upload before they get past 337 generations; and win98 machines seem to need to be rebooted every few days to recover their memory. Real OSes seem to do that if you stop and restart the client.
This client, with it's low scores during early generations and higher scores during the later generations looks like it'll be good at convincing folks to add more machines during those times when most of the machines are cranking out low generations. :) As soon as we don't have to babysit them, that should start taking effect.

Congratulations, Phil.. (/e starts chanting "more, more, more" in the background)

[phil]

:). My main issue was getting it to run consistantly on my workstation and two of my duals....there always seemed to be some problem or reason why the client wasn't running or was running poorly. I am now confident enough to fire it up on the other two duals (part time only). Looking at todays stats, it seems that once it is running well, it does actually run well (I think I even gave Mike a run for his money :)). With a bit of luck, I shouldn't have too many more problems.

A few fixes that I'd like to see:

The Gen 337 nonet error.
The giant mem leak (I hit 420MB and 380MB on one of my duals).
Contacting the server through an ISP's transparent proxy.....an issue affecting the Phase 1 client also (Linux).

[ohms18k]

Phil that mem leak is crazy I have not seen nothing go over 120MB, most of them stay around 80-90MB. I had one machine with a 37MB foot print. I restarted that client cause I thought it was broke.

What happens when you go over 337? Do you lose what's buffered?

[Hans Arne Iversen]

Originally posted by djp
This just worked for me:

I edited the filelist.txt file by deleting the line that starts with CurrentStruct and everything downward. Then I looked at the list of filenames. They all seem to have a substring within the name of "protein_??" where the ?? is some number that increments by 1 for every pair of files. In one of my crashed runs, the last file in the list didn't have a partner with the same protein_?? number, so I removed this widow from the list.

After I saved the edited text file, I ran the client with a -ut switch and it started uploading 336 files cleanly.

On a second botched upload, I didn't have an un-paired file at the end, so I just truncated the file at CurrentStruct and it is currently uploading 337 files to the server!

Oddly enough, after uploading successfully it wrote a fresh filelist.txt file and left a pair of matching *protein_86* files and a trj file. I think I'll move these over to an idle client and see if it will resume cleanly with generation 86 or 87.

http://www.free-dc.org/forum/showthread.php?s=&threadid=3452&highlight=337

[BennyRop]

what are you using for the -g setting, Ohms18k? Those with low numbers are noticing high memory usage. (gonna go switch all my win98 machines to 10, instead of the current 1s and 2s.. :)

[MikeTimbers]

Great to see your production Phil! This team has some major hitters and we're putting up some great numbers with just a few members. We're down to tenth in weekly production but given that the 1-9 places are either much bigger teams or hardware manufacturers, we ROCK!

[phil]

Yeah, considering every team above us has at least 10 more active members, I think we are doing rather well.

My daughters machine (dual XP1800's) and wife's (dual XP2500's @ 2.1GHz) adds a nice boost in performance. My wife's dual only runs during the day and the daughter's is only on for a few hours at night and at weekends. Judging by last weekend's figures, 20-25K per update is possible with all 5 machines @ 100%.