is the friggin server down?

[FoBoT]

i am not getting anything uploaded off this one box

i am using that automated restart bat file , but it is always

Uploading fileset 1/1770...


"no response from server - don't bother trying again later you loser"


:help:

i checked some others and they don't seem to be caching, but they might be stuck
guess i need to restart/check the whole farm :(

[guru]

I'm getting some in but it's a slow process. I just created a script to restart the upload after the client stops on the "no response from server - don't bother trying again later you loser" message.

My bigger hammer seems to be getting the job done. :spank:

guru

[FoBoT]

it is now on 1/1769 , holy crap

i gotta put all my offline boxen onto ECC or OGR-25 or something :bonk:

[willy1]

I'll bet their database server status doesn't look like this...

This is a 4-way version of the same server they are running.

[Fozzie]

I'm getting sometimes 32 gens uploaded before it times out.

Still crap compared to how it should be.

We shouldn't be running scripts to restart the uploads we should just be able to do it.

My boxen have been off DF for well overa week now and I'm still uploading cached work.

56K at work and 64K at home. :sleepy: :sleepy:

Got 7 processes running on each box, while one or 2 are borked the others are uploading.

Gone are the days I could hit 1 mill or more per hour though, lucky to hit half that.

[guru]

I agree! This force feeding of the servers is a real pain! :haddock:

I'm generating over 100MB of results per day and I'm running the slow client.

guru

[MerePeer]

I'm not a bigger farmer so mine just try to upload after each gen, and its obvious that they frequently sit a long time on structure 100: sometimes they get through and sometimes it gets buffered. I know that in general structure 100 takes longer because of the summary calcs it is doing, but I think we can agree that "getting a network timeout" implies you waited some # of seconds before it timed out.

If those seconds had been used for crunching, it would be interesting to know how much quicker the entire 30billion would have taken to complete. Likely we would be done by now ( I figure about 7 days left) . It is their own productivity that is being compromised.

In any case the best thing to do is get this protein over with and hope the next one is longer.
:spank:

[guru]

I think it might be faster for me to just burn the results to CD and mail it in. :crazy:

guru

[FoBoT]

:drums:

things seem to be better

this afternoon i had a box send up over 300 in a row and now my laptop is at 62/207 , that isn't bad :|party|:

[MerePeer]

I agree; not seeing buffering at all. Zoom zoom zoom.:bouncy:

[FoBoT]

they fixed something or changed something on the server, check the 8 pm statsman update, several of my offline dumps must have mostly gone through finally

:cheers:
:|party|:
:notworthy
:elephant:
:drums:
:bouncy:
:thumbs:
:jester:
:smoking:
:)
:D

holY craP !
my laptop is already caught up/dumped !!!

something definetly changed :beep: :beep: :beep:

glad i stayed late at work and added that 48 Ghz for the weekend! :rotfl: