//flex table opened by JP

Click to See Complete Forum and Search --> : seti server


meanstreak
02-20-2002, 01:59 PM
I run my wu cached (10 to be exact) and I have not been able to d/l new units all week and now I am down to 2 cached and 5 that need to be sent in (2 got in). Is any one else having this problem and are they working on the server or something?

bigparsnip
02-20-2002, 02:17 PM
I don't seem to be having any problems recently (I did last month), but I think it is just a thing where at the end of most months, the SETI project starts running low on allowed bandwidth, and so has to start dropping connections to keep to their quotas on the university connection.

itismike
02-20-2002, 10:25 PM
Yes Meanstreak, I too have had difficulties transmitting and receiving packets. I just now reset my Setihide connection mode to 'online' and I'm hoping that this will help reduce my frustration by working in the background while I browse the 'net tonight.

-Mike

BJUK
02-21-2002, 02:21 AM
Ive not been able to down load a unit all week. and frankly starting to loose intrest. they say to try between 8/11 am in the uk but this is outside flat rate access times.
If it dont improve soon i might try one of the outher projects.

meanstreak
02-21-2002, 03:18 AM
right now its 12am and the server it still busy im up to 8 to send and one has downloaded

itismike
02-21-2002, 03:32 PM
I understand that the seti@home system will send the same packet out to more than one client and just use the one which is returned first, so crunching workunits which have been stored for a while might not be very productive in reality. But is it possible to set up a mirror site or a collection of units that we can share among the group, and if so, how can we transfer them between us?

-Mike

BJUK
02-21-2002, 03:42 PM
i know from what i have read on the SETI site to set up a miror site is one option but reading between the lines i think it is cost porhibitive.

michaeln
02-21-2002, 04:33 PM
Had problems here also during last night and today. Long delays in sending results and d/l'ing new unit.

With the millions of users worldwide I can only imagine the problems faced by the Berkeley servers

BJUK
02-21-2002, 05:47 PM
It seams to me to be a victim of its own sucsess.

Godfodda
02-24-2002, 05:14 PM
Berkeley has been having bandwidth problems since the first of the year. Seems as if it's being eaten up by the mythical bandwidth ogre (aka stupid kids and mp3 download programs). Very early mornings and weekends are the best time to submit manually. Otherwise SETI Driver's Auto-Transmit option and a large cache is the way to go.

itismike: The server will send the same packet to multiple clients, but only for verification purposes. This redundant system helps to eliminate errors and to identify patched clients. If it doesn't get your packet X back in before it "expires" (28 days, I think), it will resend that one to someone else, too.

And one more note to your team, a couple of members of our team are using SETIQueue to feed WUs to the rest of the team. One person can set this up on his dedicated line to cache and serve WUs for the rest of the team. I don't use it (either serving or leaching), but it's helped quite a few of our members through this bandwidth shortage.