SETI - why such a SUDDEN flood/slowdown in bandwidth at Berkeley?

[SoulAssassin]

What we need is someone -on- Berkeley campus who's in the general bandwidth pool to setup a massive proxy that only we know about. 🙂

 

[TuffGuy]

well, OK's is working pretty well for me right now. 😀 too bad it takes so long for the results to actually get uploaded to seti.

 

[Smoke]

The people that are really suffering right now are our new recruits trying to setup their computers. Even if they try to use a SetiQueue, that 1st WU must pass through to Berkeley for them to successfully register each and everyone of their computers. I do believe I have this correct but I would like confirmation of that fact so I don't go giving bad information?

 

[Jeff7]

Geez, I can't even connect to any of the proxy servers. I was able to connect to SETI's site though and got a workunit for my laptop and my main PC. But the proxies listed above - can't connect to them.:frown:

 

[Bigwoofer]

Say no to day (seti-unit) trading... over the long haul a 2 day cache gives the Lincoln-Town-Car smoooooth ride 😉 Please don't get discouraged, and... uh...

more more more! 🙂

 

[GunDog]

I also have a setiqueue running if anyone needs WU's, currently have 53 WU's in waiting, and can add more if needed 🙂

www.setiforgundogs.dns2go.com:5517

Crunch ON

Force

edit: adding WU's at this time

 

[Def]

Is it just me or does Berkeley's network seem GHETTO for such a high priced scientific university.

I thought our network was ghetto here at Georgia Tech, but just the residents halls on 1 side of campus have 4 gigabit uplinks to the central router. I'm pretty sure those get pretty saturated during the day(since sometimes 1Mbps to the outside on a 10Mbps connection is a feat).

Well, our email server is still a POS dual Pentium Pro 180, but... 🙂

 

[Kravahn]

Smokeball I think you're right about that one... I've been using the gui on one of my boxes and recently switched over to the CLI... It took me 2 days to finally connect to the berkeley server and get the damned thing running. Even going through OK's proxy I had to initialize the account (thought the user .sah would take care of that, but no go), so the first wu had to be d/l from the berkeley server. After that pain, I got the remainder of my cache from OK's proxy with no problem. I'm a n00b to the TeAm, but I thought I'd relate my experiences with a new rig, since Smokey pointed to the issue 😉

 

[Jerboy]

<< What we need is someone -on- Berkeley campus who's in the general bandwidth pool to setup a massive proxy that only we know about. 🙂 >>




Not to mention capping file-sharing bandwidth. I think that can help somewhat

 

[Poof]

To Smokeball and Kravahn:

If you manage to get one brand new machine connected to SETI to get a user_info.sah file and if someone emails you a WU (yes, you can do this - it's like sneekernet), then you could have that machine get started on the WU, then have it point to a TeAm queue to upload the result, and doing so should "prime the pump" (meaning when the queue connects to SETI at whatever time, it will download WUs for that new client/user). Alternately, the TeAm queue-masters can move some WUs into that queue to prime it once that first machine uploads (and the Setiqueue creates the new queue). Once the client's queue has WUs in it, then the user_info.sah can be copied to all new assimilations. At that point, when they connect to the queue for their first WU, a WU should be available for them.

Setiqueue has made it easy now to manually move extra WUs around to other queues as needed. 🙂

 

[Jerboy]

Try to make file transfer during late at night in California time. If this is not possible I always have 50WU's cached up in case someone needs it(I hate seeing anyone stalled because they're out of WU's).. PM me if you need them, I can send them via ICQ.