Trying to attach to Seti


log in

Advanced search

Message boards : Number crunching : Trying to attach to Seti

Author Message
Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 6009 - Posted: 10 Jul 2004, 9:00:06 UTC

Hi

I have BOINC running predictor on all my pcs. I keep trying to attach to Seti & this is what is happening over the past 2 days:
Appears to attach to Seti & I get a few messages back,
http://setiathome.berkeley.edu/ - 2004-07-10 09:54:41 - Sending request to scheduler: http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi
http://setiathome.berkeley.edu/ - 2004-07-10 09:54:43 - Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi succeeded
SETI@home - 2004-07-10 09:54:43 - Project prefs: no separate prefs for home; using your defaults

Then that's it, it just triggers a dl of about 100 Predictor wus & no Seti wus. (Are they still called wus?)

Any idea what's wrong please?

Profile Mike
Volunteer tester
Avatar
Send message
Joined: 17 Feb 01
Posts: 23357
Credit: 31,773,941
RAC: 24,615
Germany
Message 6039 - Posted: 10 Jul 2004, 10:54:20 UTC

Hi

You have to set the share factor in the preferences.
It will take time you get your first Seti WUs.
The predictors must be finished first the sheduler will contact the server and when work is available you will get it.

greetz Mike

Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 6048 - Posted: 10 Jul 2004, 11:21:05 UTC

Thanks - still no Seti wus.

My BOINC client shows 50% to each project, but the project prefs are 100% each. Should I change my project prefs, or is it automatic?

Profile Guido_A_Waldenmeier_
Send message
Joined: 3 Apr 99
Posts: 482
Credit: 4,774
RAC: 0
Liechtenstein
Message 6051 - Posted: 10 Jul 2004, 11:31:28 UTC

Any idea what's wrong please?
I dont now maybe ;--) i hope you unterstand a joke
[/url] [/url]

Profile ghetto
Send message
Joined: 3 Apr 99
Posts: 9
Credit: 773
RAC: 0
China
Message 6053 - Posted: 10 Jul 2004, 11:34:28 UTC

Choosing One , Crunching for ...

Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 6061 - Posted: 10 Jul 2004, 11:50:10 UTC

I don't think it's a prefs issue. Reading the info the project pref site, the settings are not a percentage:

"The proportion of your computer's resources (processing time and disk space) allocated to SETI@home relative to the other BOINC projects in which you participate. The default is 100. For example, if you participate in two projects and give them resource shares of 100 and 200, the first will get 1/3 of our resources and the second will get 2/3."

So leaving both projects at a 100 resource value should be fine....?

eberndl
Avatar
Send message
Joined: 12 Oct 01
Posts: 539
Credit: 44,947
RAC: 145
Canada
Message 6114 - Posted: 10 Jul 2004, 14:07:36 UTC
Last modified: 10 Jul 2004, 14:08:24 UTC

Nick, Boinc simply makes a ratio between the two numbers that are entered. you currently have them set to 100:100, since there are 2 projects with the same priority, they both demand 50% of the project's time. if you wanted, you could set each one to 50, but as long as the two values are equal, BOINC will treat them the same.

Currently my settings are at 100:10 (SAH:PAH), so I'm at about 91% and 9%, resepctively. By entering relative number, the webpages for the different BOINC projects can remain indepenant of each other.

Pascal, K G
Volunteer tester
Avatar
Send message
Joined: 3 Apr 99
Posts: 2343
Credit: 150,491
RAC: 0
United States
Message 6125 - Posted: 10 Jul 2004, 14:35:50 UTC
Last modified: 10 Jul 2004, 14:41:40 UTC

I run 2 computers at 100 to 900, I get a few P@H and as they hit the low watermark I get a get a few S@H, rinse and repeat.... Also my faster computer is set to get new WUWUs at 3 to 5 days, the other is set at 1 to 3 days.
Seti@Home, it is really about the science, isn’t it!!

Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 6130 - Posted: 10 Jul 2004, 14:52:28 UTC

I've no problem with the science, my problem is that I'm not getting any Seti wus. :-(

Darren
Volunteer tester
Avatar
Send message
Joined: 2 Jul 99
Posts: 259
Credit: 275,618
RAC: 0
United States
Message 6141 - Posted: 10 Jul 2004, 15:21:38 UTC - in response to Message 6130.

> I've no problem with the science, my problem is that I'm not getting any Seti
> wus. :-(

Also keep in mind that it has to hit a low-water mark before there is any chance it will ask for any seti units. If it has a lot of predictor work units in the cache now, it won't try to add more to it until it starts to run low. This message -- Scheduler RPC to http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi succeeded -- tells you that you have successfully connected, so from there you just have to wait until it comes around to a normal low-water mark request.


Darren

Profile Mike
Volunteer tester
Avatar
Send message
Joined: 17 Feb 01
Posts: 23357
Credit: 31,773,941
RAC: 24,615
Germany
Message 6163 - Posted: 10 Jul 2004, 16:11:50 UTC

Hi

Nick, i´ve done this in Beta Test over four weeks before Boinc was live.
Be sure it is an fact of share.
Set sah at 200 and p@h to 50 rise the work between to 5 to 6 days at sah and you will get work when the low watermark is hit.
But you will get only a few work at the first time because the sheduler has to calculate how much work you need for one day.

The reason is ervery time the sheduler want to connect but can´t do so you will get work from the other project.

greetz Mike

Darren
Volunteer tester
Avatar
Send message
Joined: 2 Jul 99
Posts: 259
Credit: 275,618
RAC: 0
United States
Message 6167 - Posted: 10 Jul 2004, 16:31:07 UTC - in response to Message 6163.


> Set sah at 200 and p@h to 50 rise the work between to 5 to 6 days at sah and
> you will get work when the low watermark is hit.

Very dangerous idea if your trying to get seti work on a host that is already successfully downloading predictor work. If on the next low-water mark hit, seti is down for any reason at all - you will get 6 days of work from predictor.

It's much safer to LOWER the settings, instead of raise them - then when seti is down and it contacts predictor instead it will only download a few hours worth of predictor work.

Remember your cache setting propagate to every project you participate in, so if you try to tell seti to send you 6 days of work, you automatically also tell predictor to send you 6 days of work. And you have no way of knowing which connection will successfully come next.


Darren

eberndl
Avatar
Send message
Joined: 12 Oct 01
Posts: 539
Credit: 44,947
RAC: 145
Canada
Message 6196 - Posted: 10 Jul 2004, 18:27:50 UTC

I learned that the hard way.... now have 12 PAH units, and 2 SAH units, still claiming to take 53 hours...

Profile adrianxw
Avatar
Send message
Joined: 14 Jul 99
Posts: 155
Credit: 155,425
RAC: 0
Denmark
Message 6206 - Posted: 10 Jul 2004, 18:52:35 UTC

I have S@H set to 200 and P@H 100, still, I only get Predictor units. I haven't had a Seti unit for days.
<p>
------------------
Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.


Darren
Volunteer tester
Avatar
Send message
Joined: 2 Jul 99
Posts: 259
Credit: 275,618
RAC: 0
United States
Message 6208 - Posted: 10 Jul 2004, 18:59:50 UTC - in response to Message 6206.

> I have S@H set to 200 and P@H 100, still, I only get Predictor units. I
> haven't had a Seti unit for days.

What are your minimum and maximum settings for keeping work on your disk?

How many predictor wu's are currently in your queue?

What message does the scheduler give when it contacts seti?


Darren

Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 6284 - Posted: 10 Jul 2004, 23:27:33 UTC - in response to Message 6141.


> Also keep in mind that it has to hit a low-water mark before there is any
> chance it will ask for any seti units. If it has a lot of predictor work
> units in the cache now, it won't try to add more to it until it starts to run
> low. This message -- Scheduler RPC to
> http://setiboincdata.ssl.berkeley.edu/sah_cgi/cgi succeeded -- tells you that
> you have successfully connected, so from there you just have to wait until it
> comes around to a normal low-water mark request.
>

Thanks, I got the suceeded message earlier today, but got no wus from Seti, instead I got another bunch from Predictor. I have 19 predictor wus queued up right now. I'll wait for a low mark.

Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 6856 - Posted: 12 Jul 2004, 9:02:56 UTC

Hi again

I left my pc on overnight to crunch some wus. This morning, it downloaded another bunch of Predictor wus, but no Seti wus. I am running Boinc now on 3 computers all with the same user account details. 1 computer has Predictor + Seti attached, but the other 2 only are attached to Predictor. Do I need to attach all computers under the same username to the same projects?

John McLeod VII
Volunteer developer
Volunteer tester
Avatar
Send message
Joined: 15 Jul 99
Posts: 24087
Credit: 517,624
RAC: 153
United States
Message 7038 - Posted: 12 Jul 2004, 15:33:10 UTC

There are a couple of things.

Work will only be downloaded when the low water mark is hit. It will be downloaded from the project with the largest resource debt.

When you start a new project, it has a very large resource debt, expect that nearly all of the work done will be for the new project for up to a month while it plays catch up.

Nick Hill
Send message
Joined: 1 Jun 99
Posts: 18
Credit: 171,848
RAC: 0
Isle of Man
Message 7165 - Posted: 12 Jul 2004, 21:16:23 UTC

OK thanks

Message boards : Number crunching : Trying to attach to Seti

Copyright © 2014 University of California