Hey. I just insstalled the Boinc version for Mac OS X, but I just keep seeing the message:

2004-07-19 15:17:08 [SETI@home] Message from server: No work available
2004-07-19 15:17:08 [SETI@home] No work from project

What am I doing worng? Is there really no work available? Any help would be appreciated. Thanks.

why open a new Thread

so look here

http://setiweb.ssl.berkeley.edu/forum_thread.php?id=1468

http://setiweb.ssl.berkeley.edu/forum_thread.php?id=1858

http://setiweb.ssl.berkeley.edu/forum_thread.php?id=1629

and so on....

the Server are donwn ore have problems

July 19, 2004
Sometime between 5pm and 12am PDT, the splitters seem to have stopped producing new workunits. We have restarted the splitters and they are back to creating new workunits.




Greetings from Germany NRW
Ulli [/url]

WHY OPEN A NEW THREAD?????


Because we are copping "no work" ane the UC Berkley does nothing about it.

We - the distributed computing network providers do just that - we provide a free distributed computing network with a huge amount of computing power for UC Berkley to use which ever way they please.

My PC stays on just because of the seti@home project -0 I have no other need to keep the CPU chugging away 24/7.

If I can't get the WUs I need to keep my machine busy 24/7 I'll turn it off and save elctricity.

We the users deserve to be informed by Berkley - - the front page isn't good enough.

If they want a huge free computer network - the price they should pay is informing the providers about what is happening - or in the cas of WUs what is not happening.

I have written to http://www.berkeley.edu/ and chose a computing department to have a whinge at.

I encourage all other disatisfied providers to do the same - - only by doing this will we get answers and WUs.

tony
the man from the land down under

> WHY OPEN A NEW THREAD?????
>
>
> Because we are copping "no work" ane the UC Berkley does nothing about it.
>
All the stuff they've done on the front page is considered "doing nothing"? Jeez, I thought "doing nothing" involved sitting in a chair sipping a caffenated beverage and playing the Doom III alpha... On your job's time ;)

Berkeley is indeed working on the WU outage problems. They've added lots of hardware and fixed lots of problems to try to keep our caches as full as possible. But the simple fact is that they were caught off guard at how many people would join, and how fast we'd power through WUs. It was indeed a stupid thing, but they're (and we're to an extent) definatly paying for thtat underestimation now.

> We - the distributed computing network providers do just that - we provide a
> free distributed computing network with a huge amount of computing power for
> UC Berkley to use which ever way they please.
>
> My PC stays on just because of the seti@home project -0 I have no other need
> to keep the CPU chugging away 24/7.
>
> If I can't get the WUs I need to keep my machine busy 24/7 I'll turn it off
> and save elctricity.
>
Then do so. If your computer is going to be useless while Berkeley gets new WUs out, then by all means turn it off. If I didn't run SETI, this computer would be off as well. No sense in wasting electricity to have the computer do nothing.

> We the users deserve to be informed by Berkley - - the front page isn't good
> enough.
>
> If they want a huge free computer network - the price they should pay is
> informing the providers about what is happening - or in the cas of WUs what is
> not happening.
>
We are informed by berkeley, on the very front page you mentioned. If the front page is not good enough, please elaborate exactly what you're looking for (e-mails to all members, a blog where we can find quicker status updates [I think that'd be a neat idea] , etc). I find the front page to be sufficient, and FAR better than SETI 1's extremely rare outage updates. Not that it can't be improved though =)

> I have written to http://www.berkeley.edu/ and chose a computing department to
> have a whinge at.
>
> I encourage all other disatisfied providers to do the same - - only by doing
> this will we get answers and WUs.
>
We'll get our WUs complaining directly to Berkeley or not. They know they've got an angry mob of BOINCers out there who are ready to leave after enough outages, and they desperatly want that free computing power. It'd be stupid of them to not keep us happy and crunching for them. Though letting them know just how angry of a mob we can be isn't a bad idea either, and so I also encourage people to write to Berkeley and see happens.

_______________________________

I just got (yet another) work unit. That's three in the last half hour.
<A>

Eh Up Chaps,

Let's keep this snowball rolling....

Who is getting all those work units? I keep looking at the leaderboard and someone's way out front and pulling away at an astronomical rate. What about us little guys then?

Keep on chugging, I'm down to my last three units and yet another hour defer on connection.

Oh Hum!

24 hours now and always - no work from server.

I have been patient with BOINC because it is a new system, but to be honest this must be one of the worst planned systems that I have come across in all my computing experience.

I cannot understand the systems being so inadequate to cope with demand. Blind Freddy could have forecast that the new system should be able to cope with all current Seti@home members plus room for expansion, from the first time it was switched on.

The current situation is frustrating, the web site mentions changes, fixes and updates to the systems, but for me it has gotten worse not better. If the system can’t cope with the current users that have switched over, why not prevent new joins until the system can cope.
How about putting up some figures on the web site that show how the system is coping. How many work units are being produced and processed in a day.
How does the server decide who gets work units, is it programmed to provide work units for faster systems/teams or is it first come first served.

I am happy to have my system churn away at work units, so please tell us if you have any idea when the system will be able to cope with demand. When will the system be robust enough to cope with all the Seti@home users? - 1 week, 1 month, 1 year , never? Right now it looks like never.

I'd be satisfied if they would just update that front page message more often. Like say once when they arrive in the morning, and again when they go home for the day, maybe? The message out there right now is nearly 2 days old, I think.

I realize they are probably spending most of their time working on actually fixing the problems (and I think that's proper). But I think a bit more frequent communication to all those users out here would help cut down on the flood of questions they are getting. If you look at these help boards, a great deal of the questions here could be eliminated if they just did a bit more communication.

> We are informed by berkeley, on the very front page you mentioned. If the
> front page is not good enough, please elaborate exactly what you're looking
> for

> I cannot understand the systems being so inadequate to cope with demand.
> Blind Freddy could have forecast that the new system should be able to cope
> with all current Seti@home members plus room for expansion, from the first
> time it was switched on.

FYI... Per Rom Walton (one of the developers)

Their yardstick, beta, ran for about 16 months and generated 6 Million results. That's about 375,000 results per month. The project has been "production" for about one month and generated 2.7 Million results. Obviously no one on the project had a feel for the amount of horsepower they were unleashing on June 22nd.

Keep Crunching!

If I only had some work. Going on 40 hours with no work. Should I go back to classic?

Some time ago they propose to return back to classic S@H1.
I still have something to do.

> If I only had some work. Going on 40 hours with no work. Should I go back to
> classic?
>
>

[/url]
Try WAP user info http://slavko.sk/wap/seti

> If I only had some work. Going on 40 hours with no work. Should I go back to
> classic?
>

If you have an HT system you can run one instance of Classic with one instance of BOINC. That way at least you'll be working on something.

Actually, I'd go for a message attached to the messages in the BOINC client.

For instance, "No schedulers responded" could as easily be "No schedulers responded, system down for maintenance until approx. 12:00 UTC". Or "Message from server: No work available" could be "Message from server: No work available, work expected to be available at 21:00 UTC".

Sometimes the biggest gripe with problems is not the actual problem, it's the not knowing. Why, how, or when can dispel many a disgruntlement.

>>>> Tim Bonham - Posted: 20 Jul 2004 23:16:23 UTC
> I'd be satisfied if they would just update that front page message more often.
> Like say once when they arrive in the morning, and again when they go home
> for the day, maybe? The message out there right now is nearly 2 days old, I
> think.
>
> I realize they are probably spending most of their time working on actually
> fixing the problems (and I think that's proper). But I think a bit more
> frequent communication to all those users out here would help cut down on the
> flood of questions they are getting. If you look at these help boards, a
> great deal of the questions here could be eliminated if they just did a bit
> more communication.
>
Ah =) I thought you wanted something substantially more than more frequent updates (which I think everybody here wants :D). I too hope that they're just too busy fixing stuff to update, but it certianly can't take that long to update the front page...


>>>> Mac - Posted: 22 Jul 2004 1:46:13 UTC
> Actually, I'd go for a message attached to the messages in the BOINC client.
>
> For instance, "No schedulers responded" could as easily be "No schedulers
> responded, system down for maintenance until approx. 12:00 UTC". Or "Message
> from server: No work available" could be "Message from server: No work
> available, work expected to be available at 21:00 UTC".
>
That would be awesome! I don't know if the BOINC core server software could do it, but I'd love to see more in depth error messages. Would make me less apprehensive about my cache dwindling down to nothingness.... And it probably could be set up to take about as much time as updating the front page if server software is flexable enough in error messages.

_______________________________

> For instance, "No schedulers responded" could as easily be "No schedulers
> responded, system down for maintenance until approx. 12:00 UTC". Or "Message
> from server: No work available" could be "Message from server: No work
> available, work expected to be available at 21:00 UTC".
>
> Sometimes the biggest gripe with problems is not the actual problem, it's the
> not knowing. Why, how, or when can dispel many a disgruntlement.
Who knows what does it mean to develop piece of software knows that very offten to produce meaningfull error messages is much difficult as do the main work. Same for sure in this case. System doesn't know when it will work, system just know that no scheduler responded and because it did not client doesn't know why. Same with work avaialability. System just knows that there is not work and it is reported.
Ther are no little guys which are producing error messages. As I wrote already, very often it is hard to produce and create error messages.



Try WAP user info http://slavko.sk/wap/seti

Eh Up Chaps....

Me again, Just spotted something a bit strange, BOINC is under so much pressure at the moment and from this little gem I've spotted, can only get worse, here's the spiral, every time connection is deferred, the number of seconds of work requested goes up, addmittedly only by a few seconds, but multiply that by the number of current users it's a significant amount of time....

Fun here isin't it...

N.

Actually I'm wrong, it's more dramatic here's a run of figures from Monday 19th July 2004:

SETI@home - 2004-07-19 06:10:10 - Requesting 522012 seconds of work
SETI@home - 2004-07-19 07:10:25 - Requesting 565980 seconds of work
SETI@home - 2004-07-19 08:10:39 - Requesting 569781 seconds of work
and so on to today....
SETI@home - 2004-07-22 18:19:03 - Requesting 1182037 seconds of work


Hey 'ow's it gonna cope with tomorrow when I'm likely to be requesting 10,000,000 seconds???

It's really fun here!

> Who knows what does it mean to develop piece of software knows that very
> offten to produce meaningfull error messages is much difficult as do the main
> work. Same for sure in this case. System doesn't know when it will work,
> system just know that no scheduler responded and because it did not client
> doesn't know why. Same with work avaialability. System just knows that there
> is not work and it is reported.
> Ther are no little guys which are producing error messages. As I wrote
> already, very often it is hard to produce and create error messages.

That's why you leave it to advance organic AI machines to make these judgment calls (they're also known as human beings or Homo sapiens). At this point in the game the human makes the judgement on 2 or 3 items in regard to the system status, dumps appropriate text in a table, and the error generation just picks up the text. No need for "little guys" producing error messages (at least until the other AI sytems catch up to us).

> SETI@home - 2004-07-19 06:10:10 - Requesting 522012 seconds of work
> SETI@home - 2004-07-19 07:10:25 - Requesting 565980 seconds of work
> SETI@home - 2004-07-19 08:10:39 - Requesting 569781 seconds of work
> and so on to today....
> SETI@home - 2004-07-22 18:19:03 - Requesting 1182037 seconds of work

Hello,
I think these times are based on your cache preferences...

> SETI@home - 2004-07-19 06:10:10 - Requesting 522012 seconds of work
> SETI@home - 2004-07-19 07:10:25 - Requesting 565980 seconds of work
> SETI@home - 2004-07-19 08:10:39 - Requesting 569781 seconds of work
> and so on to today....
> SETI@home - 2004-07-22 18:19:03 - Requesting 1182037 seconds of work

This is a known bug that happens with a failed master fetch file.

<a> [/url]

SETI@home - 2004-08-05 11:16:35 - Deferring communication with project for 1 weeks, 5 days, 5 hours, 38 minutes, and 4 seconds


The whole is more then the sum of its particles. Aristoteles
Best wishes from Berlin(Germany), where you can't see the Milky Way ;-)! [/url]
Basti