Got a "daily quota exceeded", but on inspection only 12 wu was downloaded today (after 07 UTC). These log-entries seems to point to the problem:

(UTC +2)
--- - 2004-07-04 14:14:08 - Cache low-water mark hit; requesting more work
SETI@home - 2004-07-04 14:14:08 - Requesting 873074 seconds of work
SETI@home - 2004-07-04 14:14:08 - Sending request to scheduler: http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-07-04 14:14:21 - Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
SETI@home - 2004-07-04 14:14:21 - Message from server: No work available
SETI@home - 2004-07-04 14:14:21 - Project prefs: no separate prefs for home; using your defaults
SETI@home - 2004-07-04 14:14:21 - No work from project
SETI@home - 2004-07-04 14:14:21 - Deferring communication with project for 1 hours, 0 minutes, and 0 seconds
--- - 2004-07-04 15:14:33 - Cache low-water mark hit; requesting more work
SETI@home - 2004-07-04 15:14:33 - Requesting 876658 seconds of work
SETI@home - 2004-07-04 15:14:33 - Sending request to scheduler: http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-07-04 15:15:08 - Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-07-04 15:15:08 - No schedulers responded
SETI@home - 2004-07-04 15:15:08 - Deferring communication with project for 1 minutes and 0 seconds
--- - 2004-07-04 15:16:09 - Cache low-water mark hit; requesting more work
SETI@home - 2004-07-04 15:16:09 - Requesting 876757 seconds of work
SETI@home - 2004-07-04 15:16:09 - Sending request to scheduler: http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-07-04 15:16:20 - Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
SETI@home - 2004-07-04 15:16:20 - Message from server: Not sending work - last RPC too recent: 99 sec
SETI@home - 2004-07-04 15:16:20 - No work from project
SETI@home - 2004-07-04 15:16:20 - Deferring communication with project for 10 minutes and 0 seconds
--- - 2004-07-04 15:26:21 - Cache low-water mark hit; requesting more work
SETI@home - 2004-07-04 15:26:21 - Requesting 877628 seconds of work
SETI@home - 2004-07-04 15:26:21 - Sending request to scheduler: http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
SETI@home - 2004-07-04 15:26:31 - Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi succeeded
SETI@home - 2004-07-04 15:26:31 - Message from server: No work available (daily quota exceeded)
SETI@home - 2004-07-04 15:26:31 - No work from project

Since AFAIK the daily quota still is 50 wu/day per machine, the "no schedulers responded" followed by "last RPC too resent" indicates 38 wu never made it across the Atlantic...

Connection is ADSL + freeproxy, so shouldn't be too slow, but even getting a "too resent" seems to take over 10 seconds. Since it timed-out after 35 seconds, maybe the problem is the scheduling-server actually uses too long time assigning wu?
Looking on another machine, this got assigned 6 wu there the deadline has a difference of 18 seconds.


In either case, too slow server or lost along the way, apparently being assigned work but never being notified isn't helping on the currently low availability of work.

>Requesting 877628 seconds of work

What kind of super computer do you have that you are asking for 243 hours of work?
Isn't that a little excessive?

Try downplaying on your requested cache time. Set it to something lower than 14 days to 21 days (exaggerated example given, I hope? ;))


----------------------
Jordâ„¢

> >Requesting 877628 seconds of work
>
> What kind of super computer do you have that you are asking for 243 hours of
> work?
> Isn't that a little excessive?

Not really, seems 243 hours is 10 days of work, and considering there were 5 days this past week where there was little to no work, I'd bet there are many who are trying to get a cache of >5 days!

The "daily quota" stuff was added in the beta when some clients kept requesting work (just like the "RCP to recient" logic), and they wanted to stop the problem without fixing the original bug!

> What kind of super computer do you have that you are asking for 243 hours of
> work?
> Isn't that a little excessive?
>
> Try downplaying on your requested cache time. Set it to something lower than
> 14 days to 21 days (exaggerated example given, I hope? ;))
>
>

Asking for 5 days of work isn't really what much... ;)

> Asking for 5 days of work isn't really what much... ;)

It is when the system is having trouble keeping up with the demand. ;)

<a> [/url]

> > Asking for 5 days of work isn't really what much... ;)
>
> It is when the system is having trouble keeping up with the demand. ;)

Not when the system hasn't been able to provide work in about 7 days!

It's called "catch them when you can"!

If you got systems running 24/7, you want to keep then running, even if BOINC craps out for 3 days this week! (I'm betting on 4-5 myself!) You want work to do!

The ONLY way that users will decrease their caches, is if BOINC shows that's it stable enough to feed work on a REGULAR basis! And that ain't gona happen soon!

People, based on UCB's initial anouncement, and switched systems/farms/teams to BOINC and aren't ready to spend time "undoing" it all, but they want to have work. And will grab what they can!

Heck, BOINC/Seti has what, 5 years of WU's? Though they're in the "old WU" format, Seti/BOINC is suposed to do "more science" on each one.. The UBC folks could have had an old machine sitting in the corner generating SETI2 WU's for the last 8 months!

The lack of work is a lack of planning by UCB!

> The lack of work is a lack of planning by UCB!

This is not in dispute.

I'm simply saying that trying to grab huge caches isn't helping anyone.

<a> [/url]

> > The lack of work is a lack of planning by UCB!
>
> This is not in dispute.
>
> I'm simply saying that trying to grab huge caches isn't helping anyone.

Oh, but you miss the point! It IS helping those with large caches, as they can keep their systems crunching when BOINC is not providing work! Maybe not helping "everyone" but is helping "anyone"!

The caches will grow as the available work goes down.. Trust me.