scheduler kryten Running on the seti status page updated 3 mins ago.. but i cannot connect for ~30mins? any one got the same problem?

Me too
cannot connect for ~60 mins

> scheduler kryten Running on the seti status page updated 3 mins ago.. but i
> cannot connect for ~30mins? any one got the same problem?
>
>
>
Me three!

90mins +


Edited 22:21 UTC:- But the validators not. I just received over 500 hundred credits that is around a thousand today! Also got plenty to keep me going.

Here too...I hope that is not a big problem....:(

SETI@home - 2004-09-27 22:44:44 - Scheduler RPC to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi failed
SETI@home - 2004-09-27 22:44:44 - No schedulers responded

<p>GHz

Hardware Upgrade - Seti@home

yes schedulers are down, at this time, hope this is only a temporarly prob.

Greetings from Germany NRW
Ulli S@h Berkeley's Staff Friends Club m7 ©

> yes schedulers are down, at this time, hope this is only a temporarly prob.

So why it's running on the STATUS page??... it's an imaginary status or what? :D

i hope this is not a new Prob with the Snap...

Greetings from Germany NRW
Ulli S@h Berkeley's Staff Friends Club m7 ©

I've been trying for an hour and 15 minutes. Server status page says scheduler is running. Whaas up wit dat?

L8R....

---

Rick A. - BOINCing right along now.... It can only get better! </p>

I got my last new work unit as follows:

SETI@home - 2004-09-27 03:24:51 - Finished download of 06my04aa.15945.21650.711094.93

The scheduler stopped responding as follows:

SETI@home - 2004-09-27 12:35:00 - No schedulers responded

It's been down ever since (all times Pacific Daylight Time).

Bill

> Me thinks the snap snapped and dats dat!
> Sorry, Couldn't help myself :)

I think the SNAP is gone.. If you notice the "UL/DL moves", the dirs got moved off the SNAP with the last move (they were on SNAP, but just got moved, based on what I see!)

You'll notice that there's been no mention of SNAP in the news, and none of the sponsors are back on the main web page (they went away when SNAP started having major problems....)

LHC is open to 5000, and soon 50,000, yet I've not seen any of the "hickups" that SETI seems to display!

Predictor was good, CP is good, LHC is good, and with Seti, we have the scheduler down for hours, while the status says is still running!

The amount of users and I/O of Predictor, CP, and LHC isn't a drop in the bucket to what Seti is doing. At least credit is flowing.

>>You'll notice that there's been no mention of SNAP in the news, and none of the sponsors are back on the main web page (they went away when SNAP started having major problems....)

All the Sponsors are still listed on the Seti Home Page, including Snap.
Rom stated before on a different thread, that after they got other problems straight that he was sure Dave would get it back on there. It was lost sometime in one of the failures. If they had aboned ship, don't you think they would have removed themselves from all projects.

regards,,
Rocky

> The amount of users and I/O of Predictor, CP, and LHC isn't a drop in the
> bucket to what Seti is doing. At least credit is flowing.
>
>

LHC wants to go to 50k users this week - it blows SETI away! (seti has only got 20k "active" users at best - not those that have ever returned a result, but those that have done so in the last couple of weeks!)

> LHC wants to go to 50k users this week - it blows SETI away! (seti has only
> got 20k "active" users at best - not those that have ever returned a result,
> but those that have done so in the last couple of weeks!)

According to http://www.setisynergy.com/stats/ SETI has 47k active users and 98k active machines...

----- Rom

BOINC Development Team, U.C. Berkeley

>
> All the Sponsors are still listed on the Seti Home Page, including Snap.

Check out http://setiweb.ssl.berkeley.edu/ where the sponsors for "seti/Boinc" WERE listed until the SNAP started going toes up!


>Rom stated before on a different thread, that after they got other problems
> straight that he was sure Dave would get it back on there. It was lost
> sometime in one of the failures.

Do you know HTML? This SHOULD be a five minute fix, and hasn't been done in over a month! If I represented a "sponsor", I'd be applying pressure to get it fixed.. That is if I wanted my company associated with this project! :)

> If they had aboned ship, don't you think they
> would have removed themselves from all projects.
>

No, as Classic seems to be in the same state it has been for years.... I can see that they wouldn't want their company associated with this project, after the bad press that SNAP got!

> According to http://www.setisynergy.com/stats/ SETI has 47k active users and
> 98k active machines...


Come on Rom..

(BTW, I like the fact that you don't use Seti/boinc stats for this - you got DB access - but base it on a third party stats site!)

You know as well as I do, that the 47k are users that have one or more credits since 6/22. It doesn't mean that that they've even connected since July!

As far as the 98k machines, how many are "dups" as there was a "detach/attach" for that CPU, and the user didn't "merge"...

That's part of the problem, is that stats like this are "unbounded"

Lets bound it a bit.. How many Seti of the 47k users have gotten a credit in the last 21 days? How many of the 98k mancines have gotten a credit in the same timeframe? (I'll bet there's a boatload of CPU's that never got merged!)

That would be a real measure of "active users and hosts".. Using the count of total DB records is just as much of a guess as I make!

Bound the stats!

> Come on Rom..
>
> (BTW, I like the fact that you don't use Seti/boinc stats for this - you got
> DB access - but base it on a third party stats site!)

Well pushing a car past the red line is bad for the car... Any query I could run to give me a better idea about what is going on is only going to cause a problem somewhere else in the system.

An example here:
select count(*) from workunit where need_validate=1

Takes roughly 45 mins to execute. Even with indexes and the like. Executing such a query would probably drop quite a few queries during that 45 mins.

More equipment has been ordered, I just don't have an ETA.

> You know as well as I do, that the 47k are users that have one or more credits
> since 6/22. It doesn't mean that that they've even connected since July!
>
> As far as the 98k machines, how many are "dups" as there was a "detach/attach"
> for that CPU, and the user didn't "merge"...
>
> That's part of the problem, is that stats like this are "unbounded"
>
> Lets bound it a bit.. How many Seti of the 47k users have gotten a credit in
> the last 21 days? How many of the 98k mancines have gotten a credit in the
> same timeframe? (I'll bet there's a boatload of CPU's that never got merged!)
>
> That would be a real measure of "active users and hosts".. Using the count of
> total DB records is just as much of a guess as I make!
>
> Bound the stats!

Would be nice, but the only way to get the data at this point comes at a cost that is a little to high right now.

Being a BOINC Dev has its advantages at this point. You can spend as much time as you like arguing which project is better than which, but in the end you are still using BOINC to get the job done.

I and the other development staff will just keep making it better.

----- Rom

BOINC Development Team, U.C. Berkeley

> An example here:
> select count(*) from workunit where need_validate=1
>

I am curious how long it would take to run the following (given that
you have indexed the column need_validate):

select count(need_validate) from workunit where need_validate=1

This line of code should execute MUCH FASTER.

At last something to whinge about again
Yipee the schedulers down
I thought I was going to have to change my name to The happy worm
Every cloud has a silver lining