What happen to the Tuesday shutdown?

---

What happen to the Tuesday shutdown?

Rain delay.

Joe had an explanation for the 'result creation rate' a few days ago. The splitters make WUs, and put them in the database: but transitioners are needed to turn WUs into results (tasks). Making a WU is a slow, steady process that involves mathematics and data shuffling from disk to disk: making a result from a WU is just a bit of database record-keeping, and takes barely any time. Did you see the database queries/second go over 4,500?

Thanks for the explanation - I must have missed that one. Yes on the database queries / second. Guess they have more horsepower/capacity than I thought. V7 creation rate has dropped back to near-idle now, and pending validation is starting to drop.

---

Another Fred
Support SETI@home when you search the Web with GoodSearch or shop online with GoodShop.

Comment from Eric on Beta

Matt and Jeff are out of town

Might have something to do with it.

---

Here's the big bulge in ready to send that Richard predicted earlier in this thread:

[As of 18 Mar 2014, 17:00:05 UTC]
Data Distribution State SETI@home # Astropulse # As of*
Results ready to send 602,189 26,288 9m

The highest I happened to see RTS was 698K. Now it's down to 648K. Result creation rate is nearly nothing and what little there is is probably accounted for by resends being added to the queue.

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

With well over 300k tasks sitting in the ready to send buffer the production rate should be very low.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

...
Actually the stuck SSP shows more than 300k to send - we do know that a stuck SSP can lead to runaway splitters, when stuck below the highwater mark. I guess if it got stuck _above_ the highwater mark now, splitters are probably not gettting the signal to fire up or at least not working flat out.

Either way, I doubt it will get sorted before maintenance.

edit: small correction, my beta UPload seems to be stuck too.

Well, there's a clue - they were splitting at 15/sec when the page locked. That means they were not at high water mark - so they probably went on, and on, and on...

15 * 3600 * 20 (hours) is over a million tasks split since then. We've been drawing them down, of course, but not so many shorties as in recent weeks. I reckon we'll be bloated.

(and of course they had lots of raw material to work on)

Dr. Anderson's Changeset 2e4d561 is related to this issue. It changed BOINC's sample work generator so it wouldn't go on creating work if the Transitioner got behind or stuck.

Unfortunately, if applied to the splitters here it would mean workunit production would stop if any one of the 6 transitioners got in trouble. That would be a case of the cure being worse than the ailment, I guess.

                                                                  Joe

Half of the tasks on this computer are ap. Luv it!

---

rOZZ
Music
Pictures

No outage today - work flows - everything's fine! :)

---

Aloha, Uli

CreditNew Strikes Again:

3444282044 1455105331 18 Mar 2014, 6:32:51 UTC 19 Mar 2014, 1:47:14 UTC Completed and validated 2.12 0.50 609.32 AstroPulse v6
Anonymous platform (NVIDIA GPU)

Not bad, for 2 seconds of clock time...EAT YOUR HEART OUT!!!!

---

There seems to be a different problem visible:

Replica seconds behind master 0 31h
Results received in last hour 32h
Result turnaround time (last hour average) hours hours 32h

Replica out of sync?

---

CreditNew Strikes Again:

3444282044 1455105331 18 Mar 2014, 6:32:51 UTC 19 Mar 2014, 1:47:14 UTC Completed and validated 2.12 0.50 609.32 AstroPulse v6
Anonymous platform (NVIDIA GPU)

Not bad, for 2 seconds of clock time...EAT YOUR HEART OUT!!!!

You win the Random Number Generator Credit Lottery!!!

And still there are someones (very few at this time i belive) who continue to say: nothing is wrong with creditscrew...

---

One of the Astropulse assimilators has failed according to the server stats page, and therefore a queue of validated Astropulse work units waiting for assimilation has grown.

I like credit_new - today, at least...

3442864585 1454394478 17 Mar 2014, 10:18:44 UTC 19 Mar 2014, 15:15:16 UTC Completed and validated 2.14 0.75 621.99 AstroPulse v6 Anonymous platform (CPU)

---

Member of the 20 Year Club

Well AP splitting is almost finished so get ready for a steady flow of MB work soon.

Cheers.

Well AP splitting is almost finished so get ready for a steady flow of MB work soon.

Cheers.

Wow. Over 10 hours later and no one is wailing over the APs running out.

---

David
Sitting on my butt while others boldly go,
Waiting for a message from a small furry creature from Alpha Centauri.

Wow. Over 10 hours later and no one is wailing over the APs running out.

Many are on double secret probation. Let's all join hands and sing kumbaya.

---

Well AP splitting is almost finished so get ready for a steady flow of MB work soon.

Cheers.

Wow. Over 10 hours later and no one is wailing over the APs running out.

Ah, but the caches are still full. When the bellies get empty, the grumbling will commence :)

Well AP splitting is almost finished so get ready for a steady flow of MB work soon.

Cheers.

Wow. Over 10 hours later and no one is wailing over the APs running out.

Well if we could burn thought the MB faster then more AP would get loaded quicker, yes?

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

Well AP splitting is almost finished so get ready for a steady flow of MB work soon.

Cheers.

Wow. Over 10 hours later and no one is wailing over the APs running out.

Well if we could burn thought the MB faster then more AP would get loaded quicker, yes?

Yes and I'm trying to do my part.

Cheers.

Well AP splitting is almost finished so get ready for a steady flow of MB work soon.

Cheers.

Wow. Over 10 hours later and no one is wailing over the APs running out.

Well if we could burn thought the MB faster then more AP would get loaded quicker, yes?

As I see it they load way too much data at once, but the real culprit is really the credit flaw.