Work for AP is still a no go.
Very frustrating to see that channels are being done but no work is being sent.

So much for this run

Michael

13 AP + 13 MB channels in progress Ops, we still have a problem.

---

Every channel is ending in error

84 AP so far
Nuts

Every channel is ending in error

84 AP so far
Nuts

+1

---

Dave

Every channel is ending in error

84 AP so far
Nuts

Now over 100

---

Dave

Watching ssp for a while, it appears that all AP channels are ending in error.

If wu cannot be sent because of db off line, why split to error?

AFAIK,
the AstroPulse database could be offline during splitters are running.
The database just safe the results of the members PCs.
The AstroPulse channel errors are receiving errors from the telescope (interferences).

---

The query to the database was changed to order the results by sent time rather than ID. That will keep all tasks assigned for a single request together, but how MySQL should order them within a group sent at the same time is not specified. If it turns out to be a problem I expect Dr. Anderson would add an additional term to the query.

                                                                   Joe

That change was made after a discussion that I started with one or two others with an interest in task results. The divergence between a list of results sorted by ID and a list sorted by date is much greater at other BOINC projects - Einstein was the example in question.

It would be ideal (IMHO) if we could choose our own display order - name, ID, date, etc. - independently of the format of the display (name or ID). It's the sort of thing which BOINCstats makes look very easy.

David indicated that he wouldn't have time to make more complicated changes himself, but would be happy to accept code contributions. Any PHP/MySQL coders out there who could make some queries and page controls?

That's a disturbing situation, indicating that California might have slipped below the internationally accepted extreme poverty line of one US dollar (US$1) per day. "Please send patches, hungry"

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

I'm not sure why but I did get 1 AP task at 2.00 am and 2 AP tasks at 9:30 am
BC Canada time

The Astropulse Science DB was still disabled at the time of the downloads.

Thank god I got 120 tasks Friday.
As it is I can crunch AP for 2 days max then totally dry. About 40 hours
Right now I have 74 total and 29 of those are 601 CPU tasks

Please turn on the Science database so we can get our cache filled if that is the only problem.
Are the tapes that were loaded any good?
Seems like a awful waste of data


Michael Miles

I'm not sure why but I did get 1 AP task at 2.00 am and 2 AP tasks at 9:30 am
BC Canada time

The Astropulse Science DB was still disabled at the time of the downloads.

Thank god I got 120 tasks Friday.
As it is I can crunch AP for 2 days max then totally dry. About 40 hours
Right now I have 74 total and 29 of those are 601 CPU tasks

Please turn on the Science database so we can get our cache filled if that is the only problem.
Are the tapes that were loaded any good?
Or are they just being allowed to split to error on purpose. Seems like a awful waste of data if that is the case.


Michael Miles

Resends most likely I guess.

Cheers.

That is what I guess too. It was exciting though. LOL

Please turn on the Science database so we can get our cache filled if that is the only problem.

That the science database is down doesn't really matter to us crunchers, it just means there is a backlog of completed AP Workunits waiting for assimilation, and it'll grow as more results are validated.

Claggy

That the science database is down doesn't really matter to us crunchers, it just means there is a backlog of completed AP Workunits waiting for assimilation, and it'll grow as more results are validated.

Claggy

Just curious as to why no work is reaching us when so many are being split and channels ending in error?

Something going on with the SSP. Has someone entered the building?

                            Progress
 	             Multibeam  Astropulse  

total channels to do:	462	   442

Just curious as to why no work is reaching us when so many are being split and channels ending in error?

The majority of the work going out is going to be Seti v7, that's what's got the high result creation rate.

Claggy

For whatever reason, splitters still struggling to meet demand, let alone build up a ready-to-send buffer.

---

Grant
Darwin NT

Still limbo in Server land I see this morning : I actually got 2 tasks in the last 48 hours...

---

SSP indicates the MB splitters has 13 channels in progress (creation rate 25/sec( as the AP too (creation rate close to 0). I never see somethign like that before (normaly few MB (maybe 6 or 7) and few AP chanels in progress only). This could indicate aproblem? That starts after the last JBOD crash.

---

SSP indicates the MB splitters has 13 channels in progress (creation rate 25/sec( as the AP too (creation rate close to 0). I never see somethign like that before (normaly few MB (maybe 6 or 7) and few AP chanels in progress only). This could indicate aproblem? That starts after the last JBOD crash.

I'd take those "13 channels in progress" with a pinch of salt.

My working guess (but without actual knowledge) is that when there is a crash/restart, the splitters don't necessarily go back to working on the same tapes that were 'in progress' before the crash. So, the new tapes are genuinely 'running', but the old tapes are maybe 'waiting to run', as we see in our BOINC Managers for a task which has been started but then paused for some reason. Unfortunately, the SSP doesn't give us that fine level of detail, and it doesn't change quickly enough to make it obvious which tapes are moving and which are waiting.

Let´s see is your guess is right and wath we get in the next days. tks.

Could be an AP tape stucked? The AP creation rate is close to zero.

---

Let´s see is your guess is right and wath we get in the next days. tks.

Could be an AP tape stucked? The AP creation rate is close to zero.

No AP splitters are running (likely due to the AP database being disabled) and what you see in the creation rate for AP's is just resends going out.

Cheers.