I've noticed the different ratio of MB to GreenBank (blc) data too . I think it partially has to do with this file, 18dc09aa, being stuck. It is taking around 8 of the 11 channels in progress and nothing is coming out. Those 8 channels aren't doing anything. So there are only 3 channels working on splitting MB data.

No Panic yet, but could be a problem. The stuck Aricebo file needs to be removed.

It is nice to have two sets of data being split as if one goes into a shorty storm, then hopefully the other dataset helps stabilize the system.

It looks like they gave us a small amount of Green Bank data. blc41_2bit_guppi_58543_...
or were those files there before and I just didn't notice?

No Panic yet, but could be a problem. The stuck Aricebo file needs to be removed.
It is nice to have two sets of data being split as if one goes into a shorty storm, then hopefully the other dataset helps stabilize the system.
It looks like they gave us a small amount of Green Bank data. blc41_2bit_guppi_58543_...
or were those files there before and I just didn't notice?

. . No, they are definitely new, and they are relatively recent data from 1st March 2019 (at least they are more recent than 15th October 2018). I have never seen a 4 series blc task before so they may have commissioned another bank of recorders at GBT. I expect they will take about the same time as a blc33 task. Personally I will be more excited though when I see a new server called amigos and tasks with a Parkes identifier in their name :)

Stephen

:)

The new Green Bank data is dropping my RAC quite nicely, no reason to panic because IMO it is generally better data because it is targeted rather than the random data from Arecibo telescope.

---

. . Be careful what you wish for ....

. . I am now getting a big flow of Arecibo tasks, but they are all .VLAR WUs. That will definitely slow things down ...

Stephen

:(

https://setiathome.berkeley.edu/results.php?hostid=8572161

. . . . . .


i don't expect anything about this host, but how can the server can let it downloading all theses WU ?

---

Wow, I wonder if that is a record for most tasks sent incorrectly by the schedulers. Definitely broken code.

But there is something else strange about the host. It has two timed out tasks that apparently only had a 3 week deadline.

Normally our MB tasks have had a 6 to 7 week deadline. That made me look at my tasks and I see now that all my tasks have a much shorter 3 week deadline too.

So when did this change occur? Does everybody else have a much shorter task deadline in their cache also?

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Those shortie WUs have always had a 3 week deadline.

But I am seeing all my tasks with a 3 week deadline. Including BLC VLARs which have never been considered shorties. I call "shorties" the Arecibo VHAR tasks with AR greater than the standard 0.40- 0.44 AR range. Some of them lately have been in the .7-1.1 range and they process very fast of course. But the BLC tasks have all been normal 6-7 week deadlines.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Except for the shorties, most of my tasks on all 3 hosts have a deadline of august 6/7.

I am also seeing a lot of various short deadlines, both July 5 and July 25. There are a lot of the normal Aug 6/7 as well.

---

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours

But that is exactly what I am talking about. All my 27,700 tasks issued Jun 13-14 have a deadline date of the first week of August. So just about exactly a 3 week deadline.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

It's not universal. Across my 7 crunchers, I have a block of 4/5 July deadlines (3 weeks, certainly Arecibo, presumed VHAR, but not inspected internally). A few middling Arecibos, but every VLAR - both Arecibo and BLC - has a 6/7 August deadline as normal.

Edit

issued Jun 13-14 have a deadline date of the first week of August.

I think you've blanked July out of your calendar!

Ohh, darn. I guess I did forget about the month of July. Still frazzled in my thinking I guess. Sorry about the interruption . . . . nothing to see here.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

Wow, I wonder if that is a record for most tasks sent incorrectly by the schedulers. Definitely broken code.
But there is something else strange about the host. It has two timed out tasks that apparently only had a 3 week deadline.
Normally our MB tasks have had a 6 to 7 week deadline. That made me look at my tasks and I see now that all my tasks have a much shorter 3 week deadline too.
So when did this change occur? Does everybody else have a much shorter task deadline in their cache also?

. . I don't normally check deadlines, I have that column disabled in my manager's task screen to make room. But I just did a random survey of more than half a dozen tasks and they are all normal deadlines (7th August).

{edit} . . OK, making a mental note to send Keith a new calendar ...

Stephen

??

I've noticed the different ratio of MB to GreenBank (blc) data too . I think it partially has to do with this file, 18dc09aa, being stuck. It is taking around 8 of the 11 channels in progress and nothing is coming out. Those 8 channels aren't doing anything. So there are only 3 channels working on splitting MB data.

Yes, that particular file/tape got stuck when it was being split 9 years ago too.
It's a bad one, evil even :-) They had to remove it back then, and I guess the same will happen now.
Let's just hope it doesn't hang around for as long this time.

See: 18DC09AA Has Been Stuck More Than A Month

No need for panic more a sense of relief the multibeam tape mentioned above has been removed from the queue. I am unsure if that had human intervention, if that did thank you to whoever moved it on. Apologies people I got it wrong

I am also looking forward to the day when we see the 180.04 gig tape start to get it processed I noticed the name ended in 13. I'm not superstitious by any means I wonder is this a sign

---

No need for panic more a sense of relief the multibeam tape mentioned above has been removed from the queue.

? ? ? ? I still see it loaded. It's been stuck for days.

---

Seti@Home classic workunits:20,676 CPU time:74,226 hours

A proud member of the OFA (Old Farts Association)

No need for panic more a sense of relief the multibeam tape mentioned above has been removed from the queue.

? ? ? ? I still see it loaded. It's been stuck for days.

Thanks Keith. I have adjusted my original post

---

18dc09aa is still hanging in there, still no work from it.

---

Grant
Darwin NT

@ Unixchick

. . Hey guess what? We are now crunching the last couple of 'groundhog day' tapes, and the other older GBT tapes in blc34 and blc22 are following. By morning here I expect to be crunching the newer blc41 tapes. After 8 months we are about to see the end of them ... YYYAAAAAAAYYYY!

. . All I need now is to see Parkes tapes begin to appear ...

Stephen

:)