S5GCE, was: Beyond S5R6

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 429731459
RAC: 79335

So, you are saying that all

So, you are saying that all the ~300 000 results in database (according to status page for now) are really sent out to crunchers, are you?

BTW, I tried, but it is unable to uncheck the box of S5GCE. Only ABP (what is it?) and ABP STSP are enabled.

P.S. I currently abort any ABP 3.08 and S5GCE 3.04 tasks if I see S5R6 in queue just not to wait for deadline and to be ready to receive as many as possible for data files in host for these WUs.

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3562358667
RAC: 0

On the ~41 thousand without

On the ~41 thousand without final result are still out for crunching, or have recently timed out. At this point a large chunk are probably assigned to inactive hosts. The boinc scheduler doesn't immediately send them back out to the first host available after a user doesn't return within the deadline. It waits a bit to see if a host that already has the needed data files connects to save on bandwidth.

PS Instead of aborting WU's just suspend them temporarily. That reduces the load on the servers and network bandwidth sending data files out again for the ones you aborted.

Ver Greeneyes
Ver Greeneyes
Joined: 26 Mar 09
Posts: 140
Credit: 9562235
RAC: 0

Yeah, everything has been

Message 96845 in response to message 96844

Yeah, everything has been sent out. I can't tell you why the already returned results are still in the database - maybe they need to be combined with results which haven't been returned yet to be useful, but that's just guessing. Aborting WUs does technically give you a somewhat higher chance of getting S5R6 WUs if any are being sent out, but I doubt the Einstein@Home scientists will appreciate it.

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6588
Credit: 316693621
RAC: 354453

RE: Yeah, everything has

Message 96846 in response to message 96845

Quote:
Yeah, everything has been sent out. I can't tell you why the already returned results are still in the database - maybe they need to be combined with results which haven't been returned yet to be useful, but that's just guessing. Aborting WUs does technically give you a somewhat higher chance of getting S5R6 WUs if any are being sent out, but I doubt the Einstein@Home scientists will appreciate it.


Forgive me if I'm in error, but I'd thought that aborts changed the daily WU quota downwards for a given box [ previously discussed with respect to ABP's ]. So given that : I suppose you do get a higher chance of the desired slice, but of a smaller pie.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

DanNeely
DanNeely
Joined: 4 Sep 05
Posts: 1364
Credit: 3562358667
RAC: 0

RE: In the past there

Message 96847 in response to message 96831

Quote:
In the past there have been occasions where the outstanding results were cleared quickly and that was done by increasing the initial replication so that a whole bunch of extra copies (maybe 10 or more) of the outstanding tasks were sent out. That way at least one host was likely to send back a result quickly but this is rather wasteful of resources.
.

IIRC the last time they needed to wrap up the odds and ends they just sent all the missing WU's to their own cluster.

Ver Greeneyes
Ver Greeneyes
Joined: 26 Mar 09
Posts: 140
Credit: 9562235
RAC: 0

RE: Forgive me if I'm in

Message 96848 in response to message 96846

Quote:
Forgive me if I'm in error, but I'd thought that aborts changed the daily WU quota downwards for a given box [ previously discussed with respect to ABP's ]. So given that : I suppose you do get a higher chance of the desired slice, but of a smaller pie.


That would be a good way to discourage such behavior, so it sounds likely - you'd know better than I would!

Mike Hewson
Mike Hewson
Moderator
Joined: 1 Dec 05
Posts: 6588
Credit: 316693621
RAC: 354453

RE: That would be a good

Message 96849 in response to message 96848

Quote:
That would be a good way to discourage such behavior, so it sounds likely - you'd know better than I would!


I'm quoting second hand knowledge here. It was brought up earlier in the context of AGW aborts for those who wanted to do ABP's only - as one is not otherwise allowed to directly opt out of being served with AGW WU's. I'm assuming that logic applies here. I suppose any abort ( of whatever WU type ) reduces the perceived reliability of the box in question.

Cheers, Mike.

I have made this letter longer than usual because I lack the time to make it shorter ...

... and my other CPU is a Ryzen 5950X :-) Blaise Pascal

Ver Greeneyes
Ver Greeneyes
Joined: 26 Mar 09
Posts: 140
Credit: 9562235
RAC: 0

Speaking of stragglers - the

Message 96850 in response to message 96849

Speaking of stragglers - the last ABP1 WU is finally in! Throw a party!

Stranger7777
Stranger7777
Joined: 17 Mar 05
Posts: 436
Credit: 429731459
RAC: 79335

Oh! That's great! Let's

Oh! That's great! Let's finish S5R6 ASAP.

I found that when I abort some tasks before they start it doesn't reduce my daily quota, because it is not a computation error. So I have daily a small but stable possibility to choose what to crunch. :)

DJStarfox
DJStarfox
Joined: 25 Mar 07
Posts: 10
Credit: 2484242
RAC: 0

RE: Since the only known

Message 96852 in response to message 96840

Quote:
Since the only known pulsar close enough that it might be detected, the crab nebula pulsar, is sitting right on the edge of being detectable, this means that GCE will be much more likely to give a positive result. (I don't have a citation for this, it was from a few posters from a conference that Bernd(Bruce???) linked to a few years back.)

You would think they would do that. Looking for something that you know is there in the sky is a good test to make sure the computations you're doing are valid.

I would love to see a sky plot (with colors indicating intensity/signal strength) of the results from S5R6. Something pretty to drive home the results of all the crunching. Any project sci? Mods?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.