S5GCE, was: Beyond S5R6

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5779100
RAC: 0

RE: I thought the CC would

Message 96803 in response to message 96800

Quote:
I thought the CC would auto-hop through the list of mirrors on its own after a certain number of failed attempts.


It does so, although it seems to rotate between 4 mirrors here only.

19-Mar-10 00:25:00 Einstein@Home Started download of h1_0096.80_S5R7
19-Mar-10 00:25:00 Einstein@Home [file_xfer_debug] URL: http://einstein.ligo.caltech.edu/download/1bb/h1_0096.80_S5R7
19-Mar-10 00:25:00 Einstein@Home Started download of h1_0096.85_S5R4
19-Mar-10 00:25:00 Einstein@Home [file_xfer_debug] URL: http://einstein.astro.gla.ac.uk/download/1a1/h1_0096.85_S5R4
19-Mar-10 00:25:22 Einstein@Home Started download of l1_0096.95_S5R7
19-Mar-10 00:25:22 Einstein@Home [file_xfer_debug] URL: http://einstein.aei.mpg.de/download/fe/l1_0096.95_S5R7
19-Mar-10 00:25:22 Einstein@Home Started download of h1_0097.00_S5R4
19-Mar-10 00:25:22 Einstein@Home [file_xfer_debug] URL: http://einstein.aei.mpg.de/download/344/h1_0097.00_S5R4
19-Mar-10 00:25:45 Einstein@Home Started download of h1_0097.05_S5R4
19-Mar-10 00:25:45 Einstein@Home [file_xfer_debug] URL: http://einstein-dl.phys.uwm.edu/download/1a/h1_0097.05_S5R4
19-Mar-10 00:25:57 Einstein@Home Started download of h1_0097.05_S5R7
19-Mar-10 00:25:57 Einstein@Home [file_xfer_debug] URL: http://einstein-dl.phys.uwm.edu/download/1af/h1_0097.05_S5R7
19-Mar-10 00:26:08 Einstein@Home Started download of h1_0096.85_S5R4
19-Mar-10 00:26:08 Einstein@Home [file_xfer_debug] URL: http://einstein-dl.phys.uwm.edu/download/1a1/h1_0096.85_S5R4
19-Mar-10 00:26:30 Einstein@Home Started download of l1_0096.95_S5R7
19-Mar-10 00:26:30 Einstein@Home [file_xfer_debug] URL: http://einstein.astro.gla.ac.uk/download/fe/l1_0096.95_S5R7
19-Mar-10 00:26:52 Einstein@Home Started download of h1_0096.85_S5R4
19-Mar-10 00:26:52 Einstein@Home [file_xfer_debug] URL: http://einstein.ligo.caltech.edu/download/1a1/h1_0096.85_S5R4
19-Mar-10 00:26:52 Einstein@Home Started download of h1_0097.00_S5R4
19-Mar-10 00:26:52 Einstein@Home [file_xfer_debug] URL: http://einstein.astro.gla.ac.uk/download/344/h1_0097.00_S5R4

I'll leave it be. Eventually it'll get the stuff in, I'm sure of that. :)

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5779100
RAC: 0

Patience does it.

Patience does it. :-)

19-Mar-10 00:45:40 Einstein@Home Started download of h1_0096.85_S5R4
19-Mar-10 00:45:40 Einstein@Home [file_xfer_debug] URL: http://einstein-dl.phys.uwm.edu/download/1a1/h1_0096.85_S5R4
19-Mar-10 00:46:45 Einstein@Home Finished download of h1_0096.85_S5R4
19-Mar-10 00:49:02 Einstein@Home Started download of h1_0096.80_S5R7
19-Mar-10 00:49:02 Einstein@Home [file_xfer_debug] URL: http://einstein-dl.phys.uwm.edu/download/1bb/h1_0096.80_S5R7
19-Mar-10 00:49:56 Einstein@Home Finished download of h1_0096.80_S5R7

Alinator
Alinator
Joined: 8 May 05
Posts: 927
Credit: 9352143
RAC: 0

LOL... Yep, the old EAH

LOL...

Yep, the old EAH locomotive finally built up a head of steam and I'm all caught up as of about an hour ago! :-)

Collatz even seems to have recouped from getting bombed by having MW crap out with no one around to give it kick back into life, despite it's very low budget/low staff stature. :-D

Alinator

David Hammer
David Hammer
Joined: 15 Oct 04
Posts: 360
Credit: 1672886
RAC: 0

RE: RE: I thought the CC

Message 96806 in response to message 96803

Quote:
Quote:
I thought the CC would auto-hop through the list of mirrors on its own after a certain number of failed attempts.

It does so, although it seems to rotate between 4 mirrors here only.

...

I'll leave it be. Eventually it'll get the stuff in, I'm sure of that. :)

The Hannover mirror also plays a role in the ABP search so we are not using it as a mirror right now. I changed the server_status page to no longer list this mirror. The other four mirrors are actually running but are listed as "not running" because they fail to respond when the mirror checking script tries to download a file.

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5779100
RAC: 0

Thanks for those

Message 96807 in response to message 96806

Thanks for those explanations, David. Don't you need more mirrors (and smokescreens)? :-)

Darren Peets
Darren Peets
Joined: 19 Nov 09
Posts: 37
Credit: 97999600
RAC: 30607

One thing that's a bit odd

One thing that's a bit odd about the new tasks: the progress indicator is non-monotonic (at least on my 6.4.5 Linux BOINC manager). For instance, one task just "increased" from 92.78% to 91.85%. I haven't noticed this sort of behaviour on earlier types of workunit.

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244931581
RAC: 16394

RE: One thing that's a bit

Message 96809 in response to message 96808

Quote:
One thing that's a bit odd about the new tasks: the progress indicator is non-monotonic (at least on my 6.4.5 Linux BOINC manager). For instance, one task just "increased" from 92.78% to 91.85%. I haven't noticed this sort of behaviour on earlier types of workunit.

Was the task suspended in between, i.e. recovered form a checkpoint?

BM

BM

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5779100
RAC: 0

RE: For instance, one task

Message 96810 in response to message 96808

Quote:
For instance, one task just "increased" from 92.78% to 91.85%.


For what's it worth, I've just seen the same behaviour on a Windows machine. Nothing about it in the stderr.txt though.

It happened between 2 checkpoints. You just have to see it, by chance.

Bernd Machenschalk
Bernd Machenschalk
Moderator
Administrator
Joined: 15 Oct 04
Posts: 4267
Credit: 244931581
RAC: 16394

Thanks for reporting. I'll

Thanks for reporting. I'll take a look. Shouldn't be a showstopper, though.

BM

BM

Jord
Joined: 26 Jan 05
Posts: 2952
Credit: 5779100
RAC: 0

No, it isn't. It's just a

Message 96812 in response to message 96811

No, it isn't. It's just a little quirk and as far as I have seen it only happens once. I only checked because I read Darren's post and saw I was at 92.111%, so sat here watching what it was going to do. :-)

I'm not going to check the last 48 minutes of its run time what other flaky percentages there are, promised. :-)

Comment viewing options

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