Just being interested in Albert: What exactly does Improved (all-sky pulsar search) mean? (I don't refer to the WUs) Or is this a secret?
Thx in advance
Read the 1st post in this thread, including the Q&A section. I've got a batch of Alberts that have 17% less estimated runtime than Einsteins. 1st one should be starting ~noon tommorrow if I've got the timing right (my cores are currently half a work unit out of sync, drifting closer/farther depending on the vagarities of which core windows decides to use for housekeeping tasks.)
I read the 1st post and most of the others below. But the most are about some less important things concerning to what the program does while calculating. Reduced time says absolutely nothing important about the internal structure of the program ... the only important thing is this one:
"The differences in run times come about because we are now using a sky search grid and frequency band which depends upon frequency"
... and this is just a brief description ... so, anyone with deeper insights? :)
(btw: I am not complaining! I just want to know the important differencies)
I'm wondering about exactly what the scheduler is doing re einstien/alberts myself.
The way I *thought* it was going to work was to give E's exclusively until it had enough people assigned to each chunk of data to have reasonable computation times for getting everything done. Once that was accomplished as all the work on each dataset was assigned it would start switching people over to A's, and excepting odds/ends due to incomplete/invalid results everyone would stay with A's once they got them.
What's being put into my queue indicates this isn't the case. I've used up almost all my old E's, and am starting on a 2day supply of A's, the 2 days of work at the tail end of my queue however are a set of 'virgin' E's, most of which it appears I'm the only person to've been assigned to do so far.
I read the 1st post and most of the others below. But the most are about some less important things concerning to what the program does while calculating. Reduced time says absolutely nothing important about the internal structure of the program ... the only important thing is this one:
"The differences in run times come about because we are now using a sky search grid and frequency band which depends upon frequency"
... and this is just a brief description ... so, anyone with deeper insights? :)
(btw: I am not complaining! I just want to know the important differencies)
Slavko.sk
I see you have changed the timezone on this computer and it looks in your scheduler logs that you have the UK site on top of that list.I guess those downloads went well.
But this computer have the mid european timezon and the german server on top.
Try change this timezone too.
I guess its ok to set the timezones back to normal when you have done the downloads and start crunching.(untill the next big datablock is needed??)
Did you actaully read that link, or just conjur it out of google? IF the former, where should we be looking in it? The phrases "all-sky" and "all sky" are not in the document anywhere, and every instance of improved appears to either refer directly to sensor hardware, or other backend infrastructure; not the client app.
PS, does this board automatically tinyurl any link? The read view of the forum is showing a tinyurl that points to the tinyurl you posted, while the edit page is showing what I assume is your orginal tinyurl. I'm directly pasting the link from google's cache to see what happens.
PS, does this board automatically tinyurl any link? The read view of the forum is showing a tinyurl that points to the tinyurl you posted, while the edit page is showing what I assume is your orginal tinyurl. I'm directly pasting the link from google's cache to see what happens.
Hmmm.
Did you edit your post, JAHMAGIC? The link that showed *after* I made my reply isn't the same as the one I got backing in my browser.
Just to report back... Today I try to run E@H at my office machine... but it produce lots of client errors. It didn't even try to work with Albert but it tried to download to Einstein app and files. I get error 504 message, and signature error when downloading files. And this machine work perfectly before it was hit by Albert. Now I just can't make it to work. My home machine work nicely on Albert. The only difference is that at office I'm behind proxy. But that wasn't problem before. Hope staff will fix this nightmare soon (if this is problem on their part).
I'm still getting
)
I'm still getting errors:
2.1.2006 1:22:30|Einstein@Home|Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
2.1.2006 1:22:32|Einstein@Home|Started download of l1_1167.5
2.1.2006 1:22:58|Einstein@Home|Temporarily failed download of l1_1167.5: error 504
2.1.2006 1:22:59|Einstein@Home|Started download of l1_1167.5
2.1.2006 1:23:59|Einstein@Home|Finished download of l1_1167.5
2.1.2006 1:23:59|Einstein@Home|Throughput 108515 bytes/sec
2.1.2006 1:23:59|Einstein@Home|MD5 check failed for l1_1167.5
2.1.2006 1:23:59|Einstein@Home|expected 5aa89df74b1cdb44a5e44d69b974299f, got 83bb2e1ddb98aba4db62b7b48bb869b5
2.1.2006 1:23:59|Einstein@Home|Checksum or signature error for l1_1167.5
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.7_0.1_T08_S4lD_3 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.8_0.1_T08_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.9_0.1_T08_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.5_0.1_T09_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.6_0.1_T09_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.7_0.1_T09_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.8_0.1_T09_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.9_0.1_T09_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.5_0.1_T10_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.6_0.1_T10_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.7_0.1_T10_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.8_0.1_T10_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.9_0.1_T10_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.5_0.1_T11_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.6_0.1_T11_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
2.1.2006 1:24:00|Einstein@Home|Unrecoverable error for result l1_1167.5__1167.7_0.1_T11_S4lD_0 (WU download error: couldn't get input files: l1_1167.5 -119 MD5 check failed)
ALL GLORY TO THE HYPNOTOAD!
Do You Dare?
Potrebujete pomoc?
The Albert 4.37's are working
)
The Albert 4.37's are working perfect for me so far and I have done lots of them.
My only problem is the server seems to have picked my slowest pc and gives me Alberts all the time but my faster machines still get the other version.
Not sure why that is.
But I just sent another in (5.5hr average)
This time it loaded 4 more Alberts but this time it added a "download of skygrid_0500_r_T08.dat"
First time I have seen that.
Hopefully Alberts will start jumping on my other faster machines soon.
Happy New Year
RE: RE: Just being
)
I read the 1st post and most of the others below. But the most are about some less important things concerning to what the program does while calculating. Reduced time says absolutely nothing important about the internal structure of the program ... the only important thing is this one:
"The differences in run times come about because we are now using a sky search grid and frequency band which depends upon frequency"
... and this is just a brief description ... so, anyone with deeper insights? :)
(btw: I am not complaining! I just want to know the important differencies)
Greetings, Santas little helper
I'm wondering about exactly
)
I'm wondering about exactly what the scheduler is doing re einstien/alberts myself.
The way I *thought* it was going to work was to give E's exclusively until it had enough people assigned to each chunk of data to have reasonable computation times for getting everything done. Once that was accomplished as all the work on each dataset was assigned it would start switching people over to A's, and excepting odds/ends due to incomplete/invalid results everyone would stay with A's once they got them.
What's being put into my queue indicates this isn't the case. I've used up almost all my old E's, and am starting on a 2day supply of A's, the 2 days of work at the tail end of my queue however are a set of 'virgin' E's, most of which it appears I'm the only person to've been assigned to do so far.
RE: I read the 1st post and
)
http://tinyurl.com/8tmbt
Slavko.sk I see you have
)
Slavko.sk
I see you have changed the timezone on this computer and it looks in your scheduler logs that you have the UK site on top of that list.I guess those downloads went well.
But this computer have the mid european timezon and the german server on top.
Try change this timezone too.
I guess its ok to set the timezones back to normal when you have done the downloads and start crunching.(untill the next big datablock is needed??)
RE: http://tinyurl.com/8tmb
)
Did you actaully read that link, or just conjur it out of google? IF the former, where should we be looking in it? The phrases "all-sky" and "all sky" are not in the document anywhere, and every instance of improved appears to either refer directly to sensor hardware, or other backend infrastructure; not the client app.
PS, does this board automatically tinyurl any link? The read view of the forum is showing a tinyurl that points to the tinyurl you posted, while the edit page is showing what I assume is your orginal tinyurl. I'm directly pasting the link from google's cache to see what happens.
http://66.102.7.104/search?q=cache:by7iLtzNjZIJ:www.ligo.caltech.edu/NSF/pdf/annual_report.pdf+What+exactly+does+Improved+(all-sky+pulsar+search)+mean%3F+2006&hl=en
RE: PS, does this board
)
Hmmm.
Did you edit your post, JAHMAGIC? The link that showed *after* I made my reply isn't the same as the one I got backing in my browser.
I'm seeing some Albert
)
I'm seeing some Albert ugliness on my Mac:
5.2.13
2006-01-02 00:40:57.9653 [normal]: Start of BOINC application 'albert_4.39_powerpc-apple-darwin'.
MacOS Error -43 occured in Mac_Lib.c line 65
MacOS Error -43 occured in Mac_Lib.c line 65
2006-01-02 00:40:57.9986 [normal]: Started search at lalDebugLevel = 0
2006-01-02 00:40:59.1248 [normal]: Checkpoint-file 'Fstat.out.ckp' not found.
2006-01-02 00:40:59.1253 [normal]: No usable checkpoint found, starting from beginning.
Detected CPU type 1
2006-01-02 00:52:45.2383 [normal]: Fstat file reached MaxFileSizeKB ==> compactifying ... done.
2006-01-02 05:06:06.2487 [normal]: Search finished successfully.
And on my older P3:
5.2.8
2005-12-31 12:45:36.1250 [normal]: Start of BOINC application 'projects/einstein.phys.uwm.edu/albert_4.37_windows_intelx86.exe'.
2005-12-31 12:45:36.1250 [normal]: Started search at lalDebugLevel = 0
2005-12-31 12:45:36.8125 [normal]: Checkpoint-file 'Fstat.out.ckp' not found.
2005-12-31 12:45:36.8125 [normal]: No usable checkpoint found, starting from beginning.
2005-12-31 12:50:57.9843 [normal]: Fstat file reached MaxFileSizeKB ==> compactifying ... done.
2005-12-31 16:24:34.0937 [normal]: Search finished successfully.
Just to report back... Today
)
Just to report back... Today I try to run E@H at my office machine... but it produce lots of client errors. It didn't even try to work with Albert but it tried to download to Einstein app and files. I get error 504 message, and signature error when downloading files. And this machine work perfectly before it was hit by Albert. Now I just can't make it to work. My home machine work nicely on Albert. The only difference is that at office I'm behind proxy. But that wasn't problem before. Hope staff will fix this nightmare soon (if this is problem on their part).