Dowload errors
3 Jan 2006 12:39:08 UTC
Topic 190505
(moderation:
Last few days I'm getting a lot of DL errors. Mostly they look like this:
tcpu Einstein@Home 3.1.2006 1:32:12 1438 Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
tcpu Einstein@Home 3.1.2006 1:32:12 1439 Reason: To fetch work
tcpu Einstein@Home 3.1.2006 1:32:12 1440 Requesting 86400 seconds of new work
tcpu Einstein@Home 3.1.2006 1:32:22 1441 Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
tcpu Einstein@Home 3.1.2006 1:32:24 1442 Started download of albert_4.37_windows_intelx86.pdb
tcpu Einstein@Home 3.1.2006 1:32:24 1443 Started download of skygrid_0700_r_T09.dat
tcpu Einstein@Home 3.1.2006 1:32:47 1444 Temporarily failed download of albert_4.37_windows_intelx86.pdb: error 500
tcpu Einstein@Home 3.1.2006 1:32:47 1445 Temporarily failed download of skygrid_0700_r_T09.dat: error 500
tcpu Einstein@Home 3.1.2006 1:32:48 1446 Started download of albert_4.37_windows_intelx86.pdb
tcpu Einstein@Home 3.1.2006 1:32:48 1447 Started download of skygrid_0700_r_T09.dat
tcpu Einstein@Home 3.1.2006 1:32:52 1448 Finished download of skygrid_0700_r_T09.dat
tcpu Einstein@Home 3.1.2006 1:32:52 1449 Throughput 50639 bytes/sec
tcpu Einstein@Home 3.1.2006 1:32:52 1450 MD5 check failed for skygrid_0700_r_T09.dat
tcpu Einstein@Home 3.1.2006 1:32:52 1451 expected f6021cecf6af92e39da4f6a6f39991e3, got d4cf13ad926d41a106305165f7c9783b
tcpu Einstein@Home 3.1.2006 1:32:52 1452 Checksum or signature error for skygrid_0700_r_T09.dat
tcpu Einstein@Home 3.1.2006 1:32:53 1453 Unrecoverable error for result r1_0696.0__314_S4R2a_1 (WU download error: couldn't get input files: skygrid_0700_r_T09.dat -119 MD5 check failed)
tcpu Einstein@Home 3.1.2006 1:32:53 1454 Unrecoverable error for result r1_0696.0__313_S4R2a_0 (WU download error: couldn't get input files: skygrid_0700_r_T09.dat -119 MD5 check failed)
tcpu Einstein@Home 3.1.2006 1:33:04 1455 Finished download of albert_4.37_windows_intelx86.pdb
tcpu Einstein@Home 3.1.2006 1:33:04 1456 Throughput 226699 bytes/sec
tcpu Einstein@Home 3.1.2006 1:33:04 1457 signature verification failed for albert_4.37_windows_intelx86.pdb
tcpu Einstein@Home 3.1.2006 1:33:04 1458 Checksum or signature error for albert_4.37_windows_intelx86.pdb
tcpu Einstein@Home 3.1.2006 1:33:58 1459 Sending scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi
tcpu Einstein@Home 3.1.2006 1:33:58 1460 Reason: To fetch work
tcpu Einstein@Home 3.1.2006 1:33:58 1461 Requesting 86400 seconds of new work, and reporting 2 results
tcpu Einstein@Home 3.1.2006 1:34:03 1462 Scheduler request to http://einstein.phys.uwm.edu/EinsteinAtHome_cgi/cgi succeeded
tcpu Einstein@Home 3.1.2006 1:34:03 1463 Message from server: No work sent
tcpu Einstein@Home 3.1.2006 1:34:03 1464 Message from server: (reached daily quota of 2 results)
tcpu Einstein@Home 3.1.2006 1:34:03 1465 No work from project
Sometimes connection to DL server (in my case it's almost allways einstein.aei.mpg.de) fails.
I have a number of hosts connected to the same LAN and not all of them have the same problems.
Anybody else with similar problems?
Metod ... 
Dowload errors
)
I have similar problem with two of my machines. I just couldn't make them work on E@H since they were hit by Albert WUs for fist time. My third machine works fine and crunch both Albert and Einstein WUs.
You can read logs of similar errors here and here.
Edo
Finally, I manage to make my
)
Finally, I manage to make my office machines run as well. I change my time zone as it seems that all who have problem with Albert are from same time zone and are getting work from german server. I switch my time zone to EST (US) and my office machine downloaded Albert app and all the files without any problems.
Edo
What about those of us that
)
What about those of us that are Eastern Time zone and cannot d/l? Mine at home has yet to attempt the download, but it hasn't had problems on anything else, the one at work tho...
http://einsteinathome.org/host/431031
"Fiction reveals truth that reality obscures."
-Jessamyn West
RE: What about those of us
)
You are the second guy in this timezone who have these problems.
Everybody else is in Europe.
Your primary download server is either einstein.aset.psu.edu or
morel.mit.edu.(they have the same timezone)
Is this thing spreading or what?
or a longshot.. you couldn't connect to any of the servers and manage to hook up to the *sick* german server.. :P (in that case you should see errormessages like "Couldn't connect to hostname[ bla.bla...]" in Boinc Manager.)(or in stderrdae.txt)
Oh,your AMD has not started on Alberts yet.. That's why it's still working.
RE: Last few days I'm
)
Einstein@Home has a total of five data servers. If a download from the 'closest' one fails, then it should try from all the other data servers (twice through the list) before failing. So the fact that your downloads fail without using multiple URLs seems suspicious. Is your machine behind a proxy server?
Director, Einstein@Home
RE: Einstein@Home has a
)
Machines that I have problems with are behind proxy server. My home machine (that is not behind proxy server) never experience same problems, and I never changed time zone on my home machine.
Edo
RE: Einstein@Home has a
)
I had the same problem. Download failed with "signature verification error". A download from another download server did not take place. I'am behind a NAT router, but never had any such problems. The next scheduler request (today) triggered the download anew, successfully this time. I reported it here.
Maybe this is of interest to you.
Michael
Team Linux Users Everywhere
Just for info, from the
)
Just for info, from the developer mailing list - a bug was found in BOINC where it would not correctly go to the next download server in the list if the first one failed in certain ways, and a fix is being put in.
Bill, thanks for this info.
)
Bill, thanks for this info. So, is there any idea when this fix will came into effect? Or is it already everything fixed and we can turn back our time zones? Thanks.
Edo
RE: Just for info, from the
)
Please give a reference URL for this. I saw some traffic about a 'suspected bug' but not something about a 'confirmed, found bug'. Where is the email about a fix being put in?
Director, Einstein@Home