Advanced search

Message boards : Server and website : Download stalled

Author Message
nanoprobe
Send message
Joined: 26 Feb 12
Posts: 181
Credit: 222,317,747
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 40929 - Posted: 23 Apr 2015 | 22:28:26 UTC

Title says it all. Here's the error message from the event viewer.

5689 4/23/2015 6:27:09 PM [error] Can't create HTTP response output file projects/www.gpugrid.net


Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 912
Credit: 2,199,256,745
RAC: 781,265
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 40930 - Posted: 23 Apr 2015 | 22:54:17 UTC - in response to Message 40929.

Says it all? Doesn't tell me whether you're running the sort of operating system that might tempt you to install the sort of antivirus program that might block a download.

nanoprobe
Send message
Joined: 26 Feb 12
Posts: 181
Credit: 222,317,747
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 40931 - Posted: 23 Apr 2015 | 23:13:28 UTC
Last modified: 23 Apr 2015 | 23:13:58 UTC

XP. No A/V installed. This machine has been running fine for a long time. First time I've ever seen a stalled download from GPUGrid.

nanoprobe
Send message
Joined: 26 Feb 12
Posts: 181
Credit: 222,317,747
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 40958 - Posted: 26 Apr 2015 | 12:47:26 UTC
Last modified: 26 Apr 2015 | 12:48:53 UTC

Happened again this morning. When I tried to abort the stalled task BOINC crashed. I have no idea what's going on. This machine has never had a single issue running GPUGrid before this started. It's having no issues with WCG or Einstein. Anyone have any suggestions?

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 912
Credit: 2,199,256,745
RAC: 781,265
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 40959 - Posted: 26 Apr 2015 | 13:56:38 UTC - in response to Message 40958.

Happened again this morning. When I tried to abort the stalled task BOINC crashed. I have no idea what's going on. This machine has never had a single issue running GPUGrid before this started. It's having no issues with WCG or Einstein. Anyone have any suggestions?

I'd suggest a reboot and hard disk check.

nanoprobe
Send message
Joined: 26 Feb 12
Posts: 181
Credit: 222,317,747
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 40960 - Posted: 26 Apr 2015 | 15:29:37 UTC - in response to Message 40959.
Last modified: 26 Apr 2015 | 15:31:56 UTC

Happened again this morning. When I tried to abort the stalled task BOINC crashed. I have no idea what's going on. This machine has never had a single issue running GPUGrid before this started. It's having no issues with WCG or Einstein. Anyone have any suggestions?

I'd suggest a reboot and hard disk check.

Did a reboot. HDD is an enterprise grade Intel SSD that automatically remaps worn bits using wear leveling technology. It's only GPUGrid tasks that are giving me problems. The stuck downloads have been aborted but the event log says there are downloads still stuck so I'm guessing that's coming from the server side even though the problem tasks show as aborted in my account. I don't think there is anything else I can do but wait and see if this clears up.

EDIT: Just downloaded 2 new tasks without issues.

Richard Haselgrove
Send message
Joined: 11 Jul 09
Posts: 912
Credit: 2,199,256,745
RAC: 781,265
Level
Phe
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 40961 - Posted: 26 Apr 2015 | 15:47:02 UTC - in response to Message 40960.

The 'aborted' tasks have very strange workunit names:

Error tasks for host 190565

Other people have had problems with them too, but a couple have succeeded with 64-bit Windows hosts.

Profile Retvari Zoltan
Avatar
Send message
Joined: 20 Jan 09
Posts: 2048
Credit: 14,835,541,419
RAC: 2,690,989
Level
Trp
Scientific publications
watwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwatwat
Message 40963 - Posted: 26 Apr 2015 | 18:05:30 UTC - in response to Message 40961.
Last modified: 26 Apr 2015 | 18:05:59 UTC

Richard Haselgrove wrote:
The 'aborted' tasks have very strange workunit names:

Error tasks for host 190565

Other people have had problems with them too, but a couple have succeeded with 64-bit Windows hosts.

These are really strange names:
e56s11_e54s2_e52s3_e44s12_e43s7_e34s7_e25s8_e20s5_e19s2_e17s12_e16s10_e13s8_e11s13_e10s8_e8s8_e2s9_e1s15_7f298f162f388f357f360f382f0f459f292f273f425f492f84f153f346f11-SDOERR_villinpubKc3-0-1-RND4199_4
e56s16_e54s2_e52s3_e44s12_e43s7_e34s7_e25s8_e20s5_e19s2_e17s12_e16s10_e13s8_e11s13_e10s8_e8s8_e2s9_e1s15_7f298f162f388f357f360f382f0f459f292f273f425f492f84f153f346f198-SDOERR_villinpubKc3-0-1-RND2349_5
e56s10_e52s3_e44s12_e43s7_e34s7_e25s8_e20s5_e19s2_e17s12_e16s10_e13s8_e11s13_e10s8_e8s8_e2s9_e1s15_7f298f162f388f357f360f382f0f459f292f273f425f492f84f153f450-SDOERR_villinpubKc3-0-1-RND9134_0


nanoprobe wrote:
Did a reboot. HDD is an enterprise grade Intel SSD that automatically remaps worn bits using wear leveling technology.

This is a basic feature of all state of the art SSDs (it's called by different names like "garbage collection"). "Enterprise grade" usually refers to the larger number of the write cycles could be done on a single cell of the SSD, or the larger number of the (hidden) spare cells (so called over provisioning). It has nothing to do with the errors of the file system on the SSD, so I still recommend you to check it on your disk: from command line: chkdsk /f, or right click on the drive -> properties -> check now -> "automatically check disk errors" -> restart (and don't press any keys for 10 seconds at OS startup).
Maybe it won't help, as I think that the root of this error is not in your PC, but in your internet connection. Probably the TCP/IP packets gets fragmented. It's usually caused by improper setting of the maximum transmission unit (MTU) size (it depends on the WAN network you use, it's usually between 1492 and 1500 bytes). If it's a networking error, then it could be in the hands of a 3rd party.

nanoprobe
Send message
Joined: 26 Feb 12
Posts: 181
Credit: 222,317,747
RAC: 0
Level
Leu
Scientific publications
watwatwatwatwatwatwatwatwatwatwat
Message 40964 - Posted: 26 Apr 2015 | 21:28:29 UTC
Last modified: 26 Apr 2015 | 21:29:15 UTC

It has nothing to do with the errors of the file system on the SSD, so I still recommend you to check it on your disk: from command line: chkdsk /f, or right click on the drive -> properties -> check now -> "automatically check disk errors" -> restart (and don't press any keys for 10 seconds at OS startup).
Maybe it won't help, as I think that the root of this error is not in your PC, but in your internet connection. Probably the TCP/IP packets gets fragmented. It's usually caused by improper setting of the maximum transmission unit (MTU) size (it depends on the WAN network you use, it's usually between 1492 and 1500 bytes). If it's a networking error, then it could be in the hands of a 3rd party.


I may run a chkdsk at a later time but not now. I've done some ping tests and my internet connection seems to be fine. Since GPUGrid tasks are the only ones giving me issues at this time so I'm going to leave things as they are for now. I just find it very strange that this problem started for no apparent reason and only from this site.
Thanks to everyone for your help and suggestions. I'll keep and eye on the problem and post back later if things go good or bad.

Profile caffeineyellow5
Avatar
Send message
Joined: 30 Jul 14
Posts: 225
Credit: 2,658,976,345
RAC: 0
Level
Phe
Scientific publications
watwatwatwatwatwatwatwat
Message 40970 - Posted: 28 Apr 2015 | 0:26:30 UTC - in response to Message 40929.

XP. No A/V installed.

Sounds like an older machine, so I assume it is set to do the short tasks. I had the error on my "short only" machine.
https://www.gpugrid.net/result.php?resultid=14140396
Other people have had problems with them too, but a couple have succeeded with 64-bit Windows hosts.

My "short only" machine is a 64-bit Win8 with an A/V that is pretty much pretty passive for DLs and lets them finish all the time unless it is specifically a known bad file.

I only did see this one, so it looks like hopefully it was a passing issue they fixed shortly after it started.
____________
1 Corinthians 9:16 "For though I preach the gospel, I have nothing to glory of: for necessity is laid upon me; yea, woe is unto me, if I preach not the gospel!"
Ephesians 6:18-20, please ;-)
http://tbc-pa.org

Post to thread

Message boards : Server and website : Download stalled