Upload Problems
Message boards :
Server and website :
Upload Problems
Message board moderation
Previous · 1 . . . 3 · 4 · 5 · 6
| Author | Message |
|---|---|
|
Send message Joined: 27 Nov 11 Posts: 11 Credit: 1,021,749,297 RAC: 0 Level  Scientific publications      
|
Yea that was it I updated to the latest boinc version & it's now updating fine :) (win 7 x64) |
|
Send message Joined: 3 Oct 11 Posts: 100 Credit: 5,879,292,399 RAC: 0 Level  Scientific publications
|
I can not upload any task. HTTP transient error... |
|
Send message Joined: 16 Apr 09 Posts: 163 Credit: 921,733,849 RAC: 0 Level  Scientific publications
|
I've encountered it now four separate times -- once I lost a completed work unit -- due to a configuration issue with BOINC I suspect. Perhaps the change to HTTPS is causing a bit of confusion with the uploads. Again, each time I moved to the current BOINC client, the uploads completed. |
 Crunch3rSend message Joined: 16 Mar 09 Posts: 3 Credit: 207,697,314 RAC: 0 Level  Scientific publications
|
The only reason for the upload issues is an outdated ca-certificate. There's no reason to upgrade to a new boinc version. Simply replace the ca-bundle.crt file in the boinc program directory with the one below (extracted from boinc 7.4.36), restart boinc and that's it. download -> http://www.boincunited.org/ca-bundle.crt  Join BOINC United now! |
|
Send message Joined: 3 Oct 11 Posts: 100 Credit: 5,879,292,399 RAC: 0 Level Scientific publications
|
My upload starts ok, but after some time (10, 20, 50% progress) it stops. I think it's not related to cert issue. |
|
Send message Joined: 11 Jul 09 Posts: 1639 Credit: 10,159,968,649 RAC: 0 Level  Scientific publications 
|
The only reason for the upload issues is an outdated ca-certificate. Thanks - that did the trick for my BOINC v6.12.34 (although I pulled the v7.4.36 bundle from another machine on my own network). Didn't even need to restart the (service-mode) client. But I think we're not out of the woods yet, because I'm getting multiple re-directs (which may account for the stop/restart problem just reported) 19-Mar-2015 18:08:23 [---] [http] [ID#5324] Info: About to connect() to www.gpugrid.org port 80 (#3) |
|
Send message Joined: 15 Feb 07 Posts: 134 Credit: 1,349,535,983 RAC: 0 Level Scientific publications
|
1) Clients that are configured to use the HTTP presentation of GPUGRID shouldn't now be being redirected to HTTPS (for a while last night they were, just to see what would break) 2) If uploads you are having uploads fail, this is probably because the receiving process on the server will only run for so long. I'll need to know not the %age completion but the walltime that the upload had been processing for. 3) 301 redirects are not a problem - the client knows to follow them. Matt |
|
Send message Joined: 3 Oct 11 Posts: 100 Credit: 5,879,292,399 RAC: 0 Level Scientific publications
|
5518 GPUGRID 19.3.2015 20:17:53 Started upload of 754-NOELIA_POT-8-13-RND0792_0_0 5519 GPUGRID 19.3.2015 20:18:00 Finished upload of 754-NOELIA_POT-8-13-RND0792_0_0 5520 GPUGRID 19.3.2015 20:18:06 Started upload of 754-NOELIA_POT-8-13-RND0792_0_1 5521 GPUGRID 19.3.2015 20:18:15 Finished upload of 754-NOELIA_POT-8-13-RND0792_0_1 5522 GPUGRID 19.3.2015 20:18:34 Started upload of 754-NOELIA_POT-8-13-RND0792_0_9 5523 19.3.2015 20:19:09 Project communication failed: attempting access to reference site 5524 GPUGRID 19.3.2015 20:19:09 Temporarily failed upload of 754-NOELIA_POT-8-13-RND0792_0_9: transient HTTP error 5525 GPUGRID 19.3.2015 20:19:09 Backing off 00:04:41 on upload of 754-NOELIA_POT-8-13-RND0792_0_9 5526 19.3.2015 20:19:10 Internet access OK - project servers may be temporarily down. |
|
Send message Joined: 23 Dec 09 Posts: 189 Credit: 4,813,881,008 RAC: 42,053 Level  Scientific publications
|
Hi, All of my 3 machines have the "transient HTTP error" since yesterday evening. I have not changed anything! BUT THERE IS NOW A PARTICULAR ERRO MASSEGE: "Scheduler request failed: Peer certificate cannot be authenticated with known CA certificates" DOES ANYBODY HAVE THE SAME PROBLEM? Log of on is like this: "19/03/2015 10:07:10 a.m. | GPUGRID | Sending scheduler request: Requested by project. 19/03/2015 10:07:10 a.m. | GPUGRID | Requesting new tasks for NVIDIA 19/03/2015 10:07:12 a.m. | GPUGRID | Scheduler request failed: Peer certificate cannot be authenticated with known CA certificates 19/03/2015 10:07:16 a.m. | | Project communication failed: attempting access to reference site 19/03/2015 10:07:19 a.m. | | Internet access OK - project servers may be temporarily down. 19/03/2015 01:25:04 p.m. | GPUGRID | Started upload of e18s8_e11s4f108-GERARD_CXCL12_LIG22_CGENFF2-0-2-RND4205_0_9 19/03/2015 01:25:04 p.m. | GPUGRID | Started upload of e16s6_e1s31f254-GERARD_CXCL12_Ctl11_GAAMPGAFF1-1-2-RND7394_0_0 19/03/2015 01:25:06 p.m. | GPUGRID | Temporarily failed upload of e18s8_e11s4f108-GERARD_CXCL12_LIG22_CGENFF2-0-2-RND4205_0_9: transient HTTP error 19/03/2015 01:25:06 p.m. | GPUGRID | Backing off 3 hr 29 min 2 sec on upload of e18s8_e11s4f108-GERARD_CXCL12_LIG22_CGENFF2-0-2-RND4205_0_9 19/03/2015 01:25:06 p.m. | GPUGRID | Temporarily failed upload of e16s6_e1s31f254-GERARD_CXCL12_Ctl11_GAAMPGAFF1-1-2-RND7394_0_0: transient HTTP error 19/03/2015 01:25:06 p.m. | GPUGRID | Backing off 23 min 0 sec on upload of e16s6_e1s31f254-GERARD_CXCL12_Ctl11_GAAMPGAFF1-1-2-RND7394_0_0 19/03/2015 01:25:09 p.m. | | Project communication failed: attempting access to reference site 19/03/2015 01:25:11 p.m. | | Internet access OK - project servers may be temporarily down. 19/03/2015 01:37:28 p.m. | GPUGRID | Sending scheduler request: Requested by project. 19/03/2015 01:37:28 p.m. | GPUGRID | Requesting new tasks for NVIDIA 19/03/2015 01:37:30 p.m. | GPUGRID | Scheduler request failed: Peer certificate cannot be authenticated with known CA certificates 19/03/2015 01:37:34 p.m. | | Project communication failed: attempting access to reference site 19/03/2015 01:37:36 p.m. | | Internet access OK - project servers may be temporarily down. 19/03/2015 03:17:52 p.m. | GPUGRID | update requested by user 19/03/2015 03:17:52 p.m. | GPUGRID | Sending scheduler request: Requested by user. 19/03/2015 03:17:52 p.m. | GPUGRID | Requesting new tasks for NVIDIA 19/03/2015 03:17:54 p.m. | GPUGRID | Scheduler request failed: Peer certificate cannot be authenticated with known CA certificates 19/03/2015 03:17:57 p.m. | | Project communication failed: attempting access to reference site 19/03/2015 03:18:01 p.m. | | Internet access OK - project servers may be temporarily down. 19/03/2015 03:19:39 p.m. | GPUGRID | Fetching scheduler list 19/03/2015 03:19:44 p.m. | | Project communication failed: attempting access to reference site 19/03/2015 03:19:46 p.m. | | Internet access OK - project servers may be temporarily down." Please advice! Thanks! |
|
Send message Joined: 2 May 13 Posts: 8 Credit: 3,166,491,908 RAC: 103,485 Level Scientific publications
|
Yes, I had the same message. I upgraded to the latest version of BOINC and now all is well. |
|
Blurf Send message Joined: 20 Dec 11 Posts: 9 Credit: 120,872,506 RAC: 0 Level  Scientific publications
|
Files uploaded but now I can't report them |
|
Send message Joined: 16 Apr 09 Posts: 163 Credit: 921,733,849 RAC: 0 Level Scientific publications
|
Blurf similar problem -- on now the FIFTH of my computers and ONLY with GPUGrid. On the other four, updated to the newest version of the BOINC client resolved the GPUGrid specific problem. On this fifth computer -- it was already on that version, a repair of that version didn't resolve the problem, a uninstall reinstall didn't resolve the problem, a replacement of the credentials file didn't resolve that problem. Since it is specific to GPUGrid (hoping the GPUGrid folks realize that), that particular computer is now running Collatz for GPU processing instead. Oh a reset of the project didn't solve the problem, a delete the project and reattach didn't solve the problem either. Seems it is definitely a project specific issue. Perhaps if it affects everyone it will move to the top of the 'what the heck is going on list'. That's two reported work units I've lost in the past 24 hours - -and two that the folks at the project are not going to get unfortunately. I REALLY hope the folks at the project figure this one out. |
|
Send message Joined: 16 Apr 09 Posts: 163 Credit: 921,733,849 RAC: 0 Level Scientific publications
|
By the way, I took a ca file from another workstation which was (or at least still is) working, and brought it over to the workstation which doesn't believe in GPUGrid any more. No joy, said the server was offline (it isn't). Something clearly is weird with GPU grid and its cert handling or addressing -- in some cases it get's resolved by updating BOINC to the newest version -- though the cert file is identical. The one 'I don't believe GPUGrid server is alive' workstation uses a GTX-650 -- so I guess it isn't that much of a loss. Still I wonder what in blazes is going on here. I'm glad others are reporting the problem as that likely moves it up the 'let's look at this' list. |
|
Send message Joined: 23 Dec 09 Posts: 189 Credit: 4,813,881,008 RAC: 42,053 Level Scientific publications
|
The good news first: All my result fill have no up-loaded to the GPUGRID servers! The bad news is: They still are listed on the web site as running / calculating (in Spanish it is “En progreso”) because there is no communication with the reference site on two (02) of my three (03) computers: “20/03/2015 09:39:10 a.m. | GPUGRID | update requested by user 20/03/2015 09:39:12 a.m. | GPUGRID | Fetching scheduler list 20/03/2015 09:39:17 a.m. | | Project communication failed: attempting access to reference site 20/03/2015 09:39:21 a.m. | | Internet access OK - project servers may be temporarily down.” It is very frustrating, that the computers have finished the WUs, I will spend on the electric bill, but finally the Wus will not count as they are not reported as finished. |
|
Send message Joined: 16 Apr 09 Posts: 163 Credit: 921,733,849 RAC: 0 Level Scientific publications
|
I am guessing here in the absence of data, just judging from the timing here that something done regarding the shift back and forth from HTTP to HTTPS resulted in 'confusion' regarding the certification of the site (the CA certs). Even though the cert file is the same as it was a year ago, at least *some* workstation installations see the GPUGrid site as no longer certified (out of date certs). *Sometimes* when this has happened, particularly if you have a somewhat older version of the BOINC client, updating the BOINC client resolves this. If you have the current version of the BOINC client and the 'can't connect' problem shows up, it seems like there is no resolution. However, it might also be a timing thing. Last evening, when I tried to add the GPUGrid project back in on what was my 5th workstation with the problem (and only one that already had the current BOINC client), I was not successful -- trying each and every approach discussed in this thread. This morning on this same workstation, I was successful in connecting and downloading. I won't know if reporting works until tomorrow evening as the workstation is running a GTX 650 which has about a 34 hour process time for the large workunits. It seems though as the problem has been reported by several users and is specific to GPUGrid, that there was some change at the project, which perhaps will clear over time. It isn't clear that folks at the project have the resources to evaluate this. (If it is a persisting problem for increasing user population it might get higher priority I suspect). |
©2026 Universitat Pompeu Fabra