Mysterious error 500 still with me

Message boards : BOINC Manager : Mysterious error 500 still with me
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next

AuthorMessage
Profile Ageless
Volunteer moderator
Project administrator
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 12109
Netherlands
Message 4076 - Posted: 25 Apr 2006, 22:15:11 UTC

Your mysterious error 500 should be over with with Boinc 5.4.6. If not, let us know and again on what.
Jord
Please do not private message me for tech support. Use the forums for that. Tech PMs will be ignored.
ID: 4076 · Report as offensive
Pepo
Avatar

Send message
Joined: 3 Apr 06
Posts: 547
Slovakia
Message 4078 - Posted: 26 Apr 2006, 13:50:27 UTC - in response to Message 4076.  

Your mysterious error 500 should be over with with Boinc 5.4.6. If not, let us know and again on what.

OT - as I've read in the Boinc 5.4.6's release notes: "NTLM Authentication for HTTP Proxies disabled".
I "live" currently behind one HTTP NTLM Authentication Proxy", does this actually mean for me no connection anymore?

Peter
ID: 4078 · Report as offensive
Profile Ageless
Volunteer moderator
Project administrator
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 12109
Netherlands
Message 4079 - Posted: 26 Apr 2006, 14:07:28 UTC - in response to Message 4078.  

does this actually mean for me no connection anymore?

Peter

There's always someone special. ;)

But no, Peter, as far as I know, it shouldn't affect you. I could be wrong though.
The thing is, with the old way libCurl connected, no-one behind a proxy could get through. This should fix that.

But ask Tigher if you see him to explain better. (I'll prod him in the direction of this thread ;))
Jord
Please do not private message me for tech support. Use the forums for that. Tech PMs will be ignored.
ID: 4079 · Report as offensive
Pepo
Avatar

Send message
Joined: 3 Apr 06
Posts: 547
Slovakia
Message 4080 - Posted: 26 Apr 2006, 15:27:28 UTC - in response to Message 4079.  

does this actually mean for me no connection anymore?

There's always someone special. ;)

Sure :-)

But no, Peter, as far as I know, it shouldn't affect you. I could be wrong though.
The thing is, with the old way libCurl connected, no-one behind a proxy could get through. This should fix that.

Actually I've been experiencing only problems here until I tried the 5.0.3.27, which cured it. (Is maybe this one proxy somehow problem-inversed? :)
But ask Tigher if you see him to explain better. (I'll prod him in the direction of this thread ;))

I'll wait while I have the connection. But will sure keep the SW change in mind.

Peter
ID: 4080 · Report as offensive
Profile Ageless
Volunteer moderator
Project administrator
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 12109
Netherlands
Message 4087 - Posted: 26 Apr 2006, 19:36:48 UTC

I got a reply from Ian (Tigher) through email:

It means there is no negotiation up from basic authentication (usually done by the proxy server itself) to NT Lan Manager authentication done by windows - often on a corporate basis. Its normally easier to move "down" the complexity chain as BOINC has done here so I would expect any proxy to do basic readily.

He needs to try I guess but he should be OK to be honest.

The upshot is that libcurl is "crap" at some things. We see this with authentication and proxy support. We see it with some aspects of http keep alive. It cannot handle redirects too well by the looks (guess on my part) and somewhere in all this is the 500 problem (probably many problems all looking like 500) which basically got worse with libcurl. My guess they regret the move to libcurl. Will they say so....nope!

Hope that helps
Ian


I hope he does find the time in his busy life to explain other things further. :)
Jord
Please do not private message me for tech support. Use the forums for that. Tech PMs will be ignored.
ID: 4087 · Report as offensive
Tigher

Send message
Joined: 25 Sep 05
Posts: 62
United Kingdom
Message 4089 - Posted: 27 Apr 2006, 0:14:49 UTC - in response to Message 4087.  

I got a reply from Ian (Tigher) through email:

It means there is no negotiation up from basic authentication (usually done by the proxy server itself) to NT Lan Manager authentication done by windows - often on a corporate basis. Its normally easier to move "down" the complexity chain as BOINC has done here so I would expect any proxy to do basic readily.

He needs to try I guess but he should be OK to be honest.

The upshot is that libcurl is "crap" at some things. We see this with authentication and proxy support. We see it with some aspects of http keep alive. It cannot handle redirects too well by the looks (guess on my part) and somewhere in all this is the 500 problem (probably many problems all looking like 500) which basically got worse with libcurl. My guess they regret the move to libcurl. Will they say so....nope!

Hope that helps
Ian


I hope he does find the time in his busy life to explain other things further. :)


Is there more that you think I can explain?


ID: 4089 · Report as offensive
Pepo
Avatar

Send message
Joined: 3 Apr 06
Posts: 547
Slovakia
Message 4091 - Posted: 27 Apr 2006, 7:30:14 UTC - in response to Message 4089.  

I got a reply from Ian (Tigher) through email:

It means there is no negotiation up from basic authentication (usually done by the proxy server itself) to NT Lan Manager authentication done by windows - often on a corporate basis. Its normally easier to move "down" the complexity chain as BOINC has done here so I would expect any proxy to do basic readily.

He needs to try I guess but he should be OK to be honest.

The upshot is that libcurl is "crap" at some things. We see this with authentication and proxy support. We see it with some aspects of http keep alive. It cannot handle redirects too well by the looks (guess on my part) and somewhere in all this is the 500 problem (probably many problems all looking like 500) which basically got worse with libcurl. My guess they regret the move to libcurl. Will they say so....nope!

Hope that helps
Ian


I hope he does find the time in his busy life to explain other things further. :)


Is there more that you think I can explain?

I'll check the 5.4.6 version and then tell how it's doing...

Thanks for help
Peter
ID: 4091 · Report as offensive
Pepo
Avatar

Send message
Joined: 3 Apr 06
Posts: 547
Slovakia
Message 4175 - Posted: 2 May 2006, 10:15:31 UTC - in response to Message 4091.  

Is there more that you think I can explain?

I'll check the 5.4.6 version and then tell how it's doing...

As I feared, I'm not comming through the NTLM proxy with the actual BCC 5.4.7 :-(
I took some communication snapshots with Ethereal, maybe Ian could be interested in them?

As I noticed that Boinc is not able to upload a result to Einstein (5 minutes timeout), I started Ethereal so see the bits... 5 minutes later Boinc tried to upload the result again and failed again. Then I suspended the communication and stopped the first capture.

In second round, I restarted Boinc and updated SetiBeta (while comm suspended, Einstein upload started too), after finishing the SetiBeta update and timeouting the Einstein upload I stopped the second capture.

In the third row I tried to upload Seti and Einstein units, after some problems it succeeded.

Why did it finally succeed? As I already know from browsing experience, the proxy kicks in with the username/password request differently - sometimes immediately, sometimes after few seconds or minutes or transferred kilobytes, I don't know the key. The real problem is when downloading, with some prior Boinc versions I often had many workunits containing proxy's error messages garbage instead of scientific data.

Peter
ID: 4175 · Report as offensive
Spok
Avatar

Send message
Joined: 14 Nov 05
Posts: 49
Czech Republic
Message 4186 - Posted: 4 May 2006, 5:17:43 UTC - in response to Message 4076.  

Your mysterious error 500 should be over with with Boinc 5.4.6. If not, let us know and again on what.



So, I finaly found time to test it.
Good news: Error 500 (or later 400) disappeared.
Bad news: Uploading started, and after few seconds stopped with progress 10.87/10.15 KB and 12.31/11.59 KB. Other results have "upload pending" status. Although scheduler request succeed (yes! :) all wu's are "download pending".

***************************************************
05/04/06 07:02:22|SETI@home|Started upload of file 24ja99aa.27554.13440.990892.1.8_3_0
05/04/06 07:02:22|SETI@home|Started upload of file 04mr99aa.14657.5730.636084.1.159_1_0
05/04/06 07:02:22||Using earliest-deadline-first scheduling because computer is overcommitted.
05/04/06 07:02:22|SETI@home|Sending scheduler request to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
05/04/06 07:02:22|SETI@home|Reason: To fetch work
05/04/06 07:02:22|SETI@home|Requesting 267250 seconds of new work
05/04/06 07:02:27|SETI@home|Scheduler request succeeded
05/04/06 07:03:02|LHC@home|Sending scheduler request to http://lhcathome-sched1.cern.ch/scheduler/cgi
05/04/06 07:03:02|LHC@home|Reason: To fetch work
05/04/06 07:03:02|LHC@home|Requesting 864000 seconds of new work
05/04/06 07:07:28||Project communication failed: attempting access to reference site
05/04/06 07:07:28|SETI@home|Temporarily failed upload of 24ja99aa.27554.13440.990892.1.8_3_0: http error
05/04/06 07:07:28|SETI@home|Backing off 1 minutes and 0 seconds on upload of file 24ja99aa.27554.13440.990892.1.8_3_0
05/04/06 07:07:28|SETI@home|Temporarily failed upload of 04mr99aa.14657.5730.636084.1.159_1_0: http error
05/04/06 07:07:28|SETI@home|Backing off 1 minutes and 0 seconds on upload of file 04mr99aa.14657.5730.636084.1.159_1_0

***************************************************

Let's test 5.4.7 and 5.4.8
Will report in few minutes.

Spok

ID: 4186 · Report as offensive
Spok
Avatar

Send message
Joined: 14 Nov 05
Posts: 49
Czech Republic
Message 4187 - Posted: 4 May 2006, 6:03:55 UTC

Boinc 5.4.7

I waited for a few minutes, then upload of all results failed. After this, all new wu's were succesfully downloaded. OK, there was no error during download, but cannot tell exactly, if there will be an error when some of wu will be crunched.

Now will try 5.4.8, if something will change.
ID: 4187 · Report as offensive
Spok
Avatar

Send message
Joined: 14 Nov 05
Posts: 49
Czech Republic
Message 4190 - Posted: 4 May 2006, 7:59:53 UTC
Last modified: 4 May 2006, 8:04:39 UTC

Upload not working even in 5.4.8


**************************************
4.5.2006 9:44:08|SETI@home|Started upload of file 24ja99aa.27554.13440.990892.1.8_3_0
4.5.2006 9:44:12||Access to reference site succeeded - project servers may be temporarily down.
4.5.2006 9:49:00||Project communication failed: attempting access to reference site
4.5.2006 9:49:02|SETI@home|Scheduler request failed: a timeout was reached
4.5.2006 9:49:02|SETI@home|Deferring scheduler requests for 1 minutes and 29 seconds
4.5.2006 9:49:03|SETI@home|Temporarily failed upload of 12ja99ab.3533.18464.440904.1.168_2_0: http error
4.5.2006 9:49:03|SETI@home|Backing off 1 minutes and 0 seconds on upload of file 12ja99ab.3533.18464.440904.1.168_2_0
4.5.2006 9:49:04||Access to reference site succeeded - project servers may be temporarily down.
4.5.2006 9:49:04|SETI@home|Started upload of file 04mr99aa.14657.5730.636084.1.159_1_0
4.5.2006 9:49:22||Project communication failed: attempting access to reference site
4.5.2006 9:49:22|SETI@home|Temporarily failed upload of 24ja99aa.27554.13440.990892.1.8_3_0: http error
4.5.2006 9:49:22|SETI@home|Backing off 6 minutes and 7 seconds on upload of file 24ja99aa.27554.13440.990892.1.8_3_0
**********************************


Is Seti server down?

Spok
ID: 4190 · Report as offensive
mmciastro
Avatar

Send message
Joined: 30 Aug 05
Posts: 98
Message 4192 - Posted: 4 May 2006, 11:36:59 UTC

you sure have persistence. The seti server status page is here. It states that as of 3 hours ago it had 64 results "ready to send" and was making new work at 14/second. maybe demand is higher than the supply creation rate.

I'd stick with what you have for a bit longer and see if you get work.

tony

Boinc Wiki

Sometimes I hate my Bob Newhart style of humor.
ID: 4192 · Report as offensive
Spok
Avatar

Send message
Joined: 14 Nov 05
Posts: 49
Czech Republic
Message 4194 - Posted: 4 May 2006, 13:58:38 UTC - in response to Message 4192.  

you sure have persistence. The seti server status page is here. It states that as of 3 hours ago it had 64 results "ready to send" and was making new work at 14/second. maybe demand is higher than the supply creation rate.

I'd stick with what you have for a bit longer and see if you get work.

tony


Missunderstanding, I think.
Download is working. I already downloaded several WUs.
Upload not work.
ID: 4194 · Report as offensive
Profile Ageless
Volunteer moderator
Project administrator
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 12109
Netherlands
Message 4196 - Posted: 4 May 2006, 14:32:59 UTC - in response to Message 4194.  

4.5.2006 9:44:12||Access to reference site succeeded - project servers may be temporarily down.
4.5.2006 9:49:00||Project communication failed: attempting access to reference site
4.5.2006 9:49:02|SETI@home|Scheduler request failed: a timeout was reached

That error usually happens when you're trying to reach a mirror and you're the only mirror. But Seti has no mirrors, especially not to upload to. ;)

Yet... that last part suggests you are blocking Seti or BOINC on the way out.
Did you fully allow BOINC(.exe) through your firewall?
Jord
Please do not private message me for tech support. Use the forums for that. Tech PMs will be ignored.
ID: 4196 · Report as offensive
ThePhantom86
Avatar

Send message
Joined: 6 May 06
Posts: 16
United States
Message 4213 - Posted: 6 May 2006, 20:47:12 UTC

I'm in the same boat as Spok with my comp at work. 5.2.13 gave me the 500 errors. 5.4.6 gives me 400 errors. I'll keep an eye on this thread.
ID: 4213 · Report as offensive
Profile Ageless
Volunteer moderator
Project administrator
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 12109
Netherlands
Message 4214 - Posted: 6 May 2006, 21:27:09 UTC - in response to Message 4213.  

5.4.6 gives me 400 errors. I'll keep an eye on this thread.

5.4.6 was the first with NTLM Authentication for proxies disabled. A 400 error means a bad request, just as in HTML. You could try the latest version (5.4.9).
Jord
Please do not private message me for tech support. Use the forums for that. Tech PMs will be ignored.
ID: 4214 · Report as offensive
ThePhantom86
Avatar

Send message
Joined: 6 May 06
Posts: 16
United States
Message 4221 - Posted: 7 May 2006, 0:13:11 UTC

Still getting the 400 error with 5.4.9.

5/6/2006 8:07:52 PM||Version change (5.4.6 -> 5.4.9); running CPU benchmarks
5/6/2006 8:07:52 PM||Local control only allowed
5/6/2006 8:07:52 PM||Listening on port 31416
5/6/2006 8:07:52 PM||Suspending network activity - running CPU benchmarks
5/6/2006 8:07:54 PM||Running CPU benchmarks
5/6/2006 8:08:53 PM||Benchmark results:
5/6/2006 8:08:53 PM|| Number of CPUs: 1
5/6/2006 8:08:53 PM|| 1408 floating point MIPS (Whetstone) per CPU
5/6/2006 8:08:53 PM|| 2842 integer MIPS (Dhrystone) per CPU
5/6/2006 8:08:53 PM||Finished CPU benchmarks
5/6/2006 8:08:54 PM|SETI@home|Deferring task 01mr99ab.4149.12257.273584.3.169_2
5/6/2006 8:08:54 PM|Einstein@Home|Resuming task z1_1464.5__2543_S4R2a_0 using albert version 437
5/6/2006 8:08:54 PM||Resuming computation
5/6/2006 8:08:54 PM||Rescheduling CPU: Resuming computation
5/6/2006 8:08:54 PM||Resuming network activity
5/6/2006 8:08:55 PM|SETI@home|Restarting task 01mr99ab.4149.12257.273584.3.169_2 using setiathome_enhanced version 512
5/6/2006 8:08:55 PM|Einstein@Home|Pausing task z1_1464.5__2543_S4R2a_0 (left in memory)
5/6/2006 8:08:55 PM|LHC@home|Sending scheduler request to http://lhcathome-sched1.cern.ch/scheduler/cgi
5/6/2006 8:08:55 PM|LHC@home|Reason: To fetch work
5/6/2006 8:08:55 PM|LHC@home|Requesting 129600 seconds of new work
5/6/2006 8:09:00 PM|LHC@home|Scheduler request succeeded
5/6/2006 8:09:00 PM|LHC@home|No work from project
5/6/2006 8:09:00 PM|LHC@home|Deferring scheduler requests for 1 hours, 53 minutes and 51 seconds
5/6/2006 8:09:05 PM|Predictor @ Home|Sending scheduler request to http://predictor.scripps.edu/predictor_cgi/cgi
5/6/2006 8:09:05 PM|Predictor @ Home|Reason: To fetch work
5/6/2006 8:09:05 PM|Predictor @ Home|Requesting 20863 seconds of new work
5/6/2006 8:09:05 PM||Project communication failed: attempting access to reference site
5/6/2006 8:09:06 PM||Access to reference site succeeded - project servers may be temporarily down.
5/6/2006 8:09:10 PM|Predictor @ Home|Scheduler request failed: failed sending data to the peer
5/6/2006 8:09:10 PM|Predictor @ Home|Deferring scheduler requests for 1 hours, 58 minutes and 12 seconds
5/6/2006 8:10:20 PM||Rescheduling CPU: application exited
5/6/2006 8:10:20 PM|SETI@home|Computation for task 01mr99ab.4149.12257.273584.3.169_2 finished
5/6/2006 8:10:20 PM|SETI@home|Starting task 01mr99ab.4149.12257.273584.3.191_0 using setiathome_enhanced version 512
5/6/2006 8:10:22 PM|SETI@home|Started upload of file 01mr99ab.4149.12257.273584.3.169_2_0
5/6/2006 8:10:24 PM|SETI@home|Temporarily failed upload of 01mr99ab.4149.12257.273584.3.169_2_0: Error 400
5/6/2006 8:10:24 PM|SETI@home|Backing off 1 minutes and 0 seconds on upload of file 01mr99ab.4149.12257.273584.3.169_2_0
5/6/2006 8:11:25 PM|SETI@home|Started upload of file 01mr99ab.4149.12257.273584.3.169_2_0
5/6/2006 8:11:28 PM|SETI@home|Temporarily failed upload of 01mr99ab.4149.12257.273584.3.169_2_0: Error 400
5/6/2006 8:11:28 PM|SETI@home|Backing off 1 minutes and 0 seconds on upload of file 01mr99ab.4149.12257.273584.3.169_2_0

ID: 4221 · Report as offensive
Pepo
Avatar

Send message
Joined: 3 Apr 06
Posts: 547
Slovakia
Message 4240 - Posted: 8 May 2006, 13:25:24 UTC
Last modified: 8 May 2006, 13:27:30 UTC

Bad news, I've also observed some Error 400 and failed/rejected transfer with invalidated unit because of this:

08/05/2006 13:39:16||Resuming network activity
08/05/2006 13:39:17|SETI@home|Started upload of file 29mr99aa.2980.24961.1047168.1.116_0_0
08/05/2006 13:39:17|SETI@home|Started upload of file 19mr99aa.11608.5825.972158.1.196_0_0
08/05/2006 13:39:19|SETI@home|Temporarily failed upload of 19mr99aa.11608.5825.972158.1.196_0_0: Error 400
08/05/2006 13:39:19|SETI@home|Backing off 1 minutes and 0 seconds on upload of file 19mr99aa.11608.5825.972158.1.196_0_0
08/05/2006 13:39:59|SETI@home|Started upload of file 19mr99aa.11608.5825.972158.1.196_0_0
08/05/2006 13:40:02|SETI@home|Finished upload of file 19mr99aa.11608.5825.972158.1.196_0_0
08/05/2006 13:40:02|SETI@home|Throughput 14710 bytes/sec
08/05/2006 13:41:20|SETI@home|Error on file upload: no command
08/05/2006 13:41:20|SETI@home|Permanently failed upload of 29mr99aa.2980.24961.1047168.1.116_0_0
08/05/2006 13:41:20|SETI@home|Giving up on upload of 29mr99aa.2980.24961.1047168.1.116_0_0: server rejected file
08/05/2006 13:44:59||Suspending network activity - time of day
08/05/2006 13:50:41||Resuming network activity
08/05/2006 13:50:41|SETI@home|Sending scheduler request to http://setiboinc.ssl.berkeley.edu/sah_cgi/cgi
08/05/2006 13:50:41|SETI@home|Reason: Requested by user
08/05/2006 13:50:41|SETI@home|Reporting 2 tasks
08/05/2006 13:50:46|SETI@home|Scheduler request succeeded
08/05/2006 13:55:41||Suspending network activity - time of day

Still on 5.4.7 (as 5.4.8 and 5.4.9 contain no changes addresing the proxies).

Peter
ID: 4240 · Report as offensive
Spok
Avatar

Send message
Joined: 14 Nov 05
Posts: 49
Czech Republic
Message 4841 - Posted: 26 Jun 2006, 4:19:05 UTC
Last modified: 26 Jun 2006, 4:21:11 UTC

Just to remember this problem...

BOINC 5.5.4


******************************************************************
26.6.2006 6:14:05|LHC@home|Sending scheduler request to http://lhcathome-sched1.cern.ch/scheduler/cgi
26.6.2006 6:14:05|LHC@home|Reason: To fetch work
26.6.2006 6:14:05|LHC@home|Requesting 864000 seconds of new work
26.6.2006 6:14:07||Project communication failed: attempting access to reference site
26.6.2006 6:14:08||Access to reference site succeeded - project servers may be temporarily down.
26.6.2006 6:14:11|LHC@home|Scheduler request failed: failed sending data to the peer
26.6.2006 6:14:11|LHC@home|Deferring scheduler requests for 1 minutes and 0 seconds
26.6.2006 6:15:11|LHC@home|Sending scheduler request to http://lhcathome-sched1.cern.ch/scheduler/cgi
26.6.2006 6:15:11|LHC@home|Reason: To fetch work
26.6.2006 6:15:11|LHC@home|Requesting 864000 seconds of new work
26.6.2006 6:15:16|LHC@home|Scheduler request failed: Error 400
******************************************************************

ID: 4841 · Report as offensive
Spok
Avatar

Send message
Joined: 14 Nov 05
Posts: 49
Czech Republic
Message 5222 - Posted: 7 Aug 2006, 6:14:45 UTC
Last modified: 7 Aug 2006, 6:15:00 UTC

Boinc 5.5.10

New version, old problem...


******************************************************************
7.8.2006 8:11:28|SETI@home|Sending scheduler request: Requested by user
7.8.2006 8:11:28|SETI@home|(not requesting new work or reporting completed tasks)
7.8.2006 8:11:33|SETI@home|Scheduler request failed: Error 400
7.8.2006 8:11:33|SETI@home|Deferring scheduler requests for 1 minutes and 0 seconds
******************************************************************

ID: 5222 · Report as offensive
Previous · 1 · 2 · 3 · 4 · 5 · 6 · Next

Message boards : BOINC Manager : Mysterious error 500 still with me

Copyright © 2017 University of California. Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.