Pages: [1]
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 08:32:34

Today, when I came into my office, 10 from 12 BOINC-Clients have crashed.

Looking in stdoutdae.txt shows:

22-Jun-2010 02:06:38 [Poem@Home] Message from server: No work sent
22-Jun-2010 02:06:38 [Poem@Home] Message from server: (reached limit of 90 tasks in progress)
22-Jun-2010 02:11:55 [---] Contacting account manager at http://bam.boincstats.com/
22-Jun-2010 02:12:00 [---] Account manager: BAM! User-ID: 940
22-Jun-2010 02:12:00 [---] Account manager: BAM! Host-ID: 110192
22-Jun-2010 02:12:00 [---] Account manager: Number of BAM! connections: 17055
22-Jun-2010 02:12:00 [---] Account manager: Dear founder of team Nordlichter:
22-Jun-2010 02:12:00 [---] Account manager: You are invited to participate in the new challenge section on BOINCstats: http://boincstats.com/bam/challenge.php
22-Jun-2010 02:12:00 [---] Account manager: Delayed detach from FreeHAL, project changed URL.
After detaching project will reattach with new URL on next RPC!
22-Jun-2010 02:12:00 [---] Account manager contact succeeded

On 02:12:00 the client crashed

I checked a second client and it seems to be exactly the same: The last line is "Account manager contact succeeded" and then the client crashed

Oh, I forgot: The first mentioned client had already crashed soem hours before; same point "Account manager contact succeeded" is last line in stdoutdae.txt

Oh2, the clients vary from older versions up to the latest ones, so it is definitly not a client specific problem

Yeti
[BOINCstats] Willy
 
Forum moderator - Administrator - Developer - Tester - Translator
BAM!ID: 1
Joined: 2006-01-09
Posts: 9442
Credits: 353,172,950
World-rank: 4,877

2010-06-22 08:34:48

Can't reproduce. All my clients have the same messages and they didn't crash.
Please do not PM, IM or email me for support (they will go unread/ignored). Use the forum for support.
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 08:55:44

It seems as if it has to do with the url-change of freehal:

2-Jun-2010 10:48:06 [---] Account manager: Delayed detach from FreeHAL, project changed URL.
After detaching project will reattach with new URL on next RPC!
22-Jun-2010 10:48:06 [---] Account manager contact succeeded

After this contact, the client crashes immediatly

I can reproduce this as often as you like


Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 08:58:09

Meanwhile the client is attached twice to freehal ...
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 09:05:30

22/06/2010 10:58:14 http://freehal.net/freehal_at_home/ Fetching scheduler list
22/06/2010 10:58:17 http://freehal.net/freehal_at_home/ Master file download succeeded
22/06/2010 10:58:22 http://freehal.net/freehal_at_home/ Sending scheduler request: Project initialization.
22/06/2010 10:58:22 http://freehal.net/freehal_at_home/ Requesting new tasks
22/06/2010 10:58:28 FreeHAL@home Scheduler request completed: got 0 new tasks
22/06/2010 10:58:28 FreeHAL@home You used the wrong URL for this project
22/06/2010 10:58:28 FreeHAL@home The correct URL is http://www.freehal.net/freehal_at_home/
22/06/2010 10:58:28 FreeHAL@home You seem to be attached to this project twice
22/06/2010 10:58:28 FreeHAL@home We suggest that you detach projects named FreeHAL@home,
22/06/2010 10:58:28 FreeHAL@home then reattach to http://www.freehal.net/freehal_at_home/
22/06/2010 10:58:28 FreeHAL@home Already attached to a project named FreeHAL@home (possibly with wrong URL)
22/06/2010 10:58:28 FreeHAL@home Consider detaching this project, then trying again
22/06/2010 10:58:28 FreeHAL@home Message from server: Not sending work - last request too recent: 207 sec
22/06/2010 10:58:28 FreeHAL@home New computer location: home
22/06/2010 11:02:00 Fetching configuration file from http://bam.boincstats.com/get_project_config.php
22/06/2010 11:02:03 Contacting account manager at http://bam.boincstats.com/
22/06/2010 11:02:06 Account manager: BAM! User-ID: 940
22/06/2010 11:02:06 Account manager: BAM! Host-ID: 110192
22/06/2010 11:02:06 Account manager: Number of BAM! connections: 17061
22/06/2010 11:02:06 Account manager: Dear founder of team Nordlichter:
22/06/2010 11:02:06 Account manager: You are invited to participate in the new challenge section on BOINCstats: http://boincstats.com/bam/challenge.php
22/06/2010 11:02:06 Account manager: Delayed detach from FreeHAL, project changed URL. After detaching project will reattach with new URL on next RPC!
22/06/2010 11:02:06 Account manager contact succeeded
22/06/2010 11:02:15 FreeHAL@home Resetting project
22/06/2010 11:02:16 FreeHAL@home Detaching from project

After this, one Freehal is still remaining, but I can not detach, because the detach-button is grey
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 09:11:04

With the next contact to BAM, the client crashed

During BAM-communication the client tells that an error occured while communicating with BAM
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 09:24:40

To get my client working again I finally:

1) detached from BAM
2) detached from FreeHal (1st)
3) detached from FreeHal (2nd)
4) reattached to BAM

Now, it seems that the client is working again. Several BAM-contacts have followed but no chrash has happened again

So, I will have to do this with every client on my network :-(
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 10:45:44

If you want me to reproduce it or help to test if you could solve this problem, I have 1 client staying in last configuration so that we could use it to test ...

Please let me know if I should keep it as it is or fix it manually as my other clients

Yeti
rroonnaalldd
BAM!ID: 15220
Joined: 2006-12-20
Posts: 207
Credits: 77,733,697
World-rank: 13,612

2010-06-22 14:32:00
last modified: 2010-06-22 14:41:40

...same crashes here with boinc 6.10.56-final

I have removed freehal from my project list in BAM but this project is ever listed. Now i could reattach to freehal in BAM but with a listed freehal i got and will get the known message:
You seem to be attached to this project twice
We suggest that you detach projects named FreeHAL@home,
then reattach to http://www.freehal.net/freehal_at_home/


[BOINCstats] Willy
 
Forum moderator - Administrator - Developer - Tester - Translator
BAM!ID: 1
Joined: 2006-01-09
Posts: 9442
Credits: 353,172,950
World-rank: 4,877

2010-06-22 14:43:37
last modified: 2010-06-22 14:45:51

I just upgraded to BOINC 6.10.56 and my client also crashed after every BAM! connection. I downgraded to my previous BOINC version and the problem is gone. I suspect a bug in the BOINC code.

Edit: reported it to the BOINC devs.
Please do not PM, IM or email me for support (they will go unread/ignored). Use the forum for support.
rroonnaalldd
BAM!ID: 15220
Joined: 2006-12-20
Posts: 207
Credits: 77,733,697
World-rank: 13,612

2010-06-22 14:50:34

BOINCstats Willy wrote:

Edit: reported it to the BOINC devs.

Thanks, i was to slow...
[BOINCstats] LostBoy
BOINCstats SOFA member
BAM!ID: 231
Joined: 2006-05-12
Posts: 150
Credits: 168,139,216
World-rank: 8,072

2010-06-22 14:58:21
last modified: 2010-06-22 15:01:10

I have the same problem here with BOINC 6.10.18.

After detaching with BAM!,see here:

22.06.2010 16:43:23 http://freehal.net/freehal_at_home/ Fetching scheduler list
22.06.2010 16:43:27 Account manager: BAM! User-ID: 231
22.06.2010 16:43:27 Account manager: BAM! Host-ID: 208827
22.06.2010 16:43:27 Account manager: Number of BAM! connections: 2275
22.06.2010 16:43:27 Account manager: Delayed detach from FreeHAL, project changed URL. After detaching project will reattach with new URL on next RPC!
22.06.2010 16:43:27 Account manager contact succeeded
22.06.2010 16:43:29 http://freehal.net/freehal_at_home/ Master file download succeeded
22.06.2010 16:43:34 http://freehal.net/freehal_at_home/ Resetting project
22.06.2010 16:43:34 http://freehal.net/freehal_at_home/ Detaching from project


FreeHal is still listing in my BOINC manager???

This is what I do not understand,"After detaching project will reattach with new URL on next RPC!"
Why reattaching?
I want to detach from FreeHal.
The checkbox in BAM! is empty.
Yeti
 
BAM!ID: 940
Joined: 2006-05-29
Posts: 20
Credits: 567,440,419
World-rank: 3,450

2010-06-22 15:01:27
last modified: 2010-06-22 15:17:31

BOINCstats Willy wrote:
I just upgraded to BOINC 6.10.56 and my client also crashed after every BAM! connection. I downgraded to my previous BOINC version and the problem is gone. I suspect a bug in the BOINC code.

Edit: reported it to the BOINC devs.


HM, it has happened to a lot of different clients:

5.10.28
6.10.13
6.10.16
6.10.36
6.10.51
6.10.56

Yeti

edit: filling in Version 6.10.13
rroonnaalldd
BAM!ID: 15220
Joined: 2006-12-20
Posts: 207
Credits: 77,733,697
World-rank: 13,612

2010-06-22 15:01:59
last modified: 2010-06-22 15:02:32

LostBoy, try the "Yeti" way.
rroonnaalldd
BAM!ID: 15220
Joined: 2006-12-20
Posts: 207
Credits: 77,733,697
World-rank: 13,612

2010-06-22 18:32:14

Past detaching from FreeHAL, reattach to BAM and some BAM connections i want to be stupid. I reattached to FreeHAL again and had the same situation like before. The boinc-client is crashing in the next connection to BAM and FreeHAL is listed twice in boincmanager.
Here is a snipped from my client_state.xml:
<project>
<master_url>http://www.freehal.net/freehal_at_home/</master_url>
<project_name>FreeHAL@home</project_name>
<user_name>rroonnaalldd</user_name>
<scheduler_url>http://freehal.net/freehal_at_home_cgi/cgi</scheduler_url>
<scheduler_url>http://freehal.net/freehal_at_home/cgi-bin/scheduler</scheduler_url>

.
.
.
<project>
<master_url>http://freehal.net/freehal_at_home/</master_url>
<project_name>FreeHAL@home</project_name>
<user_name>rroonnaalldd</user_name>
<scheduler_url>http://freehal.net/freehal_at_home_cgi/cgi</scheduler_url>
<scheduler_url>http://freehal.net/freehal_at_home/cgi-bin/scheduler</scheduler_url>



I have done the Yeti way again and reconnected to BAM without FreeHAL on all hosts.
Crunch3r
 
BAM!ID: 30440
Joined: 2007-07-15
Posts: 32
Credits: 176,999,852,198
World-rank: 46

2010-06-22 18:53:24

Same problem here as well. Each time the client tried to sync with BAM, it crashed immediately.
The only way to fix it was to delete the freehal accout xml in the boinc data directory and disabling Freehal in BAM as well.
[BOINCstats] Willy
 
Forum moderator - Administrator - Developer - Tester - Translator
BAM!ID: 1
Joined: 2006-01-09
Posts: 9442
Credits: 353,172,950
World-rank: 4,877

2010-06-22 19:11:32
last modified: 2010-06-22 19:14:38

It was a combined problem of BAM! and the BOINC manager. The problem has been fixed at both sides (immediately in BAM! and in the next BOINC version) and the problem should no longer occur.
Please do not PM, IM or email me for support (they will go unread/ignored). Use the forum for support.
Pages: [1]

Index :: BAM! Bug Report :: BAM shot my boinc-network down
Reason: