lost connection to local host

Message boards : BOINC Manager : lost connection to local host
Message board moderation

To post messages, you must log in.

AuthorMessage
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6504 - Posted: 17 Nov 2006, 17:54:41 UTC

for the past few days 11/15/06 the BONIC Manager can not fine the local host. all processing stops. running 5.4.11. on 24/7. running 4 or more projects round robin. only way to get working again is to exit and re-start. any idea why this is happing. any one else have the same issue?
ID: 6504 · Report as offensive
SekeRob

Send message
Joined: 25 Aug 06
Posts: 1596
Message 6516 - Posted: 18 Nov 2006, 18:39:54 UTC - in response to Message 6504.  
Last modified: 18 Nov 2006, 18:42:33 UTC

for the past few days 11/15/06 the BOiNC Manager can not fine the local host. all processing stops. running 5.4.11. on 24/7. running 4 or more projects round robin. only way to get working again is to exit and re-start. any idea why this is happing. any one else have the same issue?


hmm did it coincide with updating/applying windows security patches?

Try typing in the actual computer name instead of local host. That's the name that i've entered in the remote.... file.



Coelum Non Animum Mutant, Qui Trans Mare Currunt
ID: 6516 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 14625
Netherlands
Message 6519 - Posted: 18 Nov 2006, 20:26:23 UTC

The Boinc daemon (boinc.exe), the GUI (boincmgr.exe) and the screen saver program (boinc.scr) all communicate among each other using TCP port 31416. If for some reason this port is taken, you can get that message.

It can also happen that for some reason, Boinc Manager starts earlier than the daemon... so it won't find the main program.

I'd also check Windows updates first. See what was installed. Or see if the Windows firewall was updated and that the three programs are still allowed through it.
ID: 6519 · Report as offensive
Thibi

Send message
Joined: 20 Nov 06
Posts: 1
Belgium
Message 6555 - Posted: 20 Nov 2006, 22:51:51 UTC

got that problem also alot lately.

only thing that helps is restarting it :/
ID: 6555 · Report as offensive
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6568 - Posted: 21 Nov 2006, 20:03:49 UTC

there have been updates done, can't say if it was happing before or after.

I would "exit" and re-start and it would lose/fail to connect with in hours to a day later. repeating again and the same would happen w/in same time frame again.

11/21/06 did a re-install/repair in the AM and have yet to loose connection as of 15:00 est.
ID: 6568 · Report as offensive
SekeRob

Send message
Joined: 25 Aug 06
Posts: 1596
Message 6570 - Posted: 21 Nov 2006, 20:19:19 UTC - in response to Message 6555.  

got that problem also alot lately.

only thing that helps is restarting it :/


Also had it a number of times in the last few days to the determent of one wu loosing heartbeat and crashing whilst almost finished.... unloading the firewall and restarting it restored stability..... the usual, new science version and a buffer overflow process it did not like. Manually set permissions and it went away.

Coelum Non Animum Mutant, Qui Trans Mare Currunt
ID: 6570 · Report as offensive
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6578 - Posted: 22 Nov 2006, 19:59:28 UTC

re-install/fix did not clear the problem. going to do a system re-boot (not been done lately). going to have a long weekend, hope it does not lock/lose connection. :(
ID: 6578 · Report as offensive
Profile Dingo
Avatar

Send message
Joined: 7 Dec 05
Posts: 9
Australia
Message 6582 - Posted: 23 Nov 2006, 1:51:30 UTC - in response to Message 6504.  

for the past few days 11/15/06 the BONIC Manager can not fine the local host. all processing stops. running 5.4.11. on 24/7. running 4 or more projects round robin. only way to get working again is to exit and re-start. any idea why this is happing. any one else have the same issue?


I have a PC where the manager cannot connect to the client. I know the problem is memory related. It is an old celeron 1.7 XP Home and it just runs out of memory, it has. I tried everything till I put another stick of memory into it and it connected immediately. While it was low I couldn't do much at all even a net time command would error with not enough resources

How much memory has the PC got and do you have the WU's kept in memory when suspended ??




Proud Founder and member of



Have a look at my WebCam
ID: 6582 · Report as offensive
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6674 - Posted: 27 Nov 2006, 18:25:41 UTC

project again shut-down over the long wk-end. pc is a P4 3.2 512mb ram. which project was causing memory problem? lost over 190 hours of crunch time.
ID: 6674 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 14625
Netherlands
Message 6675 - Posted: 27 Nov 2006, 18:28:48 UTC - in response to Message 6674.  

We don't know which projects you are attached to. So why don't you tell us which project you have the problems with?

ID: 6675 · Report as offensive
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6676 - Posted: 27 Nov 2006, 19:15:39 UTC

rosetta, climateprediction.net, einstein and SETI.
ID: 6676 · Report as offensive
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6701 - Posted: 29 Nov 2006, 15:45:38 UTC

after re-loading/repair of bonic.exe and a system re-boot, still getting lost connections a few times a day. I am now trying halting new work one project at a time to see which one is the causing the trouble. will let you know what I find.
ID: 6701 · Report as offensive
Blainer
Avatar

Send message
Joined: 30 Nov 06
Posts: 3
Canada
Message 6703 - Posted: 30 Nov 2006, 14:49:40 UTC
Last modified: 30 Nov 2006, 15:04:48 UTC

I'm having the same problem. Manager is suddenly unable to find localhost after a while. Unsure exactly how long because I only notice when I wake up in the morning, or I get home from work. Every window becomes blank as well, so I can't see anything in the Messages window that might give me a hint as to what is causing this.

Running SETI, Einstein, and Rosetta.

Have ZoneAlarm installed, with all 3 BOINC-related programs having permissions (don't use the screensaver), and the client programs. Opened port 31416 so that shouldn't be blocked. Not running other programs that might be using that port, either.

EDIT: Just noticed something else new. WindowBlinds is suddenly not skinning Boinc Manager. Used to skin it fine.
ID: 6703 · Report as offensive
Blainer
Avatar

Send message
Joined: 30 Nov 06
Posts: 3
Canada
Message 6706 - Posted: 30 Nov 2006, 17:12:45 UTC - in response to Message 6703.  

EDIT: Just noticed something else new. WindowBlinds is suddenly not skinning Boinc Manager. Used to skin it fine.


Ignore that.. it's back to skinning it properly now. <confused look>

ID: 6706 · Report as offensive
Les Bayliss
Help desk expert

Send message
Joined: 25 Nov 05
Posts: 1486
Australia
Message 6708 - Posted: 30 Nov 2006, 17:43:12 UTC

Blainer

Not much help for your problem, but the reason "all the windows go blank", is that this info is coming from the 'hidden' part of BOINC, the "worker daemon".
The function of the gui part, is to pass on to the user what the the worker is doing.
If the two parts aren't communicating, then there's nothing to pass on, so "blank".

The most common cause of the loss of communication is that something else has taken over the port that they use, 31416.
Some Windows programs do this on startup, and Zonealarm also blocks it if it's not told to let the programs through. Perhaps a recent update of some program has started the problem.

It's possible, from the number of reports lately on this problem, that something else has started to be a problem.

... so I can't see anything in the Messages window that might give me a hint as to what is causing this.

An archive of BOINC messages and errors are stored in stdoutdat.txt, which is in the BOINC folder, and a list of just the errors in stderrdae.txt
ID: 6708 · Report as offensive
Blainer
Avatar

Send message
Joined: 30 Nov 06
Posts: 3
Canada
Message 6711 - Posted: 30 Nov 2006, 20:30:08 UTC - in response to Message 6708.  
Last modified: 30 Nov 2006, 20:37:18 UTC


It's possible, from the number of reports lately on this problem, that something else has started to be a problem.

An archive of BOINC messages and errors are stored in stdoutdat.txt, which is in the BOINC folder, and a list of just the errors in stderrdae.txt


After looking through the stdoutdae.txt file, every time there's been an apparent stoppage of BOINC (the next messages are when I restarted BOINC manually), Rosetta@home downloading something was the last message saved:

2006-11-21 18:38:27 [rosetta@home] Started download of file boinc_hom012_aas014_09_05.200_v1_3.gz
2006-11-21 18:38:29 [rosetta@home] Finished download of file boinc_hom012_aas014_09_05.200_v1_3.gz
2006-11-21 18:38:29 [rosetta@home] Throughput 208015 bytes/sec
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2006-11-21 23:43:42 [---] Starting BOINC client version 5.4.11 for windows_intelx86
...
2006-11-24 15:32:02 [rosetta@home] Started download of file 1agr_1_idid_model_13_0001_idl.pdb.gz
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2006-11-24 18:49:22 [---] Starting BOINC client version 5.4.11 for windows_intelx86
...
2006-11-28 08:07:56 [rosetta@home] Started download of file PSH_0071aa1d5mG03_05.200_v1_3.gz
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2006-11-28 18:58:20 [---] Starting BOINC client version 5.4.11 for windows_intelx86
...
2006-11-29 15:08:17 [rosetta@home] Started download of file PSH_0104_descriptionfile.txt
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2006-11-29 18:52:22 [---] Starting BOINC client version 5.4.11 for windows_intelx86
...
2006-11-30 04:05:32 [rosetta@home] Started download of file PSH_0117_3913.pdb
To pause/resume tasks hit CTRL-C, to exit hit CTRL-BREAK
2006-11-30 07:53:07 [---] Starting BOINC client version 5.4.11 for windows_intelx86


Prior to the last Rosetta@home messages, BOINC appears to have been operating normally. Scheduling, uploading, downloading, etc, all happened properly.

Empirically, this seems to point to a problem with Rosetta: either Rosetta itself, or with how it and BOINC are interacting.
ID: 6711 · Report as offensive
Profile Jord
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 29 Aug 05
Posts: 14625
Netherlands
Message 6731 - Posted: 1 Dec 2006, 20:17:12 UTC

ID: 6731 · Report as offensive
RamCharger

Send message
Joined: 17 Nov 06
Posts: 10
United States
Message 6732 - Posted: 1 Dec 2006, 20:23:46 UTC

Rosetta looks to be the cause as reported below, after a day and a half of Rosetta being suspended, turn back on and w/ in a half a day BONIC would lock up, exit - restart all good again. have reset and suspended Rosetta for the weekend and next monday 12/4 will turn Rosetta back on to see if it fails again.
ID: 6732 · Report as offensive
Brian B

Send message
Joined: 18 Dec 06
Posts: 10
United States
Message 6990 - Posted: 18 Dec 2006, 7:14:07 UTC
Last modified: 18 Dec 2006, 7:14:18 UTC

Hi all. I posted a related error on this thread, Couldn't delete file.... Hope this helps.
ID: 6990 · Report as offensive

Message boards : BOINC Manager : lost connection to local host

Copyright © 2021 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.