error: (98, 'Address already in use')

[anonymous]

Hi folks,

i am stuck at an error that makes no sense to me. Please help :)

I got a debian system installed on an Icybox NAS1000 ( http://raidsonic.de/de/pages/products/external_cases.php?we_objectID=4444 ) that has an ARM working inside. Installing the cryptonas-server using apt and the *arm.deb package worked fine. After starting the daemon i get this error message. (Theres a few seconds delay between starting up and getting the error!)

Stopping CryptoBox Daemon (webinterface):. Starting CryptoBox Daemon (webinterface):. nas:~# Exception in thread CPHTTPServer Thread-1: Traceback (most recent call last):

File "threading.py", line 442, in bootstrap

self.run()

File "threading.py", line 422, in run

self.target(*self.args, **self.kwargs)

File "/var/lib/python-support/python2.4/cherrypy/_cpserver.py", line 114, in _start_http

self.httpserver.start()

File "/var/lib/python-support/python2.4/cherrypy/_cpwsgiserver.py", line 301, in start

raise socket.error, msg

error: (98, 'Address already in use')

I tried a lot of things, including setting den --host value to "nas", "localhost" and "10.0.0.100" (which is the IP number of the box running cryptonas) in /etc/default/cryptobox-server

The port (here: 80) is definitely unused! I tried netstat -lpn and i also tried a lot of other ports. From a remote machine i always see the selected port open when using "nmap 10.0.0.100"! Being very unlucky with my situation i also used "nc" and tried to establish a HTTP connection by hand, but there is no response. It is also no firewall problem, because "lynx localhost" is still "waiting for response".

There must be something i am not able to see. Light the blind, please, and thanks for all help!

[anonymous]

A hard nights results: Main error was (and still is) a bug that causes dm_mod not to load properly into the kernel.

Strange things happen...

[lars]

Hi, thanks for reporting back regarding your problem!

Maybe you can tell us, how this module loading problem interfered with the socket binding of the cryptobox? I am curious - especially if this could help us to squash an evil bug ...