[Pkg-samba-maint] Bug#341126: "Bug" still present ("smbclient fails with 'session setup failed: SUCCESS - 0'")?

[Oskar Liljeblad]

What can I say - it still happens. Here are some timestamps from last time
it occurred:

-rw-r--r-- 1 root root 12431 2011-06-30 13:10 log.smbclient-failure-15491
-rw-r--r-- 1 root root 12431 2011-06-29 11:35 log.smbclient-failure-1055
-rw-r--r-- 1 root root 12429 2011-06-23 08:05 log.smbclient-failure-85942
-rw-r--r-- 1 root root 12431 2011-06-22 21:25 log.smbclient-failure-29351
-rw-r--r-- 1 root root 12429 2011-06-21 13:35 log.smbclient-failure-55561
-rw-r--r-- 1 root root 12429 2011-06-17 11:55 log.smbclient-failure-2754
-rw-r--r-- 1 root root 45437 2011-06-15 14:20 log.smbclient-failure-53373
-rw-r--r-- 1 root root 12429 2011-06-09 11:55 log.smbclient-failure-63856
-rw-r--r-- 1 root root 12429 2011-06-09 11:20 log.smbclient-failure-17381
-rw-r--r-- 1 root root 13741 2011-05-31 15:15 log.smbclient-failure-76050
-rw-r--r-- 1 root root 13743 2011-05-31 09:45 log.smbclient-failure-85956

The servers are still W2K3.  My only problem is with the mssage - "session
setup failed: SUCCESS".  Such a message indicates a bug in the code.

Unfortunately I still don't know how to reproduce them.

Regards,

Oskar 

On Monday, June 06, 2011 at 07:32, Christian PERRIER wrote:
> tags 341126 moreinfo
> thanks
> 
> Hello,
> 
> You reported this bug a quite long time ago against smbclient. From
> the bug log you sometimes were'ntable to connect to a high-load W2K3.
> server and the faillure happened with a not very clear message.
> 
> Since then, several samba versions were released. Also, in the
> original bug, I see the horrible "socket options =
> TCP_NODELAY"...which is found on dozensof web pages but is still as
> wrong as it can be.
> 
> Samba devels have said and written many times that these very specific
> parameters should NOT be used in production environment except for
> very specific needs. Using them  randomly is often like shooting self
> in feet.
> 
> So, do you still experience these intemittent connection problems in a
> way that could be investigated?
> 
> If there is no way to reproduce the bug, I'll consider closing it.
> 
> Many thanks in advance for your answer
> 
> -- 
> 
> 
>