Hello,
I am trying to figure this out - our active/passive Clustered Windows 2003
file servers keep sending an RST (reset) packet after a random interval
(anywhere from 2-60 seconds), which causes OS X 10.4.8 & Linux clients
connected via smbclient to disconnect. See the following that was captured on
the Linux commandline:
smbclient //<servername>/<sharename>$ -W <domain> -U <username>
Password:
Domain=[<domain>] OS=[Windows Server 2003 3790 Service Pack 1]
Server=[Windows Server 2003 5.2]
smb: \> read_socket_data: recv failure for 4. Error = Connection reset by peer
What appears after "smb: \>" only shows up after the connection has been
severed. Prior to the connection reset message, it is possible to execute any
number of commands (dir, cd, etc) without failure.
I found a few references to this error showing up in samba version 3.0.6 and
later, but no one seemed to actually have an answer, except perhaps to roll
back to 3.0.5 or upgrade to Linux kernel 2.6x, which really are not options
for us. From what I was able to find, no one was able to definitively point
to why failure occurs.
How do you enable extremely Verbose logging for CIFS/SMB on a windows server?
Thanks in advance,
pj_servadmin
|