[Samba] win 7 client can't map drive: getpeername failed

Fri May 10 08:01:50 MDT 2013

I think the "Error was Transport endpoint is not connected" warnings are 
sometimes misleading. Do you have any control over the samba config 
(smb.conf) on the NAS ?    On regular samba installs, changing the 
default port settings can cause more problems.

Windows 7 will try to connect on port 445  (SMB or CIFS over tcp/ip), 
and will then reconnect to ports 137/138/139 (SMB over netbios over 
tcp/ip) since samba 3.x doesn't handle the newer 
SMB-over-tcp/ip.         Disabling 445 on the server seems to cause more 
problems than it solves.

Are you able to connect via IP ?  e.g net use \\qnap_ip\share ?

I had problems in the past when I disabled port 445 on samba servers.  
Remote users (no netbios broadcasts permitted) could connect via IP but 
not via name.     For the name only connections, packet monitoring would 
show packets getting thru the the server but the exchange between client 
and server not being completed.  For clients connecting via IP, the 
client would send packets to server, server respond, and then clients 
responded.

On 05/07/13 03:53, Ed Strong wrote:
> Hi,
>
> I'm re-posting this (with some more info) as I don't think the original got
> through as I wasn't
> signed up to the samba list.
>
> this is my first foray in samba (and newsgroups) so go easy :)
> I've started reading the o'reilly samba book but finding it hard going.
>
> Anyway I'm trying to map a network drive from a windows 7 pro client to a
> QNAP NAS with the command:
>    net use s: \\qnap\share
>
> I've posted on several forums and got good advice but the problem remains.
> Rather than repost all the detail, please see my original posts:
>
> http://forum.qnap.com/viewtopic.php?f=185&t=74639
> http://social.technet.microsoft.com/Forums/en-US/winservergen/thread/11d35b0c-ac95-489f-b5d1-0486b9774603
> http://www.edugeek.net/forums/windows-7/112309-map-network-drive-nas-but-get-error-64-58-a.html
>
> I've managed to ssh onto the QNAP via putty and found this in the logs
> (getpeername failed)
>
> [/var/log] # pwd
> /var/log
> [/var/log] # tail -f log.smbd
> [2013/05/01 09:36:17.135999,  0] lib/util_sock.c:474(read_fd_
> with_timeout)
> [2013/05/01 09:36:17.136096,  0]
> lib/util_sock.c:1440(get_peer_addr_internal)
>    getpeername failed. Error was Transport endpoint is not connected
>    read_fd_with_timeout: client 0.0.0.0 read error = Connection reset by
> peer.
> [2013/05/01 09:36:17.137700,  1] smbd/server.c:299(remove_child_pid)
>    Scheduled cleanup of brl and lock database after unclean shutdown
> [2013/05/01 09:36:17.178522,  1] smbd/service.c:1073(make_connection_snum)
>    172.24.120.139 (172.24.120.139) connect to service Staff initially as
> user DOMAIN+admin (uid=10001423, gid=10000514) (pid
>
> 25771)
> [2013/05/01 09:36:17.179093,  0] lib/util_sock.c:474(read_fd_with_timeout)
> [2013/05/01 09:36:17.179173,  0]
> lib/util_sock.c:1440(get_peer_addr_internal)
>    getpeername failed. Error was Transport endpoint is not connected
>    read_fd_with_timeout: client 0.0.0.0 read error = Connection reset by
> peer.
> [2013/05/01 09:36:17.179289,  1] smbd/service.c:1254(close_cnum)
>    172.24.120.139 (172.24.120.139) closed connection to service Staff
> [2013/05/01 09:36:37.142714,  1] smbd/server.c:272(cleanup_timeout_fn)
>    Cleaning up brl and lock database after unclean shutdown
>
>
> The QNAP's samba version appears to be 3.5.2:
>
> [/var/log] # ps -ef | grep smb
>   4016 admin      3104 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4017 admin      3728 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4366 admin      1840 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4877 admin      3300 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4902 admin      3952 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4978 admin      4132 S   /usr/local/samba/sbin/smbd -l /var/log -D -s
> /etc/config/smb.conf
>   4979 admin      3356 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4980 admin      1224 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   4995 admin      1016 S   /usr/local/samba/sbin/smbd -l /var/log -D -s
> /etc/config/smb.conf
>   5063 admin      2068 S   /usr/local/samba/sbin/winbindd -s
> /etc/config/smb.conf
>   9509 admin      1664 S   /usr/local/samba/sbin/nmbd -l /var/log -D -s
> /etc/config/smb.conf
> 25540 admin       544 S   grep smb
> [/var/log] # /usr/local/samba/sbin/smbd -V
> Version 3.5.2
>
>
> I've also installed MS network monitor on two clients and did a capture
> whilst running the command
>     net use s:\ \\saturn\staff
>
> I've posted three screenshots here:
>
> https://plus.google.com/photos/108734482620454690509/albums/5875135861918839393?authkey=CJ3lwKu2xJqMyQE
>
> Basically, Worked.png shows the SMB frames on a PC where the net use
> command worked
> and Failed.png shows the SMB frames on a PC where the net use command did
> not work
>
> It looks to me like the first 6 SMB frames are identical. Then things start
> to change
>
> On the working client we continue with frame 10113 which is a
>    Dfsc: Get DFS Referral Request
>
> but the failing client continues with some TCP frames (see
> tcp-frames154-157.png) 154 to 157
> before it seems to start the negotiation again at frame 158
>
> Not sure how to troubleshoot this further so any advice welcome.
>
> Thanks
> Ed
>
> PS I initially tried to post this on google group linux.samba but was
> rejected by the
> moderation robot which said "Please submit your message to the mailing list
> address".
> I did this with attached png's but failed due to file size so hopefully 3rd
> time lucky!