The Samba-Bugzilla – Bug 3131
connection reset by peer/getpeername failed. Error was Transport endpoint is not connected
Last modified: 2005-10-04 04:43:05 UTC
I am getting this kind of error messages, any clues?
[2005/10/03 17:26:55, 0] smbd/service.c:set_admin_user(321)
vmh logged in as admin user (root privileges)
[2005/10/03 17:26:55, 1] smbd/service.c:make_connection_snum(705)
nnc-anne-hp (192.168.1.50) connect to service vmh initially as user vmh
(uid=0, gid=524) (pid 9801)
[2005/10/03 17:27:03, 1] smbd/service.c:close_cnum(887)
nnc-anne-hp (192.168.1.50) closed connection to service vmh
[2005/10/03 17:29:12, 0] lib/util_sock.c:get_peer_addr(952)
getpeername failed. Error was Transport endpoint is not connected
[2005/10/03 17:29:12, 0] lib/util_sock.c:write_socket_data(388)
write_socket_data: write failure. Error = Connection reset by peer
[2005/10/03 17:29:12, 0] lib/util_sock.c:write_socket(413)
write_socket: Error writing 4 bytes to socket 17: ERRNO = Connection reset by peer
[2005/10/03 17:29:12, 0] lib/util_sock.c:send_smb(605)
Error writing 4 bytes to client. -1. (Connection reset by peer)
Filesaving is not working correctly.
error only exits, when user is logged to samba pdc, not without it.
The client disconnected. Note that WinXP clients will connect
to ports 139 and 445 concurrently and disconnect from 139 if 445 is
available. This in itself can generate extra noise in the logs.
You can try setting 'smb ports = 139' to quiet the noise.
Otherthan that you can try looking for bad networking hardware,
switches, hubs, mismatched duplex settings. etc....
There is no evidence to point at a bug in Samba.