getting kicked out from the console.Getting below error:

Write failed: Broken pipe

[root@home ~]# cat /etc/ssh/sshd_config |grep ClientAlive*
#ClientAliveInterval 0
#ClientAliveCountMax 3
ClientAliveInterval 60
ClientAliveInterval 60

[root@home ~]# tail -f /var/log/messages
Mar 12 17:18:30 home rpc.statd[18915]: Version 1.2.3 starting
Mar 12 17:18:30 home rpc.statd[18915]: Flags: TI-RPC

Any immediate help would be appreciated....!!!

Yeah, a broken pipe is what you would get. This line
has cat piped to grep. But cat only lasts as long as your file, and then abruptly exits, leaving the pipe to grep broken.

You don't actually need the 'cat' at all. grep will read the file.

If you're scripting, there's some switch for grep that sets whether it found it or not (exits 0 or 1).

issue is here that SSH connection dies after few minutes.Leaving the below error:

-bash-4.1$ Write failed: Broken pipe

The cat would exit nearly instantly.

1 members found this post helpful.

Older default fedora sshd configurations would time out after some minutes and disconnect the client, leaving the same error message. It seems to me I had to change a setting in the putty connection settings to send a client keep alive packet (if you're using putty). There is an options setting in linux ssh to provide TCPkeepalive maybe it can help.

Anyway, current fedora settings don't seem to have this problem any longer, maybe you can compare your sshd_config with a new default sshd_config and see if there are any big differences. Nowdays I just open a default settings ssh connection and it stays open for days.

copied ssh files from other server but no luck even tried removing ssh and installed again.
using 6.4 RHEL 64 bit
ssh -vvv o/p
Last login: Thu Mar 20 18:53:31 2014 from
[root@host ~]# debug1: client_input_channel_req: channel 0 rtype keepalive@openssh.com reply 1
debug1: client_input_channel_req: channel 0 rtype keepalive@openssh.com reply 1
Write failed: Broken pipe

[root@local1 ~]# tail /var/log/messages
Mar 20 16:57:29 local kernel: sd 1:0:1:1: [sdba] Attached SCSI disk

It can be a hardware issue?

Nope. At least not at first glance.

ClientAliveInterval 60 will send a message every 60 sec if no activity is detected for a max of 3 times (default max count) i.e. 180 seconds. So after 3 minutes of no activity your connection should be automatically broken.

You can set the value higher say 600 (15 minutes) as well as configure the client to send a similar message i.e. SeverAliveInterval 600.

Michael,
Actually i have commented those entries.
What im looking for
My console should be alive even though if im not firing any commands aleast for one hour or more...!!!

Able to resolved it after uninstalled all the ssh* packages and installed it again and rebooted the server as getting I/O error.