I did a bug lookup and found burt 187044. It says this is typically client or network side issue:
rshd is expecting a port number for standard error from the client. The client
did not return one within 60 seconds, thus rshd timed out with the error
message.
In /etc/messages, there is also:
Tue Nov 8 13:24:03 IST [rshd_0:info]: couldn't connect second port on cornice
Tue Nov 8 13:24:10 IST [rshd_1:info]: couldn't connect second port on cornice
Again this implies that there are no stderr ports on the client cornice.
%%% TITLE: rsh is timing out when reading port number for standard error
%%% DESCRIPTION:
The rsh protocol needs a stdout and stderr port. When the rsh client and rshd
on the filer are connecting, rshd asks for a stderr port from the client. rshd
waited 60 seconds for a response and did not receive one, so rshd timed out with
the message: "[rshd_0:error]: rshd: when reading port number for standard error
from <client>, it didn't arrive within 60 seconds."
%%% WORKAROUND:
There is no workaround.
%%% NOTES:
There are some causes for this problem, one is network configuration, the filer
cannot get a return path to the client. Also the client's reserved ports
may be heavily used. Do a netstat -a to determine this.
