REMSH does not work intermitantly

[timwescott]

From time to time the 'remsh' command fails to work. When executed, there is no response from the destination server and the command eventually times out. This is on HP-UX 10:20

'rlogin' and 'ping' work from the source server to destination and all commands, including 'remsh' work in the opposite direction.

Any ideas why? + and how to resolve it without having to reboot the server.

Thanks
Tim

 

[Astur]

Hello Tim,

a possible cause for this behavior is
name resolution. remsh depends heavily on this.
Review if you are using a name server and
if it is unavailable when you see these problems.

If that was the case, you can define locally
the name of the remote server in your /etc/hosts
and check also the search order for name resolution.

regards



Jose M. Blanco
MVP BrainBench for Unix admin.

 

[timwescott]

Managed to resolve this. Using 'netstat -nr' I found that the destination server had a route via our FireWall. Don't know why but it did. This did not stop connectivity using the command 'remsh server_name' but it did with the command 'remsh server_name command'. The reason for this is that when 'remsh' initiates it uses port 515 but the result of the command is returned on port 1024. If 1024 is busy then it tries 1023, then 1022 etc. These ports are blocked on our FireWall, which produced the hanging effect.

The route to this server should not be bounced off the FireWall so I used the 'route delete' command to remove the rouge route statement.

All works fine, no reboot required.

Thanks for any replies
Tim