telnet & rlogin not working for MS/SG cluster nodes.

[Gino Castoldi_2]

Hi,

2 servers: HPUX 11.11 MC/SG 11.14

We can not telnet from one cluster node to another. We also can not rlogin from one cluster node to another. But, we can run 'cmviewcl -v' and it works on either cluster node.

Any ideas on what to look for?
10 points to any good answer.

Thank you
Gino

[Steven Schweda · Accepted solution]

> We can not telnet from one cluster node to
> another.

What happens when you try? (Hint: "It
doesn't work" is not a helpful answer.)

[Gino Castoldi_2]

Yes, you are right. It doesn't connect and the message is "Trying..." then it stops.


10 points to any good answer.
Thank you
Gino

[Christian Tremblay]

check if your telnet line in /etc/services is commented out.

[spex]

Gino,

It sounds like a DNS, routing, or firewall issue. Are you able to telnet or rlogin by IP address?

PCS

[Gino Castoldi_2]

Hi,

Here is something strange.
Telnet and rlogin are working now. Not sure why it works now since no changes or reboots were made. When I run traceroute it
shows a different route then what one would expect. Observation: Since both servers are on the same network a traceoute would (normally?) not need to go over several different networks but in our case it does.



10 points to any good answer.
Thank you
Gino

[George Liu_4]

what "netstat -rn" on both systems?

[Christian Tremblay]

Looks like a temporary network glitch to me..

[spex]

If you were trying to connect by hostname instead of IP, and your DNS server(s) was/were down (and you lack entries in /etc/hosts), then the IP wouldn't have been resolved. Were you using hostname or IP to connect?

Also check for strange entries in your routing table:

# netstat -nr

You can add a static route from one node to the other using the 'route' command. See route(1M) for more information.

PCS

[George Liu_4]

I hope netstat -rn can reveal more info on the routing path. The reasons have 1. temprorary network failure like switch failover, 2. Since the systems are in the same network, so it is possible the system is multihomed and when the normal network path is down, the routing path could be altered to a dead end. The traceroute result seems verified that.

[Raj D.]

Gino,

Also dont forget to check syslog.log if at all any temporary network failure was happened. And as you have rebooted you should refer the OLDsyslog.log file.

Cheers,
Raj.

[Raj D.]

Sorry , you mentioned you didnot rebooted or changed anything , so as usual you can check the syslog.log file if you find any clue.

Cheers,
Raj.