remote shell command fails every third attempt.

[vincent Carpentieri]

Does anyone have any info why remote shell command would fail (timeout) every third attempt. This is only happens on our Tru64 cluster (v5.1B).
example : /usr/users/plat> rsh han ps -aef
han: Connection timed out

[Johan Brusche]

Is 'han' the cluster alias or a member name ?

Is the selection weigth on the alias =2 ?
Do all members have the same selw/selp/rpri
for that alias ?
'netstat -rn' from all members ...

Johan.

[vincent Carpentieri]

han was the target node. I will compare the setting you mentioned.

[Ralf Puchner]

have a look into the troubleshooting section of the cluster admin guide.

[vincent Carpentieri]

Thanks all, we have a work around in place for now. To work around the issue, we removed out_alias from shell in etc/clua_services.
HP/Compaq/DEC wants us to setup tcpdump and look at the routing . .