Service timeout after certain duration

[Stanley Nguyen]

Hi,

I have configured a package with two services using same IP address. Both of these are web-based applications. Everything works fine except that after a while I can not connect remotely to my services using the service IP in specified control script anymore (connect timeout). I still am able to ping that IP address just fine. However, if I try to access to the service from the active node (locally), or using the physical IP to access to the service, then it works fine. I think there's something wrong with the virtual IP assigned in control script but not sure what.

Thanks,

Stanley

[Serviceguard for Linux]

The IP failover does just that, it activates that IP address on that server. It is not "in control" of the applications usage of the IP address.

If you can Ping the IP on that server, then that is correct. Are both services in the same package?

scripts and logs may help in this case

[Zinky]

Wow.. Isn't SG for Linux dead already and HP is recommending the native clustering solutions that come with Linux (aka RH CLuster Suite for Redhat flavours)?

Curious though -- is SG for Linux administered somewhat the same as SG for HP-UX? (I know the minor diffies will likely be in multipathing, eth bonding and LVM2, etc..) But is it somewhat the same?

[Serviceguard for Linux]

Yes - Serviceguard for Linux has been discontinued. I still have the ID.

SGLX and SG-UX are generally managed the same. There are certain differences. You pointed out some of them, and there are others. But general things like, how packages are configured, etc, are the same. The core products even had the same general code base, with "ifdef's" for most of the differences.

[Stanley Nguyen]

Might be DNS issue but it's working now