Client thread fails to resolve IP address – goes to error proto status forever

[William Caughey]

Greetings all,

I have seen where occasionally, for unknown reason, a pdl-tcpip client thread will go into an “error” status with an Error Msg of: Can’t get IP address for ‘hostname’: WSAHOST_NOT_FOUND — where hostname is the hostname it can’t get the IP address for.

Restarting the thread will get it working until it fails again.

If the thread is configured with an IP address instead of a hostname it seems that the Cloverleaf engine will retry to connect “forever” but if this particular failure mode occurs “Can’t get IP address…” that the thread goes into an error status and never tries again.   I suspect that if the Cloverleaf engine simply retried the connection it would eventually resolve itself.

This behavior can be mitigated with an Alert to restart the thread if it goes into an “error” status but is there some other way to mitigate this?

Why does the Cloverleaf engine try to (re)connect to an IP address “forever” but only try once to connect to a hostname that can’t be resolved to an IP address?

Thanks,
Bill

 

Topic

[Vince Angulo]

I’m not sure if it will apply to a DNS induced error, but ensure the “Auto-reconnect” box is checked in the Protocol Properties options.

[William Caughey]

Thanks for the reply, Vince.

The “Auto-reconnect” box is checked.

Bill

[Author]

Replies