Web Gateway: Troubleshooting Next Hop Proxy Issues
We forward a complete internal domain to a next hop proxy. During a troubleshooting with a packet capture, we noticed that the MWG is still performing DNS requests for the hostnames accessed in that domain (receiving only negative responses since this is an internal domain and the configured DNS servers are "internet only").
That doesn't prevent the proxy chaining to happen, but we are wondering:
- why those DNS requests DO happen, since they are not needed?
- how can we avoid them?
- could this have a negative impact on performance (even if the DNS servers are quite fast to respond negatively, around 2ms) ?
To be more specific, and to avoid any interaction with the dynamic categorization process or the anti-malware, we have put the rule with the "Enable Next Hop Proxy" event as the very first rule in the MWG policy, with a "Stop cycle" action. We have been able to confirm with a troubleshooting that this rule - and only this rule - is matched.