FQDN not working vs Resolved IP address

[rockfort]

I created a new FQDN address object to facilitate a new Policy(rule).

 When tested the FQDN resolves internal to the Palo Alto Firewall.

The rule contains one destination address which is the new company.fqdn.com FQDN

The rule contains one source address

Application SSL with Application-Default Service

Action Allow

When attempts are made to connect  to this destination via the new rule with the FQDN set(destination), the traffic is denied(fails) and logs point to(identify) the "interzone-default" rule instead of the "new rule" that is set up to facilitate this connection 

But when I replace the FQDN(destination) with it's resolved IP in the new rule, it works fine(allowed) and logs point the occurrence to the "new rule" (not the interzone-default) as to be expected since that is normal behavior 

Questions:

1. Why would the interzone-default rule become a part of the failed attempt to connect to the new rule
2. Anyone know why connection fails with the FQDN set as destination rather than it's resolved IP address
3. As anyone had a similar experience

Thanks in advance.

[BPry · Accepted solution]

@rockfort 

Questions:

1. Why would the interzone-default rule become a part of the failed attempt to connect to the new rule

Because the new rule isn't properly matching the traffic. I would verify with the 'request system fqdn show' or ''show dns-proxy fqdn all' depending on your currently installed version of PAN-OS to verify that the firewall is actually properly resolving the FQDN object to the proper address.

2. Anyone know why connection fails with the FQDN set as destination rather than it's resolved IP address

99.8% of the time, this is due to the FQDN object either not refreshing properly or the rule not properly being built to accommodate for the traffic that's actually being seen by the firewall. Once you've verified the FQDN object is resolving properly, you'll want to test the rulebase entry and look at the recorded logs and make sure that your rulebase entry as configured properly accounts for the traffic. 

[rockfort]

Thank you very much

[DhananjayBhakte]

I Have similar problem however in my case on firewall when I create fqdn object it resolve to  IP and when I use same object in rule that rule doesnt work, then I need to add IPs manualy based on deny logs in same rule.

I observe one strange thing when I resolved fqdn address object it resolved to some different IPs and when I nslookup fqdn in cmd it resolved to some  completely different subnet  IPs and same IPs when I allowed in rule then rule works.

I guess  firewall is not resolving to correct IPs .

[DevonFan]

I also have the same problem...and follow your steps to try to troubleshoot the problem but encounter a problem.
1. I have verified "show dns-proxy fqdn all" in the command line and confirmed that the fqdn has a proper match to the proper IP address.
2. How can I test that the firewall's rulebase resolves to the proper ip?

[landoa]

We are also having a similar problem with the FQDN in an address object.

For a given fqdn, the result of nslookup is different from "show dns-proxy fqdn all" which is also different from the "resolve" of the fqdn in the address object of the gui.

Other fqdn's appear to work OK, but the associated IP's may be static and the IP list much shorter.

We are using DNS Proxy and the cache is activated.  We had TTL activated as well, but i've removed it as its not necessary.

Shall we try deleting the address object and recreating?  How can we resolve this?

Thanks

[landoa]

I also tried the fqdn force refresh command, to no avail.

However, I just saw this article that states that only "up to 32 addresses" are mapped to an FQDN used in an address object:  https://knowledgebase.paloaltonetworks.com/KCSArticleDetail?id=kA10g000000ClHJCA0

So, what does this mean for the fqdn we are using?  edge-teams.plcm.vc (https://mxtoolbox.com/SuperTool.aspx?action=a%3aedge-teams.plcm.vc&run=toolpage) - goes way over the 32 limit.

Thanks