Wireless Network error "Disabled gateway (bad DNS)"

[JonnoKiwi]

Clients on our wireless network has been experiencing an issue where they are not able to go out onto the internet due to a DNS error. On the Connectivity timeline reads "Disabled gateway (bad DNS)". At first we thought this meant our local DNS had some issue, but upon changing the DHCP settings of the wireless network to use Google's DNS (8.8.8.8), the error messages still read the same.

What could be the cause of this intermittent problem? 

Note: outages last around 12 minutes and may occur between 1 to 4 times within 24 hours.

[Ryan_Miles]

That indicates the AP itself is having issues reaching the DNS server. Have you tried other DNS IPs (8.8.4.4, 1.1.1.1, or whatever your provider's DNS is)? Are there other Meraki devices on the network? And if yes, do any of them have the same problems?

[JonnoKiwi]

We only use Meraki on this network. When the error had occurred, it only affected the access points (MR52, MR56). As of this week, the Bad DNS error has not returned. We are still using 8.8.8.8 and 4.4.2.2 as the DNS for the for all WiFi DHCP. I will try 8.8.4.4. and 1.1.1.1 in the scenario that the error comes back. This way I can cross reference the outcomes. Thank you for your input, Ryan!

[PhilipDAth]

Check out the DNS servers being used by the APs.

[JonnoKiwi]

This was the first thing we did. We started using public DNS servers such as 8.8.8.8 and 4.4.2.2. We did have some errors one week using these DNS server, but as of this week, there has been no problems. I am continually monitoring the network. 

[BlakeRichardson]

Has anything changed on your firewall recently that would be blocking DNS services for the device?

[JonnoKiwi]

No, there has been no changes to any of the policies nor has there been updates in the version. If there had been, wouldn't the result be more wide spread and not intermittent?

[CraigBowden]

We had this issue for a while in addition to losing connectivity to the Meraki dashboard randomly.  Strangely enough the fix was to start clearing the ARP and MAC address tables in our core switches (End of life Nexus 3K).

[randerson73]

Been having the same issue when I added in my own/public DNS severs. I changed it back to "Proxy to upstream DNS" and it seems to have stabilzed. I do have 2 WAN connections on the MX85 so maybe when traffic routes back and forth between the separate WAN links, its having issues. Not sure yet but keeping an eye on it. 

[WalterBajana]

I had the same issue. I did an extended ping to the AP and could see that I get packet loss. 

[GIdenJoe]

Usually when your AP's  have this issue they are not lying.  If you would do a packet capture on the wired port you would see the AP actually sending DNS requests periodically and if they don't get answered then the problem is upstream in the network.