Azure DNS: Intermittent Name Resolution Failures in Production Virtual Network

Problem description

I am experiencing sporadic DNS resolution failures from workloads running in my production virtual network, specifically affecting Azure Container Apps and other resources. The errors encountered include getaddrinfo returning EAI_AGAIN and EAI_NONAME, leading to unsuccessful resolution of both public domain names and Azure private endpoint names. Previously, name resolution was successful, but now it fails intermittently during certain windows.

Environment

The issue affects a production virtual network, with at least one impacted workload being an Azure Container Apps replica in that network. The failures involve both public DNS names and private-link names, relying on both public DNS resolution and private DNS integration. Specific regions and DNS configurations are not detailed in the case.

What I've already tried

I have documented the affected network, impacted workload, and the error messages. I reviewed DNS settings, resolver reachability, forwarding behavior, network rules, private DNS zone links, and DNS cache behavior. No configuration changes or tests have been conclusively linked to the issue, and no email troubleshooting outputs or diagnostic captures are provided.

Current status

The DNS failures continue to occur sporadically, with errors during name resolution for multiple namespaces, including Azure and private zones. The suspected cause is an issue within the node-local DNS or resolver path in the affected Container Apps environment. I am seeking assistance to escalate this to the Container Apps engineering team for an inspection of the internal DNS infrastructure during outage windows.