PDNS recursor server failure

[Aka217]

I am reaching out for assistance regarding an issue we have been experiencing on our two recursors servers for the past few days. For your information, our application, located on two distinct sites, queries the local recursor server on each site. Each recursor server is configured to query the authoritative server for its respective site, as well as the local database.

However, over the past few days, we have been encountering issues on both recursor servers, with "Server Failure" responses from the NS servers. No relevant information is found in the logs.

Do you have any suggestions or insights on how to resolve this issue?

Many thanks for your help

[omoerbeek]

1. Add relevant info: version and configuration.
2. Trace the query, see https://docs.powerdns.com/recursor/running.html?highlight=trace#tracing-queries and show the trace log here. Note that answers retrieved from the packet cache do not show up in the trace, so you might need to restart the recursor or clear the cache, see https://docs.powerdns.com/recursor/running.html?highlight=cache#cache-management.

[Aka217]

Hi omoerbeek,

Thanks for your reply.

We've identified that the server failure returned by the recursor originates from a failure on the authoritative server. When the authoritative responds with a failure, the recursor blacklists it for 1 minute and returns a local SERVFAIL.

We're attempting to trace the issue on the authoritative side, but the high traffic and lack of regex filtering make it challenging.

Is it possible that the authoritative server returns a local SERVFAIL when it's unable to reach the PostgreSQL DB?

Note: caching is currently disabled on the authoritative.

Thanks for your support.

[omoerbeek]

Why would you disable caching? But I'm not a authoritative server expect. Start a new discussion, showing configuration, log files and statistics.

[Aka217]

Thank you, I'will start new discussion.