I wonder why the sudden decrease and which number is correct…
1 Like
Are you running an authoritative server for the pool, right?
If so, you are serving resolvers, and resolvers can be quite peculiar when they choose which authoritative server to query.
(we did a study on this before, and the short answer is that they stick more with low-RTT authoritative servers, from the resolver vantage point. See Mueller17b.pdf).
I am talking about queries per second to the NTP pool DNS servers as shown at https://status.ntppool.org/
That doesn’t seem right. Anyway, NTP Pool DNS servers are responding normally.
The average monthly DNS queries is 45,544 per second.
indeed strange.
I have no idea.
I checked the DNS records TTLs, to see if they had been increased (and thus increasing cache hit at the resolver side), but I don’t see any TTL changes
I upgraded over the last weeks the infrastructure that’s counting and collecting these metrics to use mTLS everywhere and I had some bugs to squash for automatically reloading the certificates when they renewed. It’s finally been working properly since around the middle of last week after I fixed the last bug I found in the process.
1 Like