You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Added a DNS health check system, that temporarily removes unhealthy endpoints (for A, AAAA and CNAME RR types) from DNS responses for a host (HTTP and HTTPS checks are supported via the dns.health.http configuration property).
Improved the cache system, which is now able to serve stale cache if the Redis backend is unreachable.
Cached records are now kept up-to-date with Redis in the background, saving precious milliseconds in the main DNS responder thread (with the redis.cache_refresh_seconds configuration property).
Fixed an implementation mistake in the DNS responder, that resulted in a NOERROR with no records if Redis is unreachable, where the correct answer was SERVFAIL.
Fixed a bug in the Redis backend connector; if Redis went down and Constellation failed accessing Redis as many times as there were connections in the Redis pool, all further attempts would hang and timeout (where they should have failed fast with a Disconnected internal error).
Fixed a bug in the DNS NOERROR and NXDOMAIN response types that are sent depending on whether record exists or not (regression introduced in v1.7.0 with the cache system).
Bumped all dependencies (except trust-dns and trust-dns-server due to conflicts).