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
{{ message }}
This repository has been archived by the owner on Sep 7, 2023. It is now read-only.
A race condition problem in the linux conntrack (only recently fixed by Martynas Pumputis and available upstream only in really recent kernel versions) makes the DNS resolution hang for 5s (def timeout) and then eventually fail.
We experienced this problem as well and once we moved away from alpine our DNS resolutions error rates dramatically when down.
Use alpine should not influence consul internals but def external service definitions/checks.
Should consul consider to change base image to something like debian:stable-slim ?
The text was updated successfully, but these errors were encountered:
Alpine musl implementation is well knows for his parallel DNS resolution
https://github.com/gliderlabs/docker-alpine/blob/master/docs/caveats.md
A race condition problem in the linux conntrack (only recently fixed by Martynas Pumputis and available upstream only in really recent kernel versions) makes the DNS resolution hang for 5s (def timeout) and then eventually fail.
For reference this is just one of the hundred of blog post that talk about this problem:
https://medium.com/techmindtickle/intermittent-delays-in-kubernetes-e9de8239e2fa
We experienced this problem as well and once we moved away from alpine our DNS resolutions error rates dramatically when down.
Use alpine should not influence consul internals but def external service definitions/checks.
Should consul consider to change base image to something like debian:stable-slim ?
The text was updated successfully, but these errors were encountered: