Default read timeout in 2.11.2

I've been following the threads on github about the default read timeout introduced in version 2.11.2. v2.11.1 lingeringTimeout can break some connections · Issue #10598 · traefik/traefik · GitHub & Traefik Migration Documentation - Traefik for reference.

I'm wondering if there is a prometheus metric that would help us understand the impact of this change prior to making the upgrade? For instance traefik_entrypoint_request_duration_seconds_bucket sounds like it might, but it's not clear to me if this would capture just the phase of the request affected by the read timeout, or the entire request end-to-end.

Also curious if anyone has an example of modifying the default timeout in the Helm chart values.yaml.

Thanks!

2 Likes