Skip to content

HAProxy health check hangs with remote podman #54

@alicefr

Description

@alicefr

bink cluster start hangs at Step 8 ("Creating HAProxy load balancer") for 2 minutes and then times out when bink is run from a different host/container than where podman is running (remote podman via CONTAINER_HOST).

WaitForHealthy in internal/haproxy/haproxy.go uses a Go http.Client to poll https://localhost:/healthz from the bink process. This assumes localhost is the machine where containers are running.

With remote podman, the HAProxy port is published on the podman host, not on the machine running bink. The health check connects to localhost which has nothing listening on that port, loops with "connection refused" every 2 seconds for 2 minutes, then returns a timeout error.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions