fix(health): stop raw-probing ports of containers with own healthchecks
Some checks failed
Demo images / Build & push demo images (push) Failing after 1m34s

The health monitor opened a bare TCP connection to every published host
port each cycle. Against TLS listeners that connect+close logs noise on
the app side — LND printed 'http: TLS handshake error … EOF' endlessly.
Containers that define a podman healthcheck (LND uses lncli getinfo)
now rely on it alone; bare containers keep the port probe.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
archipelago 2026-07-16 10:06:03 -04:00
parent 47dea8cd55
commit 5cb53ade66

View File

@ -475,7 +475,12 @@ async fn check_containers() -> Vec<ContainerHealth> {
let podman_health = parse_podman_health(c, &state);
let host_ports = host_tcp_ports_from_container(c);
let host_port_ready = if host_ports.is_empty() {
// Only raw-probe published ports for containers WITHOUT their own
// podman healthcheck. The healthcheck is the better signal, and the
// bare TCP connect+close is noisy against TLS listeners — LND logged
// "http: TLS handshake error … EOF" on every monitor cycle because
// this probe hit its REST/gRPC ports and hung up mid-handshake.
let host_port_ready = if host_ports.is_empty() || podman_health.is_some() {
None
} else {
Some(host_ports_ready(&host_ports).await)