(iceberg-go) branch main updated: fix(ci): give Trino healthcheck startup grace time (#1429)

[at...@apache.org]

This is an automated email from the ASF dual-hosted git repository.

laskoviymishka pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/iceberg-go.git


The following commit(s) were added to refs/heads/main by this push:
     new d99657d9 fix(ci): give Trino healthcheck startup grace time (#1429)
d99657d9 is described below

commit d99657d9c1b9033fab25b08c5334a27f988c95d2
Author: Minh Vu <vu...@gmail.com>
AuthorDate: Thu Jul 9 22:49:03 2026 +0200

    fix(ci): give Trino healthcheck startup grace time (#1429)
    
    ## Summary
    - give the Trino service a startup grace period in
    `internal/recipe/docker-compose.yml`
    - keep `docker compose up -d --wait` from failing while Trino is still
    booting in CI
    
    ## Why
    The Spark 3.5 integration setup has been flaking before any Go tests run
    because Trino can take around 30 seconds to become ready in GitHub
    Actions. Without a startup grace period, the container's healthcheck can
    flip to `unhealthy` just before startup completes, which makes `make
    integration-setup` fail even though Trino comes up moments later.
    
    This keeps the existing readiness flow in place and just gives the
    healthcheck enough warm-up time to match observed CI startup behavior.
    
    Closes #1420.
    
    ## Testing
    - `docker compose -f internal/recipe/docker-compose.yml config`
---
 internal/recipe/docker-compose.yml | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/internal/recipe/docker-compose.yml b/internal/recipe/docker-compose.yml
index 664eab2e..313e967f 100644
--- a/internal/recipe/docker-compose.yml
+++ b/internal/recipe/docker-compose.yml
@@ -117,6 +117,11 @@ services:
       - CATALOG_MANAGEMENT=dynamic
     ports:
       - 8088:8080
+    healthcheck:
+      # Trino can take ~30s to finish booting in GitHub Actions; without a
+      # startup grace period, docker compose up --wait can mark it unhealthy
+      # just before the server reaches READY.
+      start_period: 45s
 
   fake-gcs-server:
     image: fsouza/fake-gcs-server