ci: drop the early-exit branch from the Apollo wait loop

The healthcheck in the compose file keeps the config service from dying in
the first place, so the loop no longer needs to detect a container that has
given up.

Author: lingcoder

.github/workflows/ci-main.yml

@@ -247,13 +247,6 @@ jobs:
             echo "Apollo is serving configuration."
             exit 0
           fi
-          # The container gives up on its own after 120 seconds. Once it has,
-          # nothing will ever answer, so stop burning the remaining budget.
-          if docker compose -f ".github/workflows/apollo/docker-compose.yml" logs apollo-quick-start 2>/dev/null | grep -Fq 'Config service failed to start'; then
-            echo "Apollo config service gave up during startup."
-            docker compose -f ".github/workflows/apollo/docker-compose.yml" logs --tail=50
-            exit 1
-          fi
           sleep 5
         done
         echo "Apollo did not start serving configuration within 5 minutes."