I was using healthcheck in postgres container which triggered this ominous error.
I made the following changed and the problem went away (for obvious reason, I guess) — pg_isready needs to know which user is being problem.
Following is the diff.
index 4c91dcc..89ad702 100644
--- a/compose.yaml
+++ b/compose.yaml
@@ -6,8 +6,9 @@ services:
     shm_size: 512mb
     environment:
       POSTGRES_PASSWORD: test_postgres_123
+      POSTGRES_USER: test_postgres_user
     healthcheck:
-      test: ['CMD-SHELL', 'pg_isready' ]
+      test: "pg_isready -U $$POSTGRES_USER"
       interval: 1s
       timeout: 5s
       retries: 10