fix: Bento healthchecks + patch Keycloak version

README.md

@@ -22,8 +22,8 @@ that make up the Bento platform.
 
 
 ## Requirements
-- Docker >= 24.0.4
-- Docker Compose >= 2.20.0 (plugin form: you should have the `docker compose` command available, without a dash)
+- Docker >= 25.0
+- Docker Compose >= 2.25.0 (plugin form: you should have the `docker compose` command available, without a dash)
 - Python >= 3.9 (for `bentoctl`); the services require Python 3.10 but this is included in their Docker images. 
 
 

etc/bento.env

@@ -17,8 +17,11 @@ BENTO_AUTHZ_SERVICE_URL=${BENTOV2_PUBLIC_URL}/api/authorization/
 
 BENTO_CORS_ORIGINS="${BENTOV2_PUBLIC_URL};${BENTOV2_PORTAL_PUBLIC_URL}"
 
-BENTOV2_HEALTHCHECK_TIMEOUT=5s
-BENTOV2_HEALTHCHECK_INTERVAL=5m
+BENTO_HEALTHCHECK_TIMEOUT=5s
+BENTO_HEALTHCHECK_INTERVAL=5m
+# Faster interval for services while they're starting:
+BENTO_HEALTHCHECK_START_PERIOD=75s
+BENTO_HEALTHCHECK_START_INTERVAL=5s
 
 BENTO_STD_SERVICE_INTERNAL_PORT=5000
 
@@ -44,7 +47,7 @@ BENTOV2_GATEWAY_CPUS=2
 
 # - Keycloak IdP - 'auth'
 BENTOV2_AUTH_IMAGE=ghcr.io/bento-platform/bento_keycloak_dist
-BENTOV2_AUTH_VERSION=2024.06.03
+BENTOV2_AUTH_VERSION=2024.09.26
 BENTOV2_AUTH_CONTAINER_NAME=${BENTOV2_PREFIX}-auth
 BENTO_AUTH_NETWORK=${BENTOV2_PREFIX}-auth-net
 BENTOV2_AUTH_SERVICE_HOST=0.0.0.0

lib/aggregation/docker-compose.aggregation.yaml

@@ -20,8 +20,10 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_AGGREGATION_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   aggregation-net:

lib/auth/docker-compose.auth.yaml

@@ -35,8 +35,12 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "curl", "https://localhost:8443", "-k" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      # interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      # For Docker <25 and Compose <2.24.x, start_interval doesn't work - use a shorter interval for now
+      interval: 45s
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     profiles:
       - auth
 
@@ -53,8 +57,12 @@ services:
       - ${BENTOV2_AUTH_VOL_DIR}:/var/lib/postgresql/data
     healthcheck:
       test: [ "CMD", "pg_isready", "-q", "-d", "${BENTO_AUTH_DB}", "-U", "${BENTO_AUTH_DB_USER}" ]
-      timeout: 5s
-      interval: 15s  # Faster interval for auth-db, which auth requires to be healthy
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      # interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      # For Docker <25 and Compose <2.24.x, start_interval doesn't work - use a shorter interval for now
+      interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     profiles:
       - auth
 

lib/authz/docker-compose.authz.yaml

@@ -19,6 +19,12 @@ services:
     networks:
       - authz-net
       - authz-db-net
+    healthcheck:
+      test: [ "CMD", "curl", "http://localhost:${BENTO_AUTHZ_INTERNAL_PORT}/service-info" ]
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
   authz-db:
     image: postgres:${BENTO_AUTHZ_DB_VERSION}
@@ -34,8 +40,12 @@ services:
       - ${BENTO_AUTHZ_DB_VOL_DIR}:/var/lib/postgresql/data
     healthcheck:
       test: [ "CMD", "pg_isready", "-q", "-d", "${BENTO_AUTHZ_DB}", "-U", "${BENTO_AUTHZ_DB_USER}" ]
-      timeout: 5s
-      interval: 15s  # Faster interval for authz-db, which authz requires to be healthy
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      # interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      # For Docker <25 and Compose <2.24.x, start_interval doesn't work - use a shorter interval for now
+      interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   authz-net:

lib/beacon/docker-compose.beacon.yaml

@@ -41,8 +41,10 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTO_BEACON_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     profiles:
       - beacon
 

lib/drop-box/docker-compose.drop-box.yaml

@@ -21,8 +21,10 @@ services:
     cpu_shares: 1024
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_DROP_BOX_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   drop-box-net:

lib/drs/docker-compose.drs.yaml

@@ -27,8 +27,10 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_DRS_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   drs-net:

lib/event-relay/docker-compose.event-relay.yaml

@@ -54,8 +54,10 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_EVENT_RELAY_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   event-relay-net:

lib/gateway/docker-compose.gateway.yaml

@@ -117,8 +117,10 @@ services:
         "--resolve", "${BENTOV2_PORTAL_DOMAIN}:443:127.0.0.1",
         "https://${BENTOV2_PORTAL_DOMAIN}"
       ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 
 # Lots of networks! Here, we have more or less one network per service, plus a few ones for databases

lib/gohan/docker-compose.gohan.yaml

@@ -46,8 +46,10 @@ services:
       - ${BENTOV2_GOHAN_API_INTERNAL_PORT}
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_GOHAN_API_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     depends_on:
       - gohan-elasticsearch
     profiles:
@@ -73,8 +75,10 @@ services:
       - ${BENTO_GOHAN_ES_JVM_OPTIONS_DIR}:/usr/share/elasticsearch/config/jvm.options.d
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_GOHAN_ES_INTERNAL_PORT_1}" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     profiles:
       - gohan
 

lib/katsu/docker-compose.katsu.yaml

@@ -55,8 +55,10 @@ services:
     cpu_shares: 1024
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_KATSU_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 
   katsu-db:
@@ -82,8 +84,12 @@ services:
     cpu_shares: 1024
     healthcheck:
       test: [ "CMD", "pg_isready", "-q", "-d", "${BENTOV2_KATSU_DB}", "-U", "${BENTOV2_KATSU_DB_USER}" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: 15s  # Faster interval for services which others may require to be healthy
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      # interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      # For Docker <25 and Compose <2.24.x, start_interval doesn't work - use a shorter interval for now
+      interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   drs-net:

lib/logs/docker-compose.logs.yaml

@@ -59,7 +59,7 @@ services:
     expose:
       - 3000
     healthcheck:
-      test: [ "CMD", "curl", "-k", "https://localhost:3000"]
+      test: [ "CMD", "curl", "http://localhost:3000"]
       timeout: 5s
       interval: 15s
     profiles:

lib/notification/docker-compose.notification.yaml

@@ -24,8 +24,10 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_NOTIFICATION_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   notification-net:

lib/public/docker-compose.public.yaml

@@ -32,8 +32,10 @@ services:
       - ${PWD}/lib/public/branding.png:/bento-public/dist/public/assets/branding.png:ro
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTO_PUBLIC_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     profiles:
       - public
 

lib/redis/docker-compose.redis.yaml

@@ -13,8 +13,10 @@ services:
     cpu_shares: 512
     healthcheck:
       test: [ "CMD", "redis-cli", "ping" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: 15s  # Faster interval for services which others may require to be healthy
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   redis-net:

lib/reference/docker-compose.reference.yaml

@@ -19,9 +19,11 @@ services:
     volumes:
       - ${BENTO_REFERENCE_TMP_VOL_DIR}:/reference/tmp
     healthcheck:
-      test: [ "CMD", "wget", "-q0-", "http://localhost:${BENTO_REFERENCE_INTERNAL_PORT}/service-info", "&&", "echo" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      test: [ "CMD", "curl", "http://localhost:${BENTO_REFERENCE_INTERNAL_PORT}/service-info" ]
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
   reference-db:
     image: ${BENTO_REFERENCE_DB_IMAGE}:${BENTO_REFERENCE_DB_VERSION}
@@ -40,6 +42,14 @@ services:
       - POSTGRES_DB=${BENTO_REFERENCE_DB_NAME}
     volumes:
       - ${BENTO_REFERENCE_DB_VOL_DIR}:/var/lib/postgresql/data
+    healthcheck:
+      test: [ "CMD", "pg_isready", "-q", "-d", "${BENTO_REFERENCE_DB_NAME}", "-U", "${BENTO_REFERENCE_DB_USER}" ]
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      # interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      # For Docker <25 and Compose <2.24.x, start_interval doesn't work - use a shorter interval for now
+      interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   reference-net:

lib/service-registry/docker-compose.service-registry.yaml

@@ -41,8 +41,10 @@ services:
     cpu_shares: 256
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_SERVICE_REGISTRY_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   service-registry-net:

lib/web/docker-compose.web.yaml

@@ -21,9 +21,11 @@ services:
     cpus: ${BENTOV2_WEB_CPUS}
     cpu_shares: 512
     healthcheck:
-      test: [ "CMD", "wget", "-q0-", "http://localhost:${BENTOV2_WEB_INTERNAL_PORT}/service-info", "&&", "echo" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      test: [ "CMD", "curl", "http://localhost:${BENTOV2_WEB_INTERNAL_PORT}/service-info" ]
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
 
 networks:
   web-net:

lib/wes/docker-compose.wes.yaml

@@ -49,8 +49,10 @@ services:
     cpu_shares: 1024
     healthcheck:
       test: [ "CMD", "curl", "http://localhost:${BENTOV2_WES_INTERNAL_PORT}/service-info" ]
-      timeout: ${BENTOV2_HEALTHCHECK_TIMEOUT}
-      interval: ${BENTOV2_HEALTHCHECK_INTERVAL}
+      timeout: ${BENTO_HEALTHCHECK_TIMEOUT}
+      interval: ${BENTO_HEALTHCHECK_INTERVAL}
+      start_period: ${BENTO_HEALTHCHECK_START_PERIOD}
+      start_interval: ${BENTO_HEALTHCHECK_START_INTERVAL}
     volumes:
       - ${BENTOV2_WES_VOL_DIR}:/wes/data
       - ${BENTOV2_WES_VOL_TMP_DIR}:/wes/tmp