Help with healthcheck on redis container

[jata1]

I can't work out why the redis container (named unbound-db in my setup) is saying it is unhealthy but I think it is fine.

I added an additional check to debug in healthcheck.sh and when I run the script using docker exec - it seems to be ok.

here is my updated healthcheck.sh

#! /bin/sh

SOCKET=/usr/local/unbound/cachedb.d/redis.sock
if [[ ! -S "$SOCKET" ]];
then
    echo "$SOCKET is not a file/socket so exit 1"
    exit 1
else
    echo "$SOCKET is a file/socket so do nothing"
fi

Below is from the cli - and it looks to be confirming the container is healthy. Or am I missing something (very likely)

[madnuttah]

It looks good, I'd take a look at the configs you use if you like.

[madnuttah]

Hello @jata1, could you solve your issue? My bet is the start-up order.

[jata1]

Have been in the office and haven't found the cause or solution. I think startup order is correct and the healthcheck test seems to be good.

startup order is unbound-socket --> unbound-db --> unbound

Here is my full config.

version: '3.3'

services:
  unbound:
    container_name: unbound
    image: madnuttah/unbound:latest
    restart: unless-stopped
    depends_on:
      - unbound-db
    ports:
      - 5335:5335/tcp
      - 5335:5335/udp
    networks:
      rpi5_macvlan:
        ipv4_address: "192.168.100.15"
    environment:
      ServerIP: 192.168.100.15
      TZ: ${TZ}
      UID: ${PUID}
      GID: ${PGID}
    volumes:
      - /symlinks/omv-system/appdata/unbound/unbound.conf:/usr/local/unbound/unbound.conf:rw
      - /symlinks/omv-system/appdata/unbound/conf.d/:/usr/local/unbound/conf.d/:rw
      - /symlinks/omv-system/appdata/unbound/iana.d/:/usr/local/unbound/iana.d/:rw
      - /symlinks/omv-system/appdata/unbound/log.d/unbound.log:/usr/local/unbound/log.d/unbound.log:rw
      - /symlinks/omv-system/appdata/unbound/zones.d/:/usr/local/unbound/zones.d/:rw
    volumes_from:
      - unbound-socket:rw
    healthcheck:
      test: /usr/local/unbound/sbin/healthcheck.sh
      interval: 60s
      retries: 5
      start_period: 15s
      timeout: 30s

  unbound-socket:
    image: busybox
    container_name: unbound-socket
    restart: unless-stopped
    tty: true
    command:
      - /bin/sh
      - -c
      - |
        chown -R 999:100 /usr/local/unbound/cachedb.d/
        /bin/sh
    volumes:
      - cachedb.d:/usr/local/unbound/cachedb.d/
      
  unbound-db:
    container_name: unbound-db
    image: redis:alpine
    restart: unless-stopped
    depends_on:
      - unbound-socket
    command: redis-server /usr/local/etc/redis/redis.conf
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /etc/timezone:/etc/timezone:ro
      - /symlinks/omv-system/appdata/unbound/redis/redis.conf:/usr/local/etc/redis/redis.conf
      - /symlinks/omv-system/appdata/unbound/redis/data:/data
      - /symlinks/omv-system/appdata/unbound/redis/healthcheck.sh:/usr/local/sbin/healthcheck.sh:ro
    volumes_from:
      - unbound-socket:rw
    healthcheck:
      test: /usr/local/sbin/healthcheck.sh
      interval: 10s
      retries: 5
      start_period: 5s
      timeout: 30s

volumes:
  cachedb.d:

networks:
  rpi5_macvlan:
    name: rpi5_macvlan
    external: true

[jata1]

Will try that. Thanks.

I think you know already but everything is running/working it’s just health check says unhealthy when it is healthy

[madnuttah]

I've removed that because it wasn't the issue after testing. My apologies, I should better have edited my answer.

Anyway, I've added my compose stack to the examples here, may you find something while comaring?

[madnuttah]

Being honest, I have no idea other than that, since it runs for other users and myself, too. If you found the issue and it's something I could make better, feel free to reopen this issue or create PR with your suggested improvements.

I'll close this for now.

Thank you.