Does not schedule healthcheck for images with HEALTHCHECK

[stefanb2]

Is this a BUG REPORT or FEATURE REQUEST? (leave only one on its own line)

/kind bug

Description

podman only schedules health check when command line option --healthcheck-command is given but not when the docker image has a HEALTHCHECK configuration. This is a regression compared to Docker CLI behaviour.

Steps to reproduce the issue:

1. build (docker) image with HEALTHCHECK in Dockerfile
2. create container from image using podman

Describe the results you received:

# image created by "docker build"
$ podman run --rm -d --name test docker.COMPANY.COM:5004/PATH/mongo:3.4.7
1ff89ce50e988d02bcf0393cce75b5a03e8136470fb04bd8efe5140e1cdab56f

$ podman container inspect --format='{{json .Config.Healthcheck}}' test
{"Test":["CMD","docker-healthcheck"]}

# the following never changes...
$ podman container inspect --format='{{json .State.Healthcheck}}' test
{"Status":"starting","FailingStreak":0,"Log":null}


# local image created by "podman build" from Dockerfile
$ fgrep -A1 HEALTH Dockerfile
HEALTHCHECK --interval=1m --timeout=5s --retries=3 \
    CMD curl -fsS http://localhost:8080/docs -o /dev/null || exit 1

$ podman image inspect localhost/PROJECT/mock | fgrep -i HEALTH
 "created_by": "/bin/sh -c #(nop) HEALTHCHECK --interval=1m --timeout=5s --retries=3     CMD curl -fsS http://localhost:8080/docs -o /dev/null || exit 1",

$ podman run --rm -d --name test localhost/PROJECT/mock
a416421d7718c00c6420a475ec33dfa6545f3f4de8609fecc660f0b3c2f1edda

$ podman container inspect --format='{{json .Config.Healthcheck}}' test
null
$ podman container inspect --format='{{json .State.Healthcheck}}' test
{"Status":"","FailingStreak":0,"Log":null}

(IMHO the podman build example indicates another bug, because the HEALTHCHECK issue gets lost completely...)

Describe the results you expected:

Health check service & timer should be created.

Additional information you deem important (e.g. issue happens only occasionally):

May be related to #3507?

Output of podman version:

Version:            1.4.5-dev
RemoteAPI Version:  1
Go Version:         go1.12.6
OS/Arch:            linux/amd64

Output of podman info --debug:

debug:
  compiler: gc
  git commit: ""
  go version: go1.12.6
  podman version: 1.4.5-dev
host:
  BuildahVersion: 1.9.0
  Conmon:
    package: podman-1.4.4-1.fc30.x86_64
    path: /usr/libexec/podman/conmon
    version: 'conmon version 0.2.0, commit: 41010e63c287618b1dc34ee11d10d268e2feeefe'
  Distribution:
    distribution: fedora
    version: "30"
  MemFree: 17038602240
  MemTotal: 33400737792
  OCIRuntime:
    package: runc-1.0.0-93.dev.gitb9b6cc6.fc30.x86_64
    path: /usr/bin/runc
    version: |-
      runc version 1.0.0-rc8+dev
      commit: e3b4c1108f7d1bf0d09ab612ea09927d9b59b4e3
      spec: 1.0.1-dev
  SwapFree: 32002535424
  SwapTotal: 32002535424
  arch: amd64
  cpus: 12
  hostname: beckst-lnx
  kernel: 5.1.16-300.fc30.x86_64
  os: linux
  rootless: true
  uptime: 39h 7m 49.8s (Approximately 1.62 days)
registries:
  blocked: null
  insecure: null
  search:
  - docker.io
  - registry.fedoraproject.org
  - quay.io
  - registry.access.redhat.com
  - registry.centos.org
store:
  ConfigFile: /home/stefanb/.config/containers/storage.conf
  ContainerStore:
    number: 0
  GraphDriverName: overlay
  GraphOptions:
  - overlay.mount_program=/usr/bin/fuse-overlayfs
  GraphRoot: /home/stefanb/.local/share/containers/storage
  GraphStatus:
    Backing Filesystem: extfs
    Native Overlay Diff: "false"
    Supports d_type: "true"
    Using metacopy: "false"
  ImageStore:
    number: 6
  RunRoot: /tmp/1000
  VolumePath: /home/stefanb/.local/share/containers/storage/volumes

[stefanb2]

I think I now understand the root cause for this: if a docker image doesn't provide a HEALTHCHECK option (interval, retries, timeout, start_period) then podman does not apply the standard default values like Docker does.

I.e. with an image created from the following Dockerfile snippet

HEALTHCHECK --interval=30s --timeout=30s --retries=3 CMD xyz

podman would schedule a healthcheck.

[stefanb2]

Furthermore: podman must disable the healthcheck if image has been built with HEALTHCHECK NONE. It should not create a healthcheck configuration with command list ["NONE"].