diff options
4 files changed, 31 insertions, 10 deletions
diff --git a/ansible/roles/service-check/tasks/main.yml b/ansible/roles/service-check/tasks/main.yml index 9f4cf30af..9cb17ad09 100644 --- a/ansible/roles/service-check/tasks/main.yml +++ b/ansible/roles/service-check/tasks/main.yml @@ -28,8 +28,8 @@ unhealthy_containers: >- {{ container_facts.containers | dict2items | - selectattr("value.Status", "defined") | - selectattr("value.Status", "search", "unhealthy") | + selectattr("value.State.Health", "defined") | + selectattr("value.State.Health.Status", "search", "unhealthy") | map(attribute='key') | list }} ansible.builtin.fail: msg: > diff --git a/releasenotes/notes/fix-container-healthy-checks-4ecdbe67e07819aa.yaml b/releasenotes/notes/fix-container-healthy-checks-4ecdbe67e07819aa.yaml new file mode 100644 index 000000000..2fdba6fcf --- /dev/null +++ b/releasenotes/notes/fix-container-healthy-checks-4ecdbe67e07819aa.yaml @@ -0,0 +1,5 @@ +--- +fixes: + - | + Fixes kolla-ansible check skip unhealthy containers. + `LP#2156900 <https://bugs.launchpad.net/kolla-ansible/+bug/2156900>`__ diff --git a/releasenotes/notes/fix-proxysql-database-internal-tls-b401e564ec39a633.yaml b/releasenotes/notes/fix-proxysql-database-internal-tls-b401e564ec39a633.yaml index fb3fb9db9..238a5e583 100644 --- a/releasenotes/notes/fix-proxysql-database-internal-tls-b401e564ec39a633.yaml +++ b/releasenotes/notes/fix-proxysql-database-internal-tls-b401e564ec39a633.yaml @@ -6,7 +6,7 @@ fixes: certificates (e.g. use of intermediate certificates). Users who are affected by this bug needs to create a file - ``proxysql-ca.pem`` at ``certificates_internal_dir``. + ``proxysql-ca.pem`` at ``kolla_certificates_dir``. This file needs to be a copy of the intermediate certificate (i.e. certificate that signed database internal TLS) diff --git a/tests/check-failure.sh b/tests/check-failure.sh index 19d2f071f..ecd3a10ac 100755 --- a/tests/check-failure.sh +++ b/tests/check-failure.sh @@ -6,13 +6,25 @@ set -o errexit # Enable unbuffered output for Ansible in Jenkins. export PYTHONUNBUFFERED=1 +# neutron_agents_wrappers (see ansible/roles/neutron) spawns per-namespace +# helper containers (dnsmasq/haproxy/keepalived) that are not managed by +# Kolla and can be created/removed by Neutron independently of the deploy, +# e.g. neutron_dhcp_agent_dnsmasq_qdhcp-<uuid>. Ignore them in these checks. +IGNORE_CONTAINERS_REGEX='_(qdhcp|qrouter|ovnmeta)-' + +filter_unmanaged_containers() { + grep -Ev "$IGNORE_CONTAINERS_REGEX" || true +} + check_podman_failures() { failed_containers=$(sudo podman ps -a --format "{{.Names}}" \ --filter status=created \ - --filter status=paused \ --filter status=exited \ - --filter status=unknown) + --filter status=paused \ + --filter status=removing \ + --filter status=unknown \ + | filter_unmanaged_containers) for container in $failed_containers; do sudo podman inspect $container @@ -28,7 +40,8 @@ check_podman_unhealthies() { done unhealthy_containers=$(sudo podman ps -a --format "{{.Names}}" \ - --filter health=unhealthy) + --filter health=unhealthy \ + | filter_unmanaged_containers) for container in $unhealthy_containers; do echo "Discovered unhealthy container: $container" @@ -50,10 +63,12 @@ check_docker_failures() { # are treated as failure. failed_containers=$(sudo docker ps -a --format "{{.Names}}" \ --filter status=created \ - --filter status=restarting \ - --filter status=paused \ --filter status=exited \ - --filter status=dead) + --filter status=dead \ + --filter status=paused \ + --filter status=removing \ + --filter status=restarting \ + | filter_unmanaged_containers) for container in $failed_containers; do sudo docker inspect $container @@ -69,7 +84,8 @@ check_docker_unhealthies() { done unhealthy_containers=$(sudo docker ps -a --format "{{.Names}}" \ - --filter health=unhealthy) + --filter health=unhealthy \ + | filter_unmanaged_containers) for container in $unhealthy_containers; do echo "Discovered unhealthy container: $container" |