Manager

Reset

Sometimes it is necessary to reset the entire manager service.

warning

This is a disruptive action. Data is lost in the process. For example, the database of the ARA service.

Stop the manager service

sudo systemctl stop docker-compose@manager.service

Files on the /share volume are backed up in advance and restored after the manager service is started.

docker run --rm \
  --mount source=manager_share,target=/share \
  --volume $(pwd):/backup \
  busybox \
  tar -czvf /backup/manager-share-$(date +%Y%m%d).tar.gz /share

Delete the manager service. This is a disruptive action.

docker compose --project-directory /opt/manager down -v

Start the manager service

sudo systemctl start docker-compose@manager.service

Restore the files on the /share volume.

docker run --rm \
  --mount source=manager_share,target=/share \
  --volume $(pwd):/backup \
  busybox \
  tar -xzvf /backup/manager-share-$(date +%Y%m%d).tar.gz -C /

Check that manager service is healthy

docker compose --project-directory /opt/manager ps

Depending on what the manager service looks like, this output may vary.

NAME                             IMAGE                                        COMMAND                  SERVICE                CREATED              STATUS                        PORTS
ceph-ansible                     quay.io/osism/ceph-ansible:quincy            "/entrypoint.sh osis…"   ceph-ansible           About a minute ago   Up About a minute (healthy)
kolla-ansible                    quay.io/osism/kolla-ansible:2023.2           "/entrypoint.sh osis…"   kolla-ansible          About a minute ago   Up About a minute (healthy)
manager-api-1                    quay.io/osism/osism:latest                   "osism service api"      api                    About a minute ago   Up About a minute (healthy)   192.168.16.5:8000->8000/tcp
manager-ara-server-1             quay.io/osism/ara-server:latest              "sh -c '/wait && /ru…"   ara-server             About a minute ago   Up About a minute (healthy)   8000/tcp
manager-beat-1                   quay.io/osism/osism:latest                   "osism service beat"     beat                   About a minute ago   Up About a minute (healthy)
manager-conductor-1              quay.io/osism/osism:latest                   "osism worker conduc…"   conductor              About a minute ago   Up About a minute (healthy)
manager-flower-1                 quay.io/osism/osism:latest                   "osism service flower"   flower                 About a minute ago   Up About a minute (healthy)
manager-inventory_reconciler-1   quay.io/osism/inventory-reconciler:latest    "/sbin/tini -- /entr…"   inventory_reconciler   About a minute ago   Up About a minute (healthy)
manager-listener-1               quay.io/osism/osism:latest                   "osism service liste…"   listener               About a minute ago   Up About a minute (healthy)
manager-mariadb-1                index.docker.io/library/mariadb:11.3.2       "docker-entrypoint.s…"   mariadb                About a minute ago   Up About a minute (healthy)   3306/tcp
manager-netbox-1                 quay.io/osism/osism-netbox:latest            "osism worker netbox"    netbox                 About a minute ago   Up About a minute (healthy)
manager-openstack-1              quay.io/osism/osism:latest                   "osism worker openst…"   openstack              About a minute ago   Up About a minute (healthy)
manager-redis-1                  index.docker.io/library/redis:7.2.4-alpine   "docker-entrypoint.s…"   redis                  About a minute ago   Up About a minute (healthy)   6379/tcp
manager-watchdog-1               quay.io/osism/osism:latest                   "osism service watch…"   watchdog               About a minute ago   Up About a minute (healthy)
osism-ansible                    quay.io/osism/osism-ansible:latest           "/entrypoint.sh osis…"   osism-ansible          About a minute ago   Up About a minute (healthy)
osismclient                      quay.io/osism/osism:latest                   "sleep infinity"         osismclient            About a minute ago   Up About a minute

When the manager service is healthy, the inventory and the fact cache must be rebuilt.
```
osism reconciler sync
osism apply facts
```

Manager

Reset​

Reset