Shutdown and Startup

Shutdown and Startup Procedures¶

This guide provides the safe and validated procedures for shutting down and starting up the entire homelab. Following these steps is critical to prevent data corruption, especially for services relying on the OCFS2 cluster filesystem.

The core principle is to treat the NAS as the first thing on and the last thing off. All cluster nodes must have storage gracefully detached before the NAS is powered down.

Shutdown Procedure¶

This procedure powers down all services and nodes in a controlled sequence using dedicated Taskfile commands.

Step 1: Stop All Docker Stacks

This command runs an Ansible playbook to remove all running stacks, which also handles the unmounting of CIFS volumes.

task ansible:teardown -- -K

Note: This step can take 10-15 minutes as it waits for network volumes to unmount cleanly.

Step 2: Unmount OCFS2 Filesystems

This task runs a playbook to unmount all configured OCFS2 filesystems on all nodes.

task ansible:ocfs2:umount -- -K

Step 3: Disconnect iSCSI Storage

This task logs out from the iSCSI targets on all nodes.

task ansible:iscsi:logout -- -K

Step 4: Shut Down Cluster Nodes

Shut down the workers first, then the managers, by passing the inventory group name to the task.

1. Shut down worker nodes:

   task ansible:nodes:shutdown -- workers -K
   

2. Shut down manager nodes:

   task ansible:nodes:shutdown -- managers -K
   

Step 5: Shut Down the NAS

After confirming all cluster nodes are powered off, the NAS is no longer serving files or block storage and can be safely shut down via its administrative interface.

Startup Procedure¶

This procedure brings the homelab back online, ensuring storage is available before services attempt to start.

Step 1: Power On Hardware

1. Power on the NAS and wait for it to be fully booted and available on the network.
2. Power on all cluster nodes (managers and workers).

Step 2: Verify Node Connectivity

Use the ansible:ping task to confirm that all nodes are online and reachable by Ansible.

task ansible:ping

Step 3: Connect iSCSI Storage

This task runs a playbook that discovers and logs into the iSCSI targets defined in your Ansible inventory.

task ansible:iscsi:login

Step 4: Mount OCFS2 Filesystems

This task mounts all configured OCFS2 filesystems.

task ansible:ocfs2:mount

Configuration Note: For this step to work reliably, you must define the UUIDs of your OCFS2 filesystems in your Ansible inventory (e.g., inventory/group_vars/all.yml). This ensures the correct devices are mounted every time.

# inventory/group_vars/all.yml
OCFS2_MEDIA_UUID: "your-media-partition-uuid"
OCFS2_APP_DATA_UUID: "your-app-data-partition-uuid"

You can find the UUID of a formatted partition by using the lsblk -f or blkid command on one of the nodes after connecting to the iSCSI target.

Step 5: Deploy All Services

Use the ansible:deploy task to bring all your application stacks online.

task ansible:deploy

Ansible will deploy all stacks defined in your stacks.yml file, and services will start in a controlled manner.