On-Premise Deployment using Filesystem
Warning
Note
- If the user chooses
backup_config
asfile_system
inconfig.toml
backup is already configured during the deployment, and in that case the below steps are not required. Ifbackup_config
is left blank, then the configuration needs to be configured manually.
Overview
A shared file system is always required to create OpenSearch snapshots. To register the snapshot repository using OpenSearch, it is necessary to mount the same shared filesystem to the exact location on all master and data nodes. Register the location (or one of its parent directories) in the path.repo
setting on all master and data nodes.
Setting up the backup configuration
Configuration in OpenSearch Node
- Mount the shared file system to the base mount path which is mentioned in
backup_mount
on all OpenSearch and Frontend servers.
Note
/mnt/automate_backups
is the default value for thebackup_mount
, which is also used in this document page as reference backup path.
Apply the following steps on all of the OpenSearch server node
Create an OpenSearch sub-directory and set permissions (only if the network mount is correctly mounted).
sudo mkdir /mnt/automate_backups/opensearch sudo chown hab:hab /mnt/automate_backups/opensearch/
Configuration for OpenSearch Node from Bastion Host
Configure the OpenSearch path.repo
setting by following the steps given below:
Create a .toml (say os_config.toml) file in the Bastion host and copy the following template with the path to the repo.
[path] # Replace /mnt/automate_backups with the backup_mount config found on the Bastion host in /hab/a2_deploy_workspace/a2ha.rb repo = "/mnt/automate_backups/opensearch"
Following command will add the configuration to the OpenSearch node.
chef-automate config patch --opensearch <PATH TO OS_CONFIG.TOML>
Healthcheck commands
Following command can be run in the OpenSearch node
chef-automate status
Following command can be run in the OpenSearch node
hab svc status (check whether OpenSearch service is up or not) curl -k -X GET "<https://localhost:9200/_cat/indices/*?v=true&s=index&pretty>" -u admin:admin (Another way to check is to check whether all the indices are green or not) # Watch for a message about OpenSearch going from RED to GREEN journalctl -u hab-sup -f | grep 'automate-ha-opensearch'
Configuration for Automate Node from Provision Host
Configure Automate to handle External OpenSearch Backups.
Create an
automate.toml
file on the provisioning server using the following command:touch automate.toml
Add the following configuration to
automate.toml
on the provisioning host:[global.v1.external.opensearch.backup] enable = true location = "fs" [global.v1.external.opensearch.backup.fs] # The `path.repo` setting you've configured on your OpenSearch nodes must be a parent directory of the setting you configure here: path = "/mnt/automate_backups/opensearch" [global.v1.backups.filesystem] path = "/mnt/automate_backups/backups"
Patch the
automate.toml
config to trigger the deployment from the provision host.chef-automate config patch --fe automate.toml
Backup and Restore commands
Backup
To create the backup, by running the backup command from bastion. The backup command is as shown below:
chef-automate backup create
Restoring the Backed-up Data From the file system
To restore backed-up data of the Chef Automate High Availability (HA) using External File System (EFS), follow the steps given below:
Check the status of Automate HA Cluster from the bastion nodes by executing the
chef-automate status
command.Execute the restore command from bastion
chef-automate backup restore <BACKUP-ID> --yes -b /mnt/automate_backups/backups --airgap-bundle </path/to/bundle>
.
Note
- If you are restoring the backup from an older version, then you need to provide the
--airgap-bundle </path/to/current/bundle>
.
Troubleshooting
While running the restore command, If it prompts any error follow the steps given below.
- Check the chef-automate status in Automate node by running
chef-automate status
. - Also check the hab svc status in automate node by running
hab svc status
. - If the deployment services is not healthy then reload it using
hab svc load chef/deployment-service
. - Now, check the status of Automate node and then try running the restore command from bastion.