But restoring a logical backup is pretty straightforward anyway. WALG_DISABLE_S3_SSE: "true" # I disable this because it's not supported by ScalewayĪs far as I know, the operator doesn’t support restoring from logical backups directly in an automated way, yet. As I mentioned earlier, this config map includes the information required for both archiving of WAL to S3, and clone/point-in-time recovery from S3 using the WAL archives. Then you need to create a config map which will be used by the database pods. To create a cluster, first create a namespace, e.g. In the meantime I have build a custom image for the logical backups using the relevant Dockerfile in the repo, and adding the -region parameter to the aws CLI command in dump.sh. Please note that there currently is a missing setting for the region, so I have submitted a PR to add the setting logical_backup_s3_region since this is required with some s3 compatible services like Scaleway, otherwise the upload of logical backups fails. The logical backups settings should be self explanatory. I am not 100% sure whether the configAwsOrGcp section is actually needed when using s3 compatible storage. Setting that to * makes it possible to create clusters in different namespaces, which is likely what you want. watched_namespace is set to an empty string by default, which means that the operator will only watch its namespace for the custom resources required to create the clusters. Pod_environment_configmap specifies the name of a config map where we define some settings for the cluster to enable WAL archivation to object storage, and cloning/point in time recovery.
0 Comments
Leave a Reply. |