Utility: schema_backup

schema_backup reads the CSV2 database and creates backups of both table structures and non-ephemeral table data to the standard and user data backup directories (see configuration below). It is normally invoked during the upgrade process by the db_upgrade utility but should also be run regularly as a stand-alone utility to backup user data.

Synopsis: schema_backup

What gets backed up and where?

Tables are divided into four categories (see configuration below and Tables) for security, and upgrade purposes and are backed up as follows:

  • All table structures and all global configuration data is saved in a host specific subdirectory of the CSV2 backup directory (configurable). By default, the CSV2 backup directory is …/cloudscheduler/schema/schema_backup.
  • All user data (and table structures) is saved in the user data backup directory (configurable).
  • Ephemeral data is ignored and not saved.

Configuration

schema_backup, and other utilities, are configured by the YAML file * …/cloudscheduler/etc/schema_backup.conf. This file identifies the backup directories, database service provider and consumers, and the categorization of the tables within the CSV2 database. The following parameters are recognozed from this file:

  • csv2_backup_dir - specifies the full path of a local filesystem directory to be used as the CSV2 backup directory. It is recommended that this directory is also within a git/github repository. If the parameter is omitted, set to null, or set to “default”, the default value of …/cloudscheduler/schema/schema_backup is assumed.
  • user_data_backup_dir - specifies the full path of a local filesystem directory to be used as the user data backup directory. This directory may also be within a git/github repository, but it should be noted that this data contains private/senesitive information which should be kept secret. If the parameter is omitted, user data will not be backed up.
  • providers - specifies a list of service names of services on the local host providing relational database services for CSV2. Normally, the list consists of one name, “mariadb” or possibly “mysql”, depending on your host configuration.
  • consumers - specifies a list of service names of services that depend on the CSV2 database. Normally, this is a list of all CSV2 service components plus the web server.
  • global - specifies a list of tables names within the CSV2 database that comprise global configuration data. These tables do not contain any sensitive data.
  • local - specifies a list of tables names within the CSV2 database that comprise user, group, and cloud configuration data. These tables will contain sensitive data.
  • ephemeral - specifies a list of tables names within the CSV2 database that comprise the state of the distributed cloud system. This data is in a continual state of flux and is therefore never saved. However, the structure of the tables is important and is saved in the CSV2 backup directory.
  • ignore - specifies a list of tables names within the CSV2 database used by other supporting services and which can be ignored by schema_backup and other CSV2 utilities.

Any other parameter within the configuration file will be ignored.