Configurations for RPO

Updated

RPO (recovery point objective) backups are based on intelligent automatic scheduling. To get started, you define the following options in a plan:

  • Backup frequency: How often incremental backup jobs are run. You can specify the frequency in number of hours or days. If you specify the frequency in days, then specify the start time for the jobs. The time is according to the time zone of the client.

  • Add full backup: Enable full backup jobs. This option is applicable only for server backup plans.

  • Full backup frequency: If you enable a full backup job, specify the frequency at which the system should start the job. You can select the frequency to run the job daily, weekly or monthly. For a weekly frequency, specify the days of the week. For a monthly frequency, specify the week and the day of the month.

  • Start time: If you enable a full backup job, specify the time of day to start the job. The time is according to the time zone of the client.

  • Backup window: The days and times when an incremental backup job can run.

  • Full backup window: The days and times when a full backup job can run.

After the above options are defined, automatic schedules and dynamic priorities are used to achieve your RPO.

Disk Caching of Database Log Backups

If your current log backup schedule is hourly or several times a day, you can use disk caching of log backups to decrease the load on the CommServe computer. You can run log backups at a greater frequency, in a scalable manner, without adding to the workload of the CommServe computer. Log backups continue to work even if the CommServe computer is down for maintenance or if connectivity to the CommServe computer is disrupted. With disk caching enabled (Use disk cache for log backups option) the log recovery point objective (RPO) can be set to as low as 5 minutes.

Any interactively submitted job that results in a log backup uses disk caching when the feature is enabled. The job history of these jobs displays an application size of 0 to prevent the size from being counted twice if capacity licensing is enabled.

Backup windows, which define when backup operations are to be run, are not applicable to database log backups using disk caching. This is because one goal of disk caching is to run disk caching backups as defined in the schedule, regardless of a CommServe computer's maintenance or downtime.

All clients, except the clients that have the SQL Server Agent, must be running Indexing Version 2 to enable disk caching of log backups. For more information about Indexing Version 2, see Indexing Version 2.

Disk caching of database log backups is available for the following agents: Informix, Microsoft SQL Server on Windows, Oracle, Oracle RAC, and SAP HANA

Schedules

  • An incremental backup schedule is configured to automatically convert to a full backup job based on Commvault backup conversion rules.

  • A synthetic full backup schedule is configured to be an automatic schedule.

  • A full backup schedule is configured manually. This schedule is applicable only for server backup plan.

  • By default, the plan includes a continuous incremental schedule set to the RPO value for the plan.

Automatic Backup Level

The jobs from the incremental schedule are automatically converted to a full backup job based on the following backup conversion rules:

  • For SQL agents:

    • If the most recent full backup job is older than 30 days and the operation window allows a full backup job, the job is converted to a full backup job.

    • If the most recent full backup job is within 30 days and the most recent differential backup job is older than 7 days, the job is converted to a differential backup job.

  • For other agents: If the most recent full backup job is older than 7 days and the operation window allows a full backup job, the job is converted to a full backup job.

Dynamic Priority

Using machine learning, the dynamic priority intelligently prioritizes backup jobs using the following criteria:

  • Strike Count: The strike count for a subclient is defined as the number of failures that happened since the last successful backup job of the same backup level. The strike count is computed dynamically.

  • Estimated Completion Time: The estimated completion time is calculated based on previous backup job patterns of the same backup level on the subclient. The estimated completion time is calculated using machine learning.

  • Priority Calculation: When backup jobs have equal priorities at the operation, client, and agent levels, the subclient with the greater strike count is prioritized. If backup jobs also have equal strike counts, the subclient with the higher estimated completion time is prioritized.

Important Things to Know About a Recovery Point

Here are a few things to know about a recovery point:

  • The recovery point for a subclient is the start time of the last successful backup job. The recovery point for a client is the earliest recovery point of all subclients. You can recover all the data on the client computer that was created before the recovery point.

  • Jobs that start later than the recovery point and fail are not included in the calculation of the recovery point.

  • The creation of a new recovery point is neither dependent on the previous recovery point nor related to the RPO.