When/why would you schedule Data Replication?
In the ideal world data replication is only in real time, is turned on once, and never has any problems. Most real-world replication has yet to reach the ideal world, and as a result there are many reason why it makes sense to schedule jobs in a real-time data replication environment.
Think of the following scenarios in the context of data replication:
- One time load
- A batch load runs on your source system and you only want to capture the result.
- There are some unlogged (dangerous!) direct path loads that took place and the new data must be captured.
- You want to move data between systems that have no log-based capture mechanism, and data does not have to be in real-time.
- For the initial load the source system cannot be overloaded so run the data extract during the weekend rather than on a week day.
- Compare: schedule data compare jobs when the system is relatively idle so that over time you continue to feel comfortable that the systems are in sync.
- Choose to run CDC in steps rather than in real-time.
Every operating system has its own scheduler so why should a database replication tool have its own scheduler? HVR has its own scheduler that
- Is identical on every platform,
- Ensures appropriate restarts if processes fail,
- Centralizes logging.
HVR does use the operating system scheduler, but only to run maintenance tasks that may include verifying that its own scheduler is still running.
To learn more about our software, contact us.