Data Recovery Protection

What is data recovery?

Data recovery involves restoring data that has been lost or corrupted due to a security breach, technical failure, human error or disaster.

What is a data storage discipline?

The data storage discipline focuses on the management of persistent data stores throughout the organization.  It covers such topics as database selection guidelines, backup & recovery procedures, data retention/archiving policies, and disaster recovery planning.

Database administrators in this discipline require extensive experience working in a data warehousing/business intelligence environment.

They are also usually heavily involved with integration testing and with test data acquisition.

Which environments should have data recovery protection?



Development

• Personal schema's are created for individual developers who use their personal schema’s to develop and unit test individual code before moving to system testing in the development environment. These schema's need to be backed-up on a regular basis to avoid time loss recreating test data in the event of human error;
  

• A system test environment is sometimes  used to system test several code components before they are moved to the integration test environment. These environments require backup after they are created so that they can be restored to a "vanilla" state prior to each system test.

• An integration test environment is used by the integration test team to complete integration and load performance testing. These environments will require:

• Test data backup, to eliminate time loss to recreate test data;
• Frequent checkpoint backups so that data ca be restored part way thru a test thus eliminating th need to restart the entire test (which might be very time consuming);
  

• Quality assurance (QA) environment is used by an independent QA team to ensure that the application meets specified requirements. This environment will require:

• Test data backup, to eliminate time loss to recreate test data;
• Frequent checkpoint backups so that data can be restored part way thru a test thus eliminating th need to restart the entire test (which might be very time consuming); and
• Rapid data recovery procedures, e.g. back-ups to disk so that restores do not take "forever" loading archive tapes;

• User acceptance test environment is used by end users so they can test that the application meets expectations. These environments require backup after they are created so that they can be restored to a "vanilla" state prior to each user acceptance test. Frequent checkpoint/back-ups are less important since most real bugs should have been caught during integration and quality assurance testing.

• Release test  environment is used to test the operational readiness of new application software before it moves into production. This environment requires a backup before each release and may require checkpoint  backups after any initial data conversions have  occurred to  avoid lengthy downtime due to data issues;

• Production support (DEV/Test) environment is used by the production support team to fix and test any production issues before moving them into production. This environment should have normal daily backups but it is expected to be a mirror image of production so data recovery, if needed, could come from the production environment;

• Production environment contains all production data. This environment needs the most data recovery protection which might  include:

• Daily incremental backups;
• Weekly "Cold" backups;
• Off site data backup storage;
  
• Data mirroring e.g. a copy of the production environed is created as changes occur using tools such as Oracle Data Guard; and
• Data center back-up facilities, and associated procedures to ensure data backups are current, to provide disaster protection.

• Data architect, who has overall responsibility for data warehousing solution architecture; and
• Database administrators, who are responsible for defining and ensuring backup & recovery procedures and processes.