Disaster recovery (DR) and failover drills are a must for any enterprise dealing with critical data delivery and the ongoing protection of that data. For Oracle, these drills drink up valuable time that enterprise teams can be allocating elsewhere, all while dramatically increasing in complexity.
Danny Higgins, global practice lead for Oracle databases at Pure Storage, joined DBTA’s webinar, Modern Non-Disruptive Approaches to Oracle Disaster Recovery Drills, to discuss the ways in which Oracle teams can adopt certain best practices and technologies to reduce the time, complexity, cost, and risk of Oracle disaster recovery drills.
Higgens stressed that for many of today’s organizations, data is their business. In the event of a natural disaster, outage, or hardware failure, data downtime means business is halted, incurring massive losses in enterprise resources and finances.
Disaster recovery drills developed in response to this looming threat of business loss. However, these drills can require more than forty different roles, most likely needed twice in one year; though necessary, disaster recovery drills’ required manpower causes significant pressure on enterprise finances. Massive teams for drills mean thousands of hours and money down the drain.
Today, conventional HA and DR models for Oracle feature Data Guard replication between app servers, which ultimately creates, maintains, manages, and monitors one or more standby databases to enable production Oracle databases to survive disasters and data corruptions.
In the event of a disaster or DR drill, it’s going to be down to the DBAs to go through each and every one of the databases and conduct a Data Guard failover. Despite being the most common method of DR for Oracle, Higgens remarked that it’s quite the “heavy lift” on behalf of the teams conducting it, often disturbing crucial production environments.
Is this a necessary evil? According to Higgens, it doesn’t have to be. With ActiveDR, an asynchronous replication technology, the solution enables the testing of DR processes without experiencing an interruption in production. Meaning, DR drills can be done during the normal work week, reducing the time and complexity usually necessary to ensure DR is successful in the event of a failure.
ActiveDR also delivers the following the advantages:
- Near zero RPO (recovery point objective)
- Fast recover/failover time
- Test failover without losing RPO
- No journals to manage
- Pre-connect hosts
- No additional licensing or software
Higgens continued the discussion by walking webinar viewers through a detailed demo of ActiveDR, highlighting the technology’s ability to conduct non-disruptive DR drills for Oracle, as well as controlled failover for Oracle.
He then cautioned viewers to make the distinction between near zero RPO and zero RPO, as asynchronous technologies will only achieve near zero, not complete zero.
For enterprises that do require true zero RPO, Higgins recommended the need for a synchronous solution—namely, the ActiveCluster Business Continuity solution. This solution is fully automated, driving active/active, zero touch DR as opposed to active/standby complex failovers.
For an in-depth discussion, review, and demo of these Oracle DR technologies, you can view an archived version of the webinar here.