Adaptive Solutions Inc.

Database availability - An Oracle Corporation study

There was a time when an average DBA with junior level skills could administer the average database. However, maintaining high availability is no small task in a client/server environment. In fact, providing around the clock, (24 hours a day, 7 days a week), availability requires your database to be virtually available without a margin for downtime. What is the cost for downtime? A recent survey shows that, businesses, on average, incur $1,300.00 of mean revenue loss per outage minute. This survey was performed on 450 companies in manufacturing, securities, health insurance, transportation, banking, retail, and telecommunications industries. Taking this study into account, a business that maintains 99% availability in a 24x7 environment (1% downtime), incurs approximately 5,000 minutes of downtime per year. This small amount of downtime equates to over $6 million dollars of lost revenues per year.
What causes downtime? Many operational issues come contribute to this problem. The most notable contributor is hardware failure. Second is operational outages. Operational outages are caused by human intervention frequently attributed to poor DBA skills, user errors, inappropriate database setup, or inadequate backup procedures. Can operational outages be completely eradicated from your list of downtime culprits? No, but operational outages as a surce of downtime can provide the greatest opportunity for downtime cost reduction.

According to Oracle Corporation studies of client downtime, 34% of downtime can be attributed to operational outages, or outages that you can control. Another 36% of downtime is caused by design-related issues. Over half of the design-related issues are due to a lack of adequate documentation. Approximately 5% of downtime is caused by configuration issues and problems associated with the setup and installation of the database. All of these outages are issues that you can control and many times are attributed to poor DBA skills.
This study would indicate that the lion’s share of outages could be prevented. The average corporate out-of-pocket cost for downtime could be reduced by over $4.5 million per year. When adequate preventative measures are applied through the use of appropriate database monitoring procedures, a huge savings can be achieved. Proper expert advice on database configuration, and backup and recovery strategies is equally important in reducing this number to the smallest amount possible.
Finally, the survey showed that the average Mean Time Between Failures (MTBF) was 102 days, and the average Mean Time to Recover (MTTR) was calculated to be 17 hours and 53 minutes. These statistics could be the reality of your users and the company as a whole. Can you afford to have these unexpected outages?