Today’s database administrator (DBA) has many responsibilities related to protecting and maintaining access to enterprise data resources. There are many diverse tasks and activities required to keep databases up and running smoothly. They range from the mundane, such as creating login credentials for new employees, to conducting deep explorations to determine the reasons that database performance is lagging. One of the characteristics that make a good DBA is the flexibility to do whatever is needed on a given day to manage and protect their systems.
Why Data Protection is Critically Important
An organization’s databases contain its most valuable resources. Losing access to its data assets and applications unexpectedly can cripple a business. With the prevalence of eCommerce solutions and the need to maintain twenty-four-hour availability for customers and employees, even a short outage can cause multiple problems. A reliable backup and recovery strategy is vital to protecting enterprise data.
Outages can be categorized as either being planned or unplanned. Activities like upgrades and infrastructure maintenance may necessitate that systems are inaccessible for an expected length of time. These types of outages may be disruptive, but proper communication with the affected parties should ensure they do not cause serious problems.
Unexpected outages are more problematic and as the name indicates, often happen at the most inopportune time. They can be caused in many different ways that all need to be addressed quickly. The impact of extreme weather events, failures of the power grid, and cyberattacks are just a few of the ways that system outages occur. One of the major issues related to unexpected outages is that recovery time can be extremely variable. The average downtime in the wake of a ransomware attack has increased to over 16 days. Many businesses would be hard-pressed to handle a lengthy outage that affected mission-critical databases and applications.
Creating a Backup and Recovery Strategy
Among the most important tasks that a database team performs is developing a viable backup and recovery strategy to protect against unexpected outages. This is not an activity that can be done haphazardly or without the input of other stakeholders. The first step in designing a backup and recovery plan is to answer the questions that will enable a recovery time objective (RTO) and recovery point objective (RPO) to be defined. These two objectives will be used to determine how backups and recovery are performed in your environment.
The RTO specifies how long your organization can afford to be without a system or application. This value will vary widely based on the business importance of the system in question. Extremely critical systems may not be able to tolerate any downtime and should be implemented with automatic failover and mirrored storage to maintain constant availability. The choice between restoring or rebuilding a system and the methods used need to take into account its RTO.
An RPO is a measure of how much data loss can be tolerated when planning a backup and recovery strategy. Unless the system is mirrored with a hot failover, some data will inevitably be lost when an unexpected outage occurs. This is due to the time interval between when a backup was taken and when the outage began. If backups are taken once a day and systems are impacted just before the next iteration begins, up to 24 hours of data could be lost.
Critical systems may need to be backed up more frequently than daily to achieve an RPO that is acceptable to the organization. A combination of full and differential backups along with backing up transaction logs may be what is needed to sufficiently protect your SQL Servers.
In complex environments, there may be multiple RTOs and RPOs that address different aspects of the infrastructure. The time constraints on these elements are influenced by SLAs and business considerations. A one-size-fits-all mentality may not be suitable for protecting your data resources. RTOs and RPOs should be developed for each system or related groups of systems that reflect their importance.
Regardless of the values determined for the RTO and RPO, a good backup strategy should follow the 3-2-1 backup methodology. This method proposes that you need to have at least three copies of your data. So at a minimum, you want to have the live data that resides on your servers and at least two backup copies. The backup copies should be on at least two different media formats, for instance, a disk and tape backup. Finally, one copy at a minimum should be stored offsite, perhaps in the cloud, to afford protection in the event of a physical disaster that destroys your environment.
An extremely important aspect of any backup and recovery plan is that it be fully documented and tested so that when it is needed there is no confusion as to what activities need to be done and who will perform them.
Using SQL Safe Backup
An IDERA Webcast is available that goes into further detail regarding backup and recovery strategies for your SQL Servers. It includes a thorough discussion of how SQL Safe Backup can be a valuable tool in protecting your SQL Server environment. It contains valuable information for DBAs looking to safeguard their SQL Servers. The application contains many features designed to make it easy to protect your enterprise data resources.
SQL Safe Backup helps minimize the pace and time required to perform backups with multi-threading, dynamic compression, and encryption. Compression versus available space is automatically adjusted to provide the fastest and smallest backup files. The tool also provides multiple recovery options including the ability to bring up databases instantly to handle emergencies. A combination of a good backup and recovery strategy with a powerful backup application will keep your SQL Servers protected from unexpected downtime.