A common fact in the IT sphere is that backup and restore failures are inevitable. With failure comes expense and possible data loss—something no one likes to experience. Do you want to do your best to avoid this problem? Below, I describe how you can reduce your chances of failure and strengthen your backup environment in five easy points. Though these points don’t guarantee constant success, you can be sure that if you address certain concerns, you will be much more likely to experience high backup success rates and more-favorable restore performance.
The Five Points to Improving Backup Performance
The five points below address the main areas of concern for most backup professionals. Like many problems, after addressing the areas directly, professionals will be better equipped to achieve the high results they want and need.
Point 1: The Monitoring Process
Monitoring in and of itself doesn’t cause failures during the backup process, but it can certainly make a failure more difficult to spot and unravel when it is happening. As the world of IT becomes more and more vast, backup systems must be altered to view and manage many backup servers. As it stands today, most systems were not designed to monitor so many servers at once, and if an issue does arise, the whole unraveling process can become tedious and difficult.
What is needed is a monitoring system with automated capabilities that will compile data and graphically provide a user interface, offering a comprehensive look at the whole environment. Also visible should be individual servers and clients. To streamline operation even further, this system must also be able to monitor the multivendor backups in use by a particular business.
Point 2: Missed Alerts!
Though email alerts sent to the proper administrators are generally a reliable way to communicate a problem, life doesn’t always stay the same. Servers, applications, backup devices and people change as time goes by, making it necessary to frequently verify and ensure that alerts are getting to the right people quickly.
A real-time alert is a great solution. This type of warning could be sent to a variety of people through email, SNMP integration and SMS, pinpointing the appropriate person to receive the information and communicating quickly and efficiently.
Point 3: Systematic Errors in Command-Line-Driven Operation
Though administrators may prefer the command-line interface to finish a job quickly, it is prone to errors! The reason is because of a lack of consistency in backups among various administrators when using this method. Best practices should be reinforced and codified through timely updates, but it doesn’t always happen, making errors a common occurrence.
A user interface that allows GUI operation of backup features must be added to the backup systems of reputable IT departments. Doing so decreases the chance for human error and improves the repeatability of current operations.
Point 4: Reports and Planning Receive Insufficient Attention
As tempting as it is for backup professionals to focus mainly on the report from the system that sent an alert, it is vital to remember that this information is only one piece of the puzzle in managing the backup environment. All too often, administrators miss many other important reports that are specific to their department and procedures.
When the backup server begins to hold the transferred alert and monitoring data, problems can soon follow. Generally, the data on the primary backup server is only saved for a short of amount of time, potentially resulting in it no longer being accessible and thus making the task of understanding and preventing the failure next to impossible.
Following the recommendation of best practices, it is prudent to compile the data from primary and distributed backup servers into individual databases, which will help to keep daily backup operations running smoothly. This way, the data can be analyzed and used in various reports relating to your department’s specific needs.
Point 5: Misconfiguration
Though IT departments are well-versed in backup and recovery systems, sometimes things can go wrong. Misconfigurations are one example; they often occur because of enlarged data and server environments. Here are some common problems:
- Recovery logs are size inaccurately: Such recovery logs can lead to lost information. This special space must be manually enlarged and restarted to avoid disaster.
- Mismatch from disk to tape: When using a small disk pool, there is a chance that new data won’t be accepted, delaying backups and causing missed backup windows. Only one thread can be written from a disk pool to a tape device, and if tape can’t handle the speed that data must be written from disk, the disk pool won’t be able to take backup data.
- Overload of simultaneous backup sessions: It’s easy to exceed the maximum client number of the backup system, not to mention missing the backup window. This problem can occur as data environments grow and add backup clients.
Regardless of mistakes, many IT professionals are using larger monitoring systems, which offer a more comprehensive view. This view grants an immediate way to find errors and identify a changing factor in the environment. Backup software and monitoring systems used together are the solution for IT departments to accurately assess the needs of the backup environment.
A Good Backup Environment: Art or Science?
A good backup environment comprises many things in the spheres of both art and science. Some are tools that assist administrators in being aware of trends and predicting the possibility of problems ahead. Another feature that a well-functioning backup environment has is monitoring and reporting tools that work well with backup software and with products from a variety of vendors. Along with these more science-based factors, keep in mind that backup administration is also an art that is perfected over time, and the knowledge of it must be appropriately shared to keep the art alive and well.
About the Author
Michael Johnson is a technical writer for Rocket Software.