Why are Data Backup Operations Often the Source of Performance Problems?
When considering performance enhancements or changes in data retention methods, it’s important to not only carefully plan, but also to examine why it may be necessary.
Performance Improvement
One common reason for evaluating data retention methods is performance improvement, including time necessary to complete the operation. Backups perform best when the data is sequential in nature. Just like the mail carrier who picks up your letters, he or she goes up and down the street visiting each home in sequence to pick up the letters. Imagine how inefficient it would be for the mail carrier to randomly go from house to house picking up your letters. It’s easy to see how this is akin to your backup process gathering up the bits and pieces of a file to copy, but it gets worse. Instead of taking your letters to the local post office, imagine if the mail carrier had to drive all over town dropping off letters at different post offices. This is what happens to the backup container file over time: you’re taking one performance hit on gathering up the data and another when writing the data. This randomization naturally occurs in the Windows NTFS file system and causes longer backups that eventually interfere with interactive user operations.
Other causes of backup performance problems include server configurations, network speeds and storage tiers. The leading backup vendors can address many of these issues, however the randomization of data is something only a few software vendors solve.
Other Pitfalls
Backups, Snapshots and Shadow Copies are a necessary part of any disaster recovery plan, but there are some other pitfalls. Backups do not capture changes made since the last backup, so work performed during that gap is not captured. Snapshots and Shadow Copy can fill some of these gaps, however scheduling constant snapshots and managing space utilization adds to the system administrator’s workload. There is nothing more aggravating to system administrator or help desk operator than a user calling in and saying, “Help! I’ve just deleted the major presentation I was working on and I can’t get it back!”
Fortunately, there are software solutions to automatically capture files as they are deleted and are placed into a recovery bin, including files created or modified since the last backup. This even includes capturing individual file versions so inadvertent document changes can easily be reverted. These solutions allow recovery without having to spend time crawling through a backup.
Many of these solutions are non-disruptive and easy to install. Given the importance of backups to the integrity and availability of an organization’s data, it’s worth looking into tools that can help improve performance and increase user satisfaction.
