If SQL Backup isn’t top in performance tests, I’ll eat my hat

Compared to using native SQL Server to perform database backups, third party tools can create fast backups which are compressed and encrypted. This is the main premise for SQL Backup, performance. How quickly are your backups created and how much space do they take up? There are also other factors involved when reviewing the other products on the market, not to mention how much it will cost or ease-of-use, but fundamentally the performance has to be up there as a deciding factor.

There are 3 compression levels to choose from in SQL Backup. It’s a trade-off between optimising for speed or compression, so compression level 1 offers will compress the backup at high speed, while compression level 3 will take the longest (still quicker than native backups) but will result in the smallest backup file. In the middle, compression level 2 will provide a fast backup which is well compressed if both factors are equally important. It’s an easy 3-way decision for users to make, is it speed, compression or both which takes priority?

As part of the testing during development of new versions of SQL Backup, we regularly run tests on our test databases to monitor SQL Backup performance, recording speed and compression at all 3 compression levels. After boasting about the recent performance enhancements to the current released version of SQL Backup the idea was hatched to get an independent test carried out to compare performance with the other tools in the SQL Server Backup market.

The Tolly group have produced a report comparing the performance of Red Gate’s SQL Backup with Litespeed from Quest and Idera’s backup tool. The highlights of the independent test showed SQL Backup delivered throughput of up to 176 MB/sec (compared with 153 MB/sec from the nearest competitor) whilst still managing to achieve the best level of compression of all 3 applications.

They used a public data source, (the data from wikipedia) so that the tests can be reproduced by anyone on their own servers. The tests were required to conform to the fair test charter.

Thankfully for the development team the results they came back with justified the boasting since with or without encryption SQL Backup out performed both competitors on both compression and speed. No eating of hats required.