So, what can you do to restore MySQL database scalability and throughput?
One option is to scale up using bigger hardware, but this will hit a point of diminishing returns. Perhaps you’re already running on Amazon’s largest available configuration. Application re-architecture is another possibility, but the time and expense can be a real problem.
At ScaleBase, we believe automated and simplified scaling out — or MySQL database sharding — is the way to go; enabling limitless MySQL scalability by leveraging the power of multiple database servers acting as one. Many companies, including Tumblr and Facebook, have successfully scaled out their MySQL databases.
Historically, MySQL sharding used to be hard. In fact, some vendors still use the tag line “sharding sucks” to exploit that out-of-date reputation. Nowadays, policy-based data distribution (horizontal partitioning or creating MySQL shards — there are many terms for much the same thing) can be automated and simplified.
In the next series of blog entries I’d like to discuss how today technology eliminates ten challenges historically associated with do-it-yourself, home-grown database sharding.
Today let’s look at three that all relate to issues that used to exist because of multiple MySQL databases.