Improving storage performance with fast external cache systems

Send to friend

Database applications are at the core of almost every major business today. The unparalleled growth of large database applications however, has created severe performance issues for end users. Too many users, too many transactions, and too many queries are slowing down systems, costing additional business for many and grinding business to a halt for some.

Often, additional processing power alone will do little or nothing to improve database performance. This is because the processor, no matter how fast, finds itself constantly waiting on mechanical storage devices for its data. While every other component in the data chain moves in terms of computation times and the raw speed of electricity through a circuit, hard drives move mechanically, relying on physical movement around a magnetic platter to access information.

When servers wait on storage, users wait on servers. This is I/O wait time. RAM-based cache is the most common remedy for I/O wait time on database applications, but comes with its own set of challenges. Large monolithic RAID systems that incorporate high capacity caches are prohibitively expensive and cannot be upgraded in a modular fashion.

To solve this problem, more RAID manufacturers are partnering with solid state disk companies such as Texas Memory Systems to deliver a fast external cache system alongside high-capacity RAID. The RamSan-330 cache system, when installed alongside a RAID device, provides the fastest access to data for applications and users.

The combination of RAID and fast RamSan cache accelerates data access by providing a high-speed data pipeline. By temporarily writing data to fast DDRRAM instead of conventional rotating disks, the solution drastically increase burst bandwidth and I/O rates, ensuring that existing storage systems will survive under immediate pressure. Smart caching algorithms pre-fetch data that users are likely to access, which increases overall performance by enabling applications to read popular data directly from cache.

Additionally, the system accelerates the transfer rate of the RAID itself through a process called data smoothing. Random data access behaviour is extremely detrimental to conventional hard disk performance because rotating drives are optimised for sequential data reads and writes. In tests, while data was reading and writing to cache at the speed of RAM, the RAID itself was writing data at The reason is that the external storage cache is smoothing the data by converting random data patterns to nearly sequential data patterns. Because the RAID is responding faster, the external cache is more effective, and as a result the application is responding faster to the user.

This overall improvement in storage performance will eliminate storage performance bottlenecks for database applications, accelerating user concurrency, query times, and transaction rates. As applications and data volumes continue to expand storage environments, most organisations face tough choices: data migration challenges, matching capacity and performance requirements, and the prospect of massive hardware upgrades. Independently scaling storage performance by adding external cache devices to the storage network alleviates these pains.

Cache systems already have a permanent place in modern IT environments. As business applications become more demanding, these external storage cache systems will become more necessary. Monolithic RAID configurations are not scalable to meet enterprise application needs and are too expensive for the datacentre looking for performance. Combining the worlds fastest RAID controllers with the worlds fastest external storage cache offer high capacity and high performance while limiting overall cost.

Comments (0)

Add a Comment

This thread has been closed from taking new comments.