Kickfire took a major step for a small vendor last year by passing the most credible benchmark in data warehousing: the TPC-H. For those not familiar with it, TPC is a non-profit association whose benchmarks require all vendors to use the same workload, thereby producing comparable results. Of course, performance is a key measure; but perhaps the more important benchmark result is the price vs. performance ratio. While a vendor could throw large amounts of hardware at a workload to produce high-performance results, making it cost-effective has always been the biggest challenge. TPC-H rates the vendors' solutions using its own metric, called the "Composite Query-per-Hour Performance Metric" (QphH). The metric is a comparative measure of effective query throughput and processing power relative to data warehouse workloads.
Kickfire's results are impressive. For a 300 GB workload, it currently ranks fourth on the performance list (and second on the non-clustered performance list). But what stands out is its price/performance ratio. Kickfire has the lowest cost per QphH of any vendor, at $0.89. Compare that to the fifth-placed product on the performance list, an SQL server-based solution that costs $5.40 per QphH, and the sixth-placed product, an Oracle-based solution that costs $18.67 per QphH!
Kickfire announced last week on its blog that it had shipped its first appliance to a Web 2.0 customer. As many Web 2.0 businesses are finding out, killer features alone do not determine success or failure. Success also depends on how well the vendor understands its customers' evolving needs and how it generates revenue by addressing those needs. Kickfire's Web 2.0 customer is using the appliance to do click-stream analysis to better understand its own users' behavior, so that it can target relevant advertising offers. According to Kickfire, the customer has around 500 GB of imported data, but this is expected to grow at a rate of 1 GB per day.
The traditional enterprise space is less of a focus for Kickfire at the moment, partly because that space is already relatively well served by specialized offerings, but also because MySQL has less of a presence there. While Sun is pushing MySQL to break through those enterprise walls, most corporate data platforms remain largely the domain of Oracle, Microsoft, IBM, and niche vendors, such as Teradata. If Sun does manage to break through, then the road will be paved for Kickfire to follow.
Kickfire is a stand-alone solution and can of course be loaded with data from any data source, but the ease of adoption for existing MySQL customers, combined with its strong price/performance ratio, makes Kickfire a compelling option for Web 2.0 businesses looking to add a data warehouse platform to improve their analytics capabilities.
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteEnterprise posts
Good article. Tweeted about it (I'm @vasudevram) here:
http://twitter.com/vasudevram/status/1141318168
- Vasudev
Thanks for a great post. We use mysql a lot and have been having troubles working with large tables. Good to know about kick fire and I will be looking into this more. Do you know if it is available outside the US?
Hi Alex, I spoke with Kickfire and in these early stages they are focusing just on the US market. But it can't hurt contacting them if you are elsewhere and interested, I'm sure they will make it available once the demand is there.