Reliability and availability have always been critical reasons why enterprises select Riak as their database for highly scalable applications. In 2016, Riak is making a concerted effort to communicate what these attributes mean in terms of business value to our customers.
We are all well aware that application downtime is not an option. It can have a dramatic impact on revenue and lead to negative customer experiences and real costs. Gartner published a survey that reported the average cost of downtime is $5,600 per minute, or approximately $300,000 per hour. To further understand the Riak customer experience, we reached out to customers directly and also via the third-party research platform TechValidate. Following are some of the highlights from both the research and our conversations with customers.
- Nearly half of the Riak customers surveyed said that downtime costs them $50,000 to more than $100,000 per hour in lost revenue.
- The biggest factors that Riak customers look for when choosing a NoSQL solution are:
- Riak KV means resiliency – your application stays functional and performant. 72% of Riak customers who responded to our survey said that they saw no increase, or only a slight decrease, in the cluster processing power of their applications when a node in their Riak cluster went down.
The question “What would you tell others about the value of Riak to your business when it comes to minimizing the costs associated with data downtime?” elicited the following responses:
- “Riak is very dependable. The documentation and support are excellent. WAN replication was simple and straight-forward to setup. The ability to tune buckets and bucket-types between high availability (AP) and high consistency (CP) is a plus. The CRDT types that support predefined merge strategies is a feature that sets Riak apart from its competitors.”
- “Riak can be managed by a single person. With other solutions, it is recommended to have a team of people to manage and keep a solution up. The reduced overhead of managing Riak allows us to focus on other solutions, because we’re confident that the data will remain available.”
- “[The] Riak platform is very reliable and easy to understand, operate and use. It nicely handles high loads and node failures. The documentation and support (both community and commercial) are excellent. For developers, there is a variety of client libraries to choose from.”
- “[Riak’s] Top-notch stability can be relied upon to support mission-critical, customer-facing applications.”
“Find Zen in Zero Downtime” is the title of an infographic we’ve created that captures some of our survey results and other industry data – click the graphic to expand.
Matt Davis, site reliability engineer, from OpenX, a leading online advertising platform, made it clear that “choosing a NoSQL database with masterless architecture that minimizes downtime is extremely important.” You can hear more from Matt about why he chose Riak KV in a live webinar he will give on March 24th. Registration is here, and as long as you register we’ll send you the link to the on-demand recording so that you can watch at your convenience.
Whitepages’ Sr. Architect, Devin Ben-Hur, stated, “The combination of technical excellence, right design principles, excellent engineering, and meeting all of our scaling and resiliency needs made Riak KV a perfect choice. The product just works.”
We have coined the term RiakABILITY as a description for the resilience of the Riak products and launched a campaign stressing the importance of database resiliency to mitigate the cost of application downtime and eliminate operational nightmares.
Want to find out more about why Riak is so resilient? Read about Riak KV Enterprise in this technical overview or hear from Devin Ben-Hur in this short customer video about why he made the shift from Postgres and Hbase to Riak KV. We sat down with Matt Davis and discussed OpenX’s experience with Riak KV Enterprise in this video.
What the Riak are you waiting for? Get some RiakABILITY into your environment and sleep better like Matt:
“…a single node goes down, even up to three or four nodes going down, depending on whether it’s one of our big clusters, we can wait until the next morning to fix that stuff or we may not even fix it for a couple days. You can let that stuff go and Riak is resilient, so prone to staying available that we just don’t have to worry.”
Peter Coppola
Vice President, Product and Marketing
@pbcoppola