Data is really hard to move. It becomes pretty much intractable as it increases in mass. Sure it can be pulled out, but that takes time and bandwidth and incurs a host of costs typically associated with using a cloud service. To really get the most of all that stored data, it’s increasingly apparent that moving it is not really a good idea.
Joyent’s new Manta Storage Service puts the compute together with the data in the cloud where it can be processed in one place. The compute is available directly on the object store, meaning that the data can be queried immediately without having to manage all the underlying infrastructure.
The new storage allows customers to analyze log data, financials and other data-intensive functions without moving data into separate clusters, which can take hours depending on the amount of data needed to be processed. Services like Amazon Web Services, by contrast, separate compute from storage, which can mean a lot of time and cost spent just moving data around.
The ramifications of the Manta service are considerable and make Joyent more relevant as an infrastructure and a services provider. In an interview last week, CTO and Co-Founder Jason Hoffman said Joyent will make a number of announcements in the coming months about new data services.
The Joyent news is another example of the disruption to the network attached storage market. Joyent Manta Storage means a customer can spin up instances without waiting. This means that the company can charge by the second as opposed to by the minute or hour. It also means data gets processed in one place without the need to spin up any number of servers to keep a service running.
That’s the big trick here, and it took Joyent four years to research, develop and make it happen. Now we’ll have to see how much traction Joyent can get in a market that includes AWS, Microsoft, IBM and a host of other competitors.
