It’s been staring me in the face for a while, all the training, advising and talking about this Hadoop thing. It needs an AppStore, a seamless platform for one click installing and running of Hadoop applications within the framework and cluster.
Up until now most MapReduce work has been hand coded by in house development teams working to a specific requirement. Hadoop2 and it’s Yarn component extends on that and makes applications easier to distribute across the cluster, like I said in yesterday’s post about deploying point of sale software across kiosks.
The issue is that Hadoop takes an amount of skill to get things running, thinking of getting data in and running scripts at so on. We’re a way off from a one touch solution for app and data deployment and processing.
It’s coming, I’m sure it is. The next two years will be very interesting.
MapR got the ball rolling with an application gallery, not an AppStore par se but a great start.