Last summer in T100442, we expanded the Hadoop cluster. Analytics reserved some FY2015 hardware budget for future expansion. The future is now!
We'd like to spend the remainder of our budget on Hadoop nodes. These nodes will be used for regular Hadoop expansion, as well as to help support productionization of realtime streaming analytics jobs, using YARN based streaming frameworks like Spark Streaming.
T124945 and T124947 will come out of Analytics FY2015 HW budget first. Once those are quoted, we can allocate the remaining funds to buy a few more Hadoop nodes.
We can order nodes with the same capacity as we did last summer:
Dell PowerEdge R720
- 12 core EW-2620 @ 2.00GHz
- 64G RAM
- 12 x 4TB 3.5 HDDs
- 2 x 2.5 small HDDs (in chassis?) for / partition
Note that this expansion is explicitly for this FY. Next FY will come with more hardware requests. :)