This blog is part of a series on Data ONTAP 8 operating in Cluster-Mode, and focuses on what comes with the cluster. The first blog in this series was “Why Scale-Out & What’s New”, the second blog was on “Unified Scale-Out – What’s Supported & Why It’s Unique” the third blog focused on “Nondisruptive Operations – What Does it Mean? and the fourth blog was on Enterprise Ready Data ONTAP 8 Cluster-Mode.
For background, a cluster node is a storage controller, and nodes in the cluster can be different models and sizes; for example, a FAS2240, FAS3270, FAS6280, FAS3040 and V-Series (front-end 3rd party storage from EMC, HDS, IBM, etc.) could all be in the same cluster. Individual disks are grouped into an aggregate, which is a group of disks of a particular type that are protected using NetApp RAID-DP, which protects against double disk failure.
A 10GbE dedicated, redundant, high-throughput network is used for communication between the cluster nodes and for Data Motion (lets you easily and quickly move data from one logical or physical storage device to another while maintaining continuous access to your applications), and is included with every cluster configuration. Management network switches are also included.
The figure below depicts a cluster that supports a mix of controller types, storage protocols and supported disk types to match the performance and cost of the storage to the requirements of the data and workloads. The left of the diagram depicts a performance data processing area within the cluster built using high-end, faster controllers (e.g. FAS6280) and high performing SAS and/or SSD drives. The right is an archive built with midrange controllers (e.g. FAS3240) and high capacity SATA drives for the best $/GB. If the requirements change, data can be nondisruptively moved within the cluster for load balancing performance or capacity. For example, when a high performance project such a semiconductor chip reaches completion, it can be moved to lower cost storage. When it is time to make the next rev of the chip, it can be moved back into the high-performance processing area.
The storage cluster can be expanded, contracted, and subdivided nondisruptively into secure logical partitions or Vservers. The next blog in this series will focus on Vservers.