Looking Ahead
If you look at the combined picture of data networking and storage networks, it is obvious that something must change. Storage traffic constitutes different protocols, such as Internet Small Computer System Interface (iSCSI), Network File System (NFS), Server Message Block (SMB), and others. Merging such protocols with data traffic over an oversubscribed infrastructure can have huge implications on the stability of storage access and the performance of applications. Also, networking in general needs to move from complex manual configuration to a more automated and dynamic way of network deployment. The advances in switching and routing hardware and software allow you to dramatically simplify the network and minimize the tiers. Advances in server virtualization have changed the way servers and compute are deployed in the data center, which challenges the existing storage networking designs. Hyperconvergence collapses the compute, networking, and storage tiers, offering a pay-as-you-grow approach to storage.
However, the basic needs of the enterprises do not change. Simplifying network deployments should not come at the expense of having highly available and robust networks. To understand the benefits of hyperconvergence, it helps to understand the current storage deployments and their challenges. You cannot solve a problem if you do not understand what the problem is.
Chapter 2, “Storage Networks: Existing Designs,” discusses some storage basics that might be familiar to a storage administrator but challenging for a virtualization or a networking engineer. It covers the storage multitier architecture; block, file, and object file systems; storage connectivity; storage protocols and protection methods; and advanced storage functionality. Part II of this book, “Evolution in Host Hardware and Software,” discusses the evolution in compute, storage, and networking that allows you to build next-generation hyperconverged data centers.