The word “tier” is used frequently in the context of data centers. As we explained in a previous blog post, data center facilities are ranked by Tier depending on their infrastructure and redundancy. But tiers are also used to describe the resources and infrastructure of a virtual machine or server, and are often mentioned in billing or deployment. This post aims to clear up any remaining confusion over the two types of tiers.
When deploying a new virtual machine or application, developers often use “tiers”, each geared for different performance or storage types depending on the expected audience and usage. In the case of Green House Data, Tier 1 is connected to Fibre Channel SAN, Tier 2 is SATA SAN, and Tier 3 is Isilon NAS. All storage tiers are RAID 5 configured. These tiers allow IT deployments based on performance needs. For example, development and test environments can be built on Tier 3 NAS infrastructure, while client-facing production environments can be deployed on Tier 1 Fibre Channel SAN. In combination with a pay for use billing cycle, this strategy can considerably lower the cost of infrastructure deployment cycles.
Developers and system administrators may design their own tier systems depending on the project at hand, customizing the operating system, development tools, routing systems and network settings appropriately. A new application might feature a front-end tier with load balancing for users, a middle application tier that runs the app, and a back-end tier loaded with a database, for example.
These deployment tiers are quite different from the facility rank tiers we previously discussed. Check out our previous post for an in-depth assessment of how facilities are ranked Tier I – Tier IV. Have you seen “tier” used in any other data center context? If so, let us know!
Posted By: Joe Kozlowicz