alex_aldo - Fotolia


Implementing a tiered storage architecture with hybrid cloud

By offering lower costs, higher performance and reduced data loss, hybrid cloud helps some organizations optimize their tiered storage systems.

Tiered storage architecture is nothing new. It's been a common practice for years to create systems with primary, secondary and archived storage to protect enterprise information. However, traditional tiered storage models and technologies proved to be both expensive and complex. As a result, many organizations now look to hybrid cloud to provide cheaper and more effective storage options.

Tiered storage is a storage model in which two or three separate storage systems work together. The first tier is the primary storage system and is typically the highest performer. The first tier is also closest to the application and is the most expensive. The second tier, or secondary storage, tends to be lower-performing and less expensive. It's designed either as an active storage backup system for the primary tier, or as a system that stores old or infrequently accessed data that is removed from primary storage. Finally, the archived storage tier is the cheapest -- but slowest -- system and stores data that is rarely used. 

Establishing a tiered storage architecture in hybrid cloud

In a hybrid cloud, the primary storage system typically exists on the private cloud if the applications are also on the private cloud. This storage system stores files and data for the applications, and should be the highest-performing storage system within the tiered storage architecture.

Because the price of local and cloud storage has fallen dramatically, using secondary storage as an active backup system for primary storage -- or as an up-to-date copy of primary storage -- is often the best approach.

The secondary storage system exists in the public cloud. Data is replicated or moved to a public cloud storage system, such as Amazon Web Services' (AWS) Simple Storage Service, to provide active backup for the primary storage system. In other cases, an enterprise might move older data to the public cloud to free up space in primary storage. 

Because the price of local and cloud storage has fallen dramatically, using secondary storage as an active backup system for primary storage -- or as an up-to-date copy of primary storage -- is often the best approach. If you do use secondary storage to store data moved from primary storage, be sure to design your applications to look for data in both storage systems.

Archived storage is optional, but tri-redundancy is typically a good idea. These are low-cost and low-performance systems designed for long-term storage of data that is no longer needed by core business processes, or to back up other backup systems. AWS Glacier and Google Cloud Storage Nearline, for example, are purpose-built for archival storage functions.

Weighing private vs. public cloud for data storage

The private cloud versus public cloud decision for cold data storage hinges on data amounts, length of storage and access needs, one consultant advises.

The benefits and challenges of a tiered storage architecture

Using hybrid cloud for a tiered storage architecture offers several benefits to an enterprise. First, because it establishes two or three storage systems at different price and performance levels, tiered storage helps developers and admins create application-specific storage strategies. For example, some data analytics systems interact almost exclusively with current data. As a result, an organization could move older data into a cheaper storage system.

Additionally, tiered storage in hybrid cloud creates a backup environment that spans two separate platforms -- public and private cloud -- to provide automatic redundancy that is geographically separate. Finally, tiered storage in hybrid cloud should reduce the risk of data loss, caused by disk failure or human error, to almost zero. 

The challenge, however, is to create a tiered storage strategy that accounts for all enterprise use cases and properly serves business processes. For instance, in the data analytics example above, that system might need to support other use cases, such as fraud detection through machine learning. 

Enterprises must carefully design and plan both their tiered storage architecture and automation process. Conduct testing to ensure your hybrid tiered storage system lives up to expectations. 

Next Steps

High-performance cloud storage becomes a reality

Cloud storage features you should consider

How to choose the right cloud storage system

Dig Deeper on Building and maintaining a hybrid cloud