This article can also be found in the Premium Editorial Download "Storage magazine: Storage Products of the Year 2010."
Download it now to read this article plus other related content.
In host-based replication products, the replication is performed by so-called filter drivers on servers that intercept write I/Os so they can forward file or block changes to replication targets. Host-based replication has the lowest entry cost and complexity, but these increase proportionally as the number of replication nodes expand. While it's easy to manage for a small number of servers, managing a large number of nodes, from initial installation and rollout to ongoing support and monitoring, can be a daunting task.
Compared to array- and network-based replication, host-based replication is less isolated and its execution environment less controlled and, as a result, it can be adversely impacted by other applications and server events. Virus infections, resource shortages and application crashes can't bring down array- or network-based replication, but they can definitely stop host-based replication. Standalone host-based replication products are available from the likes of CA with ARCserve, Double-Take Software Inc. (now part of Vision Solutions Inc.), Neverfail Ltd., Quest Software Inc. and SIOS Technology Corp. (formerly SteelEye Technologies Inc.). The products differ in platform support, with Windows supported by all, and features, such as throttling, compression, deduplication, encryption, high-availability (HA) capabilities
and management options.
Host-based replication is the nimblest of the three approaches. It's software only; works with any type of storage, including DAS and cloud storage; and is capable of supporting a wide range of platforms, depending on the replication product and vendor. Even though it competes in scenarios where array- and network-based replication may be used, the fact that it's pure software allows it to extend into areas where the other two simply can't compete. First, it's the ideal replication method for cloud storage. Cloud services like the Amazon Elastic Compute Cloud (EC2) simply can't deal with hardware-based replication, but they allow running replication software to exchange data between clouds, and between the cloud and on-premise servers.
Secondly, host-based replication software can be incorporated into other apps. Prime examples are backup apps that have been adding replication to provide and manage both replication-based and traditional data protection. Replicas aren't usually standalone images, but are likely part of a larger data protection process. As replication-based data protection moves mainstream, being able to manage and monitor them along with traditional backups becomes more relevant. As a result, CA, EMC, IBM, Symantec Corp. and others have been adding replication-based data protection to their backup suites.
The evolution of replication
Replication has been used for high availability and data protection of critical data and applications for a long time, and it has been slowly eating away at tape as the media of choice for data protection. As a consequence of shrinking recovery time objectives (RTOs) and increased need for 24/7 availability of applications and data, this trend is likely to continue, if not accelerate. On the technology side, cloud computing, the virtualization of IT infrastructure, and a flurry of replication options and offerings are aiding this trend.
BIO: Jacob Gsoedl is a freelance writer and a corporate director for business systems. He can be reached at firstname.lastname@example.org.
This was first published in February 2011