Some of our Windows 2000 hosts do not run Veritas Volume Manager. We would like to SRDF those volumes to a remote site. Will there be any problems in mounting those remote volumes on the remote hosts? What is the advantage of having Veritas Volume manager, as far as SRDF is concerned?
With the wide variety of volume managers and remote data replication products on the market, it's no wonder you're confused. The two products that you are asking about serve different purposes, but can actually be used to create similar solutions. The challenge is that each solution acts at a different level.
Veritas Volume Manager is a host-based product, while SRDF is an array-based replication product. A host-based volume manager is usually used to create concatenated or striped volumes from a set of disks, as well as creating mirrors of the set of disks locally. Veritas Volume Manager has an optional toolset called Veritas Volume Replicator that works on top of Volume Manager, creating remote mirrors of the local disks to another host licensed with Veritas Volume Manager and Volume Replicator. From a data availability perspective, you can create this remote mirror in a variety of ways, including synchronous and asynchronous. SRDF can also provide remote mirrors that are sync or async, as well as adaptive, which is a certain number of 32K tracks out of phase with the original data set. This is all user-configurable.
The key difference here is that Veritas' solution can be used with disparate disk vendor solutions. You can have EMC disk arrays at the primary site and HP disk arrays at the remote site, or even locally connected JBOD storage, while with EMC's SRDF, both ends have to be EMC solutions and similarly configured disks. Veritas Volume Manager and Volume Replicator only care that the amount of disk space is the same, the underlying disk configuration isn't as important. Also, Veritas Volume Manager provides snapshot capabilities and integration with Windows from the host perspective, while SRDF is not aware of anything going on at the host level and requires TimeFinder integration.
When considering which to use, think about the specific challenge that is before the organization and lay out the requirements for a solution upfront. Cost and complexity of management are just a couple of the criteria, but you should also look at the requirements of the application data that is being served. Is the data critical to the organization, how fast a recovery is required? If you need sync replication then also look at the network costs and complexity and evaluate accordingly.
Related Q&A from Brett Cooper
We are currently using an HDS9960 array and are planning to upgrade early next year to another enterprise array. We're in the process of evaluating ...continue reading
We are looking at purchasing a new storage system for our financial application, what should we be looking for? The system needs to provide very high...continue reading
Brett Cooper answers the question: What is the best practice for moving AIX 5.2 data, including an Informix database, over to a SAN that is attached ...continue reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.