Organizations of all sizes implement virtualization to gain efficiencies for both IT and the company. Typically, the server admin group leads this charge by running other virtualized servers on “virtual hosts,” or existing physical servers. In other words, it is possible to implement server virtualization without ever changing the server, storage, or networking hardware. From a cost perspective, this is a huge benefit! However, before you get started you must analyze your server and storage hardware to find out:
- Are your server and storage hardware compatible?
- Is a more efficient solution available?
You may decide to purchase a few new servers and consolidate many older servers on them, or you could replace an old physical server by leasing a new server; using the newly leased server as your virtual host. Either way, when you analyze hardware, physical servers usually get most of the attention and storage is left as an afterthought.
Considerations When Virtualizing Storage
Traditionally, physical storage pools are tied to specific applications. In a virtualized environment, CPU resources are allocated dynamically and utilized by mission-critical applications. Although virtualization reduces the number of physical servers required, it actually increases the amount of traditional storage capacity needed. When the server side is dynamic, the storage pools must also be dynamic – since they can no longer be tied to a particular set of physical servers, so when virtualizing mission-critical applications, keep these factors in mind:
- The most critical data should be served from fastest storage.
- Adding high-performance, high cost disk drives is an expensive solution.
- There is a limit to just how fast mechanical disks can spin.
- Using intelligent storage balances high performance and low budget requirements.
- If you optimize servers and not storage, you don’t fully realize the benefits of virtualization.
Since traditional storage does not have dynamic capabilities, you need to anticipate application performance metrics and add storage as-required with more expensive 15,000 RPM SAS or FC disk drives. Adding expensive disk storage eats away at the savings generated by reducing physical servers.
5 Reasons to Consider Storage Tiering
The alternative to purchasing high-end storage is storage tiering. A storage tiering method mixes Flash and Hard Disk Drives (HDDs) with intelligent, automated tiering software. Using Flash drives in your storage virtualization strategy offers:
- Easy virtualization integration – Your array is optimized for virtualized environments.
- CapEx reduction – With fewer overall drives on smaller, more powerful arrays.
- OpEx savings – By tiering data to the cache and across the storage pool.
- Performance improvement – 30 times the IOPS of a mechanical drive and less than 1 millisecond response time.
- Faster application performance – Tune Microsoft SQL Server takes 80% less time; Oracle database OLTP applications see a 100% increase in transactions per minute.
Any datacenter design and virtualization consolidation plan should take a holistic approach to the datacenter, it’s not just “how many VMs can I put on a host”. By utilizing today’s faster, more efficient storage array with data tiering, you increase application performance and reduce both capex and opex at the same time.
The moral of this story is that, whether you are interested in storage or not, you must analyze storage as part of your move to virtualization.
To learn more, download our ebook or go to the Softchoice Advisor blog to read about the technologies that make storage tiering possible. Please feel free to contact me to find out how storage tiering can make a difference in your environment. You can leave a comment below and we’ll respond as quickly as we can.
About David Davis
David Davis is the author of the best-selling VMware vSphere video training library from TrainSignal.
Recent Conversations