Home > Blogs > Virtual Blocks > Author Archives: Duncan Epping

Author Archives: Duncan Epping

Duncan Epping

About Duncan Epping

Duncan is a Chief Technologist working in the Storage and Availability Business Unit at VMware, serving as a partner and truster adviser to EMEA customers. He is the co-author of several books, including the vSphere Clustering Deepdive series and Essential Virtual SAN. He is the owner and main author of virtualization blog yellow-bricks.com.

Two brand new customer case studies published!

Over the past few months I’ve spoken with many customers about Virtual SAN. These conversations usually result in “unofficial” success stories that are published in the form of a blog post. In addition to these informal stories, businesses that successfully implement Virtual SAN are featured in customer case studies. Below are two new customer case studies that highlight how Virtual SAN transformed their IT infrastructure and positioned their business for future growth.

The first study describes how highway infrastructure concessionary Autopista Vespucio Norte (AVN) has leveraged VMware virtualization solutions since 2008 to reduce capital and operational expenditure. vSphere 5.5 and Virtual SAN was them the next step in this process. The complexity of their project was mainly due the fact that as this was an unprecedented technology in the market: since they were literally one of the first to implement Virtual SAN they could not rely on previous experience or proven success cases to validate the VMware VSAN solution. By converging layers of complexity and delivering over 65TB of storage capacity with just IBM 3 hosts all connected using 10GbE Arista switches they managed to reduce capacity expenditure with 50% and operational expenditure with 60%. Click here to read the full AVN customer case study.

The second case study talks about Keck Medical Center of USC which was in dire need of a new data warehouse for business and medical analytics but didn’t have room for the data in its conventional SAN. The center’s IT team quickly chose QlikView Business Discovery as its data management solution. Unfortunately, they also discovered that the center’s SAN lacked the necessary capacity and horsepower and that adequate funds to expand it were unavailable. With no available SAN capacity, a limited budget, and an impatient cohort of analysts, the Keck team urgently needed a more flexible and affordable storage solution. Storage they could deploy quickly with just the capacity they needed to launch. Storage they could scale easily and affordably as application requirements changed. Storage that would be easy to manage and that wouldn’t lock them into expensive proprietary hardware. The Keck team’s solution was to create a pair of storage clusters, building out existing VMware vSphere hosts (Dell PowerEdge R715 servers) with internal hard drives and Fusion-io flash drives. Click here to read the full Keck Medical Center of USC customer case study.

Support for non-uniform/heterogeneous VSAN clusters? Yes!

Over the last couple of weeks I’ve been asked this question by various field people and potential customers: does VMware support non-uniform / heterogeneous VSAN clusters? The question typically comes during the pre-sales stage and by the sounds of it is one of those myths that competitors use to give their product the edge in a bake-off.

For those who just care about the facts I can be short: Yes we support non-uniform / heterogeneous VSAN clusters. Our very first datasheet has the following statement in it:

Hardware independence – Virtual SAN can be deployed on hardware from any server manufacturer. This gives you the flexibility to build out customized storage systems in heterogeneous hardware environments.

Anyone telling you VMware VSAN does not support non-uniform / heterogenous environments is badly informed, or not being honest. Some of you may say, hold on… I’ve seen posts where people are making a recommendation for uniform clusters. A strong recommendation even in some cases, does that mean VSAN doesn’t function properly when you use a non-uniform cluster? Again, I can be short: No it does not. But why did they make this recommendation?

The reason is simple: lower operational complexity. Just like with vSphere HA and vSphere DRS, balancing resources and doing calculations to see if your cluster has sufficient spare resources to reprotect VMs after a failure is easier when you have a uniform cluster. Not that it is rocket science when the cluster is non-uniform, as you will just need to assume that if something fails it will be the host providing the most storage capacity. Luckily with VSAN the whole cluster acts as “rebuild capacity”, which means that mitigating this risk is relatively easy.

As an architect I usually prefer to avoid risks though, but also understand that situations like these occur and there is no reason to fear them. I spoke with a customer who is managing a rather large non-uniform environment last week. They started with 1U hosts, but for various reasons decided to go with 2U hosts more recently. Using the same cache to capacity ratio per host they’ve had no challenges whatsoever in this environment (they even went with more cache as SSD prices lowered). Performance and management has been a breeze, they are very keen on standardization but the ability to be flexible when needed is actually what they love about VSAN.

Virtual SAN going offshore

Over the last couple of months I have been talking to many Virtual SAN customers. After having spoken to so many customers and having heard many special use cases and configurations I’m not easily impressed. I must say that half way during the conversation with Steffan Hafnor Røstvig from TeleComputing I was seriously impressed. Before we get to that lets first look at the background of Steffan Hafnor Røstvig and TeleComputing.

TeleComputing is one of the oldest service providers in Norway. They started out as an ASP with a lot of Citrix expertise. In the last years they’ve evolved more to being a service provider rather than an application provider. Telecomputing’s customer base consists of more than 800 companies and in excess of 80,000 IT users. Customers are typically between 200-2000 employees, so significant companies. In the Stavanger region a significant portion of the customer base is in the oil business or delivering services to the Oil business. Besides managed services, TeleComputing also has their own datacenter they manage and host services in for customers.

Continue reading