Given how central IT is to most 21st century organizations and its criticality in achieving true "digital transformation," it is so very surprising that so many companies are still failing to adequately prepare for IT-related disasters.  To gain a better understanding, I spoke with DH2i's Connor Cox, Director of Business Development, on this important and timely topic.

VMblog:  In your experience, why do companies fail to prepare for IT-related disasters?

Connor Cox:  I see three primary reasons companies fail to prepare.  It's not that they are unaware of the issue. Instead, a main issue is the cost. Comprehensive, fail-proof disaster recovery (DR) comes with a significant price tag. It requires an extremely complex system of hardware, software, and personnel. It's also very difficult to configure a solution that protects both legacy and newer applications - onsite and in the cloud.

There are also issues of latency. In short, traditional approaches to true high availability (HA) and DR are complex and costly, especially for the most highly available solutions.

VMblog:  What traditional approaches to high availability and disaster recovery are most common?

Cox:  For Microsoft SQL Server specifically, one of the most highly deployed platforms in the world, traditional approaches include Always On Availability Groups, which provides database level management, as well as replication for critical databases. Another approach is Failover Cluster Instances. Also, you can use virtualization in combination with either of these approaches, or independently.

VMblog:  Are these traditional approaches sufficient to meet organizations' needs?

Cox:  They both present challenges - that's why we are still seeing organizations unready to cope with IT disasters.

A central challenge preventing optimal HA or DR with all of these approaches is the cost. In order to achieve higher availability for SQL Server, the most expensive software editions are often required. The Availability Groups or FCI approaches are also complex - everything needs to be the same across the environment, and this can cause additional downtime if you have to upgrade or update, or if you use software or hardware from different vendors.

VMblog:  Are there any newer approaches companies can utilize for disaster recovery?

Cox:  For heavy duty applications like SQL server, as well as containerized applications, an all-inclusive software solution is now possible. It starts with a Vhost - essentially an IP name and address - which abstracts and encapsulates instances, providing a consistent connection string.

Note that this entails built-in high availability, as well as simplified disaster recovery: it encapsulates instances and allows you to move them from host to host locally, or between sites for disaster recovery. This relies on a means of replicating data from one site to another, while managing the failover component of rehosting the instances themselves to the other site. This gives the user many choices around replication, including selecting the most common array replication, and vSAN technology or Storage Replica.

VMblog:  How does this differ from the traditional approaches to disaster recovery?

Cox:  First, it's more flexible. Since it offers high availability with disaster recovery built in, it's a completely new approach compared to the traditional disaster recovery approaches. It's infrastructure-agnostic, and can work on bare metal, virtual machines, or any combination in between. It can run on the cloud, so can be useful if you have a cloud-based workload you want to protect. It's simple to layer this into existing cloud frameworks and is not restricted to being "like for like." You can use different versions of SQL server from 2005 up, Docker containers, Windows server back to 2008R2, and even SQL Server on Linux.

Since you can mix and match, updates require less system downtime.

This also breaks the cost and complexity trade-off because it enables significant consolidation. It allows anywhere between 5 to 15 SQL Server instances per server, with no additional licenses needed, significantly reducing costs associated with management and licensing - savings are 25-60% on average.

Finally, there is no restriction on the edition of SQL Server you have to use for this type of clustering, so there is no need to use the much more expensive Enterprise version. If you've already purchased the licenses, you can save them for future use.

VMblog:  I see there are significant benefits to this new software-based approach to disaster response.  Could you tell us a bit about what it looks like in practice?

Cox:  Here's an example: you can install this software tool on two servers, add a SQL Server instance under management, then fail that instance over for local HA. You can then add a third node in a different subnet from the first two nodes, then move that instance over to the other site - either manually, or in the event of an unplanned fault.

This leverages standalone instances, and decouples workloads, file shares, services, and Docker containers from the infrastructure. It requires no standardization of the entire database environment on one edition of the OS and database, allowing complete mobility from host to host. It allows for close-to-zero planned and unplanned downtime, simplifies management, enables peak utilization and consolidation, and reduces costs significantly.

VMblog:  Is there anything else that comes to mind that you'd like to share?

Cox:  The key point is that, unlike traditional approaches to DR for SQL Server, this software-based solution allows for total optimization of the environment. It unifies HA/DR across SQL Server, Windows, Linux, and Docker, and gives an all-inclusive approach for management and optimization of your environment. This results in a much lower cost, less stress for your IT team, and ultimately a more consistent and failsafe system.

##

About Connor Cox, Director of Business Development, DH2i

Connor Cox is a technical business development executive with extensive experience assisting customers transform their IT capabilities to maximize business value. As an enterprise IT strategist, Connor helps organizations achieve the highest overall IT service availability, improve agility, and minimize TCO. He has worked in the enterprise tech startup field for the past 5 years. Connor earned a Bachelor of Science in Business Administration from Colorado State University and was recently named a 2017 CRN Channel Chief.