In late 2017 a client of BTS’ located in Seattle, WA was faced with several circumstances which required resolution by year end. BTS was asked to develop a strategy and solution set to be implemented in one overarching project.
Firstly, the company had database servers that were hosted by a provider that was no longer offering those services. We needed to, at minimum, replicate the existing environment and get the databases out of the colocated data center in only 30 days while working around their heavy workload. Additionally, the client requested a higher level of availability for these databases as they are the backbone of their analytics reporting business. The hope to publish the reports to a publicly accessible and secure location, rather than manually emailing them, was discussed as a future consideration.
Secondly, their board was requiring a business continuity plan be developed and implemented. While the first requirement had a time component, this second requirement was the priority for the executive team.
The foundation of the final solution was based on linking, via a VPN tunnel, the on-premise infrastructure with an Azure virtual machine environment that included domain controllers, a SQL cluster that could failover from on-premise to the cloud, and other critical application servers. This allowed for the near real-time replication of Active Directory, server-stored code and scripts, database maintenance plan backups, and other critical data such as the company financial system’s files. Essentially, the two environments were configured to replicate bi-directionally, keeping a synched copy in both locations. The client already had a presence in the Microsoft ecosystem, with its utilization of Office 365 and the use of SQL Server, making Azure their preferred cloud provider.
This solution not only created an elegant structure for business continuity, but provided 99.999% availability for all mission-critical business systems. Additionally, the cloud infrastructure positioned the client to readily implement a publicly accessible reporting solution when the time comes.
The solution consisted of the following components:
- A virtual network and “route-based” virtual gateway which allows both site-to-site and point-to-site VPN connections
- Virtual Machines: domain controllers, SQL Servers, file servers, etc.
- An AlwaysOn SQL Cluster
- A load balancer to forward requests to the SQL Cluster
- A Recovery Services Vault to store backups of non-critical on-premise servers to be restored during a disaster
- An AlwaysOn SQL Cluster built on an existing Hyper-V server infrastructure
- A System Center Virtual Machine Manager implementation to mange the virtual machines and enable backups to the Recovery Services Vault
- BTS identified a disaster recovery office space vendor with the ability to support the client’s communication and data systems requirements
- BTS developed the client’s processes to include:
- Emergency Response Procedures
- Backup Operations Procedures
- Recovery Actions Procedures
- Disaster Action Checklist
- Emergency Communications Plan