Success Story – Global Scale Business Continuity
Client needed to improve the availability for the databases underneath a globally distributed physical security platform for thousands of world-wide locations and used by hundreds of thousands of employees around the clock.
Heraflux reviewed the virtualized SQL Server architecture proposed for this application server upgrade, and implemented SQL Server Availability Groups and log shipping to provide a platform for local high availability with a globally replicated disaster recovery architecture for each of the five centralized datacenters around the globe.
Client achieved a new level of availability in the data platform underneath the security system, the ability to implement routine patches to the OS and SQL Server layers without disrupting the business, and a disaster recovery strategy that exceeded the expected RPO and RTO for the data layer.
This platform allows me to sleep well at night knowing that our high availability solution provides insulation to routine operational requirements, and better protects us from system problems and natural disasters.
Client is a leading provider of consumer technology hardware devices and a platform for software development. The Client was using a two-tier third-party application that handles perimeter and in-building security for objects such as badge scanners, security alarms, and entry points. The SQL Server back-end was leveraging stand-alone SQL Server instances at each primary datacenter, and did not provide for means to perform routine patching and other operational maintenance tasks without taking down the security platform, creating an outage for the critical security system.
The Client was about to complete a software upgrade of the security platform, and wished to leverage improved availability features of SQL Server to improve the high availability of the SQL Server layer to improve the uptime of this security system.
Heraflux discussed the advantaged of SQL Server Availability Groups with the Client, and after a short POC to validate the reference architecture, the decision was made to deploy Availability Groups in each of the client datacenters worldwide to employ this feature as part of the upgrade.
The Client needed an outside specialist to review the architecture and inner workings of the application platform and the way it used the SQL Server layer, with the application server, database server, virtualized platform, and networking requirements all to be scrutinized. A working proof-of-concept would then be constructed of a SQL Server Availability Group, and based on the testing results, construct a reference architecture for SQL Server Availability Groups that could be implemented globally for each datacenter in the Client network as a base platform for the security system as part of this upgrade.
Heraflux performed a deep-dive systems review of the VMware-based virtualization systems platform powering the database and application servers to determine the overall configuration and performance characteristics of the on-prem hosting platform and the current SQL Servers.
Heraflux then provisioned a number of SQL Server 2016 database servers inside virtual machines at their primary location, and performed application load testing with a test copy of the application. The proof-of-concept was then approved. To determine the sizing of the new destination servers, performance monitoring was put into place on each of the current SQL Servers in the client network, and an analysis showed that the current SQL Server VMs’ configuration of eight virtual CPUs could be cut in half with no performance penalty after the upgrade, saving the client over $200,000 in SQL Server licensing.
The destination SQL Server VMs were mapped out and deployed in each of the client datacenters ahead of the migration. Availability Groups were established in each datacenter. Log shipping on each Availability Group was used to perform one-minute data synchronization of the critical data streams to at least two other global data centers to improve the business continuity strategy.
The application upgrade process and database migration were rather uneventful, due to the planning and diligence up-front in the project, and afterwards, the platform worked as advertised to provide improved high availability and disaster recovery for this critical system.
The new Availability Group platform allows the Client to perform operational maintenance on the SQL Server databases without fear of disrupting the business.
The “right-sizing” of the SQL Server VMs also allowed the Client to save over $200,000 in Microsoft SQL Server licensing that would have been unnecessarily spent.
The improved RTO and RPO of the SQL Server business continuity plan with log shipping replicated globally helped the Client exceed their stated internal SLAs.