The challenge
A hospital warehouse that had to be built right the first time
The client is a Canadian hospital that needed a SQL Server data warehouse. Data platforms in healthcare don't get many second chances, so high availability had to be designed in from the start, not added later.
That meant the plan came before the servers. The project opened with a kickoff, a SQL lab, scripts, and a design document that was revised, then revised again after a review meeting. DNS and directory checks happened in the same early pass, so the new platform would have a clean place to land.
Our approach
BITSUMMIT built the warehouse in a deliberate order: design and lab first, then the servers, then high availability, then maintenance.
The build
Windows Server was set up on two servers, and SQL Server was installed on them with high availability as part of the same install. The team then validated and tuned the high-availability configuration in follow-up sessions.
Failover readiness and handover
Failover testing was set up, and the team held a SQL Server failover and technical walkthrough session. Maintenance was set up on both servers, and a spreadsheet report was produced for the client. The project then moved through its validation and handover phase and was closed out on a call with the client.
The outcome
- A SQL Server warehouse with high availability in place. Built on two servers, with high availability designed into the install rather than bolted on.
- Failover readiness built in. Failover testing was set up, and a SQL Server failover and technical walkthrough session was held before the closeout call.
- Maintenance set up before handover. Both servers had maintenance in place when the project was handed over.
- A clean close. The project finished with a validation and handover phase and a closeout call.
A data warehouse that exists only in a design document is still a plan. This one was designed, installed, set up for maintenance, and closed out.
For how BITSUMMIT approaches the platform around the data, see data and analytics and hybrid infrastructure.
Talk with a senior engineer about the warehouse, the servers, and how they should be run. Book a data platform review.
