Home > Articles > Data > SQL Server

Making Your SQL Server Apps Highly Available: First, Do The Assessment

  • Print
  • + Share This
Want your SQL Server apps to stand up to industrial use? In this sample book chapter, you'll learn a rather formal approach to assessing and choosing a high availability solution. As you'll learn, the key to success is doing as complete a job as you can on the assessment and using some of your best folks to do it. They will interpret the technology and the business needs with the most accuracy.
This chapter is from the book

The key topics in this chapter are

  • Launching a Phase 0 (zero) High Availability assessment
  • A hybrid high availability selection method
  • Scenario 1: Application service provider (ASP) assessment
  • Scenario 2: Worldwide sales and marketing (brand promotion) assessment
  • Scenario 3: Investment portfolio management assessment
  • Scenario 4: Call before you dig assessment
  • Cost justification of a selected high availability solution—ROI
  • High availability considerations embedded in your development methodology

Moving Toward High Availability

Up to this point, we have described most of the essential elements that need to be defined for you to properly assess your application's likeliness of being built utilizing a high availability solution. Formally conducting the Phase 0 (zero) High Availability assessment (and within this, completing the High Availability Primary Variables Gauge) is the primary tool to start this effort with and is geared toward garnering results very quickly.

The mere fact that you are considering a Phase 0 HA assessment for an application points to the likeliness of some type of high availability solution. But, which one? And, which one is the BEST one?

In order to make the best possible decision of which high availability solution matches your business requirements, a simple four-step process can be followed. The steps are

  • Step one calls for a launch of a brief Phase 0 HA assessment effort to gather all the needed information as quickly and as accurately as possible. Or, perhaps not so brief if you can drill down a bit further in each requirement area.

  • Step two will require that the High Availability Primary Variables Gauge be done as completely and accurately as possible. This gauge is actually a deliverable of the Phase 0 HA assessment, but is more than worth calling out individually as a separate step since it can be used as a high-level depiction of your application's HA requirements and is easily understood by management-level folks in the organization.

  • Step three will use the assessment and the gauge information to determine the optimal high availability solution that technically matches your business needs. A hybrid decision-tree selection method will be described to help in this step.

  • As an added bonus, we will run through a basic Return on Investment (ROI) calculation as an optional Step four of this high availability decision process. The ROI calculation is optional because most folks don't bother with it—they are already losing so much money and goodwill during downtime that the return on their investment can be overwhelming. Very often, the ROI cannot be clearly measured and no financial impact can be calculated.

This four-step process is, in effect, a mini-methodology designed to yield a specific high availability answer that best matches your needs.

Design Note

ROI Calculation: We will try to describe a fairly simple and straightforward method of calculating the ROI when deploying a specific high availability solution. Your calculations will vary because ROI is extremely unique for any company. However, in general ROI can be calculated by adding up the incremental costs of the new HA solution and comparing them against the complete cost of downtime for a period of time (I suggest this be a one year time period). This ROI calculation will include

Maintenance Cost (for a one year period):

+ system admin personnel cost (additional time for training of these personnel)

+ software licensing cost (of additional HA components)

Hardware Cost (add +)

+ hardware cost (of additional HW in the new HA solution)

Deployment/Assessment Cost:

+ deployment cost (develop, test, QA, production implementation of the solution)

+ HA assessment cost (be bold and go ahead and throw the cost of the assessment into this to be a complete ROI calculation)

Downtime Cost (for a one year period):

If you kept track of last year's downtime record, use this number; otherwise produce an estimate of planned and unplanned downtime for this calculation.

+ Planned downtime hours x cost of hourly downtime to the company (revenue loss/productivity loss/goodwill loss [optional])

+ Unplanned downtime hours x cost of hourly downtime to the company (revenue loss/productivity loss/goodwill loss [optional])

If the HA costs (above) are more than the downtime costs for one year, then extend it out another year, and then another until you can determine how long it will take to get the ROI.

In reality, most companies will have achieved the ROI within 6 to 9 months of the first year.

  • + Share This
  • 🔖 Save To Your Account