Metrics to prove or disprove the need for additional DBA’s?

Posted on

Question :

We currently have 2 DBA’s, me and one other guy and we both feel the need for an additional resource (or two). I have had the conversation with my direct manager on this topic a couple different times, but am having difficulty in getting the notion sold. Most recently our manager shared with us that it is really good to come up with our “soft values” list of things that we want or feel we need to be doing, but what will really help is to come up with “hard values”, showing dollar savings.

Could you help me in understanding first of all if our perceived need is legitimate, and if so, how do we go about coming up with this “hard values” list. I feel like we could come up with those things that we need to do in order to save the company money, but that could just get put at the front of our priorities list, and that much more stress on us in the process.

To give you a high level look at our environment, here are the things that I am looking at in terms of what we are providing support for.

Production servers – 30+
Production Databases – 200+
SQL Versions – 2008/2008R2/2012, looking into 2014 now, and 2016 later in the year
Applications(3rd party and home grown) – 20+
Application Teams Supported – 6
Virtualization – 75% virtualized 25% physical
Clusters – 3+ more planned
Replication – 1 distribution server, 2 subscription servers, 24 subscriptions, 6 publication servers, 12 publications
Log shipping – 8 primaries +more planned, 4 secondaries + more planned, 49 log shipped databases + more planned
Availability Groups – None at the moment, but exploring the possibility
Average application upgrade/installs per year that drive version change or DBA resources – 2-3
Patching (SP, CU’s) – Non-existent at the moment unless an issue comes up
Cumulative database size – 14TB+
Reporting Servers – One scale-out deployment consisting of 2 servers, neither of us are well versed in SSRS
Analysis Server – Two servers, neither of us are well versed in SSAS

What metrics should we use to prove or disprove the need for additional DBA’s?

My initial intention was demonstrating our current capacity vs. our planned projects and repeating tasks, but we need to demonstrate the ROI on hiring additional resources.

Answer :

Demonstrating ROI on hiring additional DBAs is hard since DBAs don’t generate revenue by themselves. They do however allow the organization to function and generate revenue, just like any other supporting staff. A lack of staff in any department makes the organization function less optimal and lose money or lose opportunity. So there is a cost from lack of investment rather than a return on investment.

If you feel you need additional DBAs you must know what tasks you are unable to perform because of lack of time/skill, if it’s lack of skill (such as the SSAS stuff you mentioned) that can be trained, if it’s lack of time it could be temporary or structural. I’ve been in situations where you just know that your life would be a lot easier if you could just automate this and that or set up some alerting but just never get around to because of too much reactive or repetitive work and general fire fighting.

Once you have determined why your team isn’t functioning as well as it should it’s time to document it.

First of all, document what you do now, why that takes up all your time and make clear up front to your manager the existing problems that need solving are absolutely none of your team’s fault. Make sure you don’t give the impression that the problem would be solved by hiring ‘better’ DBAs.
Make an excel sheet showing your workload for the coming weeks if you can, charts impress, and it can show your workload in a single slide.

Once you are clear on that, document all additional tasks and everything that needs improving, but especially document what the business risks and possible costs are, and the savings could be. Also document how much man-hours you think would be needed to make those improvements and perform those tasks.

Things that go on the investment side of the equation are

  • Salary of the newly hired DBA
  • Price of training

Things that go on the savings side of the equation are

  • Projects not finished on time
  • Preventing downtime that makes the business lose money
  • Preventing legal risk
  • Preventing data loss
  • Consulting fees could be reduced
  • Cost of burning out employees because of too much work

Once you have compiled the list as complete as you can it should be clear what the best price/reward option is

  • Getting additional training
  • Getting additional DBAs
  • Hiring a temporary consultant to assist in automating stuff, monitoring stuff, performing a migration so that the existing team has more time
  • Changing the priorities put on you by the business so you get time to get things in place that will allow you to work more efficiently later on

Leave a Reply

Your email address will not be published.