There are many reasons for migrating your enterprise data. Whether it’s to consolidate physical data centers following a series of mergers or acquisitions, take advantage of new applications or efficiencies such as cloud capabilities, or leverage greater capacity for strategic growth, the relocation of your company’s most precious asset needs careful thought.
With so many moving parts and interdependencies, planning a data center migration can take months. To be successful and minimize business interruption, the plan needs to include:
The right information at the right time
Get to know your current physical and virtual asset requirements and future needs.
An extensive checklist: Including multiple plans, requirements applications and event dependencies, workflow steps, blackout dates and other relevant time constraints such as lease expirations, task owners, and the tools you’ll use.
Workflow execution and tracking: Showing how workflow changes and delays impact the step (or most likely) multiple steps down the line.
Analytics and reporting: It’s vital to keep an eye on progress and let stakeholders know what’s going on.
Step 1
Compile an inventory of your physical and virtual assets and current requirements, running reports and pulling information from tools, databases, and spreadsheets. Talk to key stakeholders and get to know your company’s governance and security policies (if you don’t know them already).
Understand future requirements to right-size your investment: Why are you moving? Use that as a starting point and factor in future growth demands, by talking to business leaders, before working out what hardware you need to replace.
Compile and clean up the information you have (there will be duplications) to define the data center model you should adopt.
Step 2
Identify the locations or providers you want to move to using your current and future requirements: think about where you place your workloads, what locations you need to be in and which CSPs align most closely with your needs.
Analyze dependencies across all your components to develop a wave rollout plan using:
Create sub-plans to your project rollout plan including:
Get final sign-off from your business leads and tweak plans where necessary.
Run through lab simulations and execution scenarios with teams where you can.
Step 3
At last! You can get your move underway. At this stage you’ll need to:
Make sure everyone knows their part in the rollout and that they are available. Check your equipment is where it needs to be and that you know when data needs to be backed up.
Be flexible: The best laid plans, etc. Make sure you can quickly implement mitigation plans if any issues occur.
Stay in contact with your project team and wider stakeholders to make sure every task is checked off and chase progress where required.
Create reports to communicate the project outcome with business leaders and other stakeholders.
Step 4
Once the data center migration is complete you should monitor and optimize performance across your physical and virtual environment as an ongoing program to maintain costs and end user satisfaction. There are many factors that can lead to data center outages and they can be expensive, costing on average US$9,000 per minute. Any data breach can have far-reaching effects and changes in performance can impact critical users or business applications. To mitigate risks, focus on:
Your physical environment and status of any uninterrupted power supply (UPS) batteries to ensure changes don’t affect hardware.
Application usage and network traffic to identify anomalies that could signify security issues. While CSPs have stringent security in place you should be aware of any fluctuations that could signify a breach.
Weather patterns and environmental issues in data center locations.
Your hardware status across its lifecycle.
