IT Tips & Tricks
Published 1 July 2016
Updated 13 April 2022
5 Common (and Dangerous) Data Migration Mistakes
Do you have a data migration looming in your future?
Check out this list of the five common data migration mistakes. Avoid them and you could save yourself a lot of time and frustration — and your company lots of money, which just might lead to them giving you some of that saved dough in the form of a raise. (Just sayin’.)
Mistake #1: Trusting That Your Backup Will Be There for You
“I have backups! I have so many backups! We always back-up before we change anything around here!”
We don’t doubt that you back up your data. The problem is that many (too many) IT professionals trust back-ups that turn out to be unreliable or problematic — and they didn’t realize it until it was too late. Cue the hypothetical thought bubble: Do you test those backups?
When was the last time you fully tested your backup system? When was the last time you did a test deployment of your back-up in exactly the same manner as you would have to do it, should you actually lose some data?
Too many IT professionals trust back-ups that turn out to be unreliable or problematic — and they didn’t realize it until it was too late.
Here’s where the mistake hides, like many, in plain sight. Many times, during the crunch of time of a data migration, people are quick to back up, and keep moving. Many times, people have to refer back to those backups, at least to recover some part of their data. Many times the recovery mission fails, and only then do they realize that something was wrong with either the back-up itself or with the recovery system.
You’re not going to get very far if you don’t know what your destination is or how to get there. A data migration is no different.
Make a copy of your data before migrating it. You want to be able to see everything as it was originally, on its original system, so that you can spot any corruption or anything that looks out of place.
Make sure that you change as little as possible between the “back-up” phase, and the actual migration phase, so that you don’t lose any changes that you might make before the migration.
Oh, yeah, and test your back-up. Run an actual recovery and see what happens.
Mistake #2: Getting Started Without a Solid Strategy
Planning is everything in data migration. It will help you prepare for possible issues along the way, and it can save you boatloads of time and money. Without a solid strategy or outline of the steps you’ll be taking, the migration process could be prone to data losses, security issues, and more.
So what are the steps in a data migration? They can vary slightly depending on your individual needs, but the following are the most basic steps involved in the task. To conduct a successful migration, you need to:
- Develop a detailed strategy
- Analyze the data
- Profile and organize the data
- Test the data for quality and accuracy
- Execute migration
By following these steps, you’ll be able to account for every piece of data and its validity before initiating the move. You’re free to make adjustments, operate within this framework, and enlist data migration tools to meet your goals. Just remember to start with a full strategy — the glue that will hold everything together. Even if you’re tempted to let some of the details slide, take our advice and outline them anyway. You’ll be glad you did later.
Mistake #3: Leaving Home Without a Roadmap
Look, you wouldn’t leave your house and try to fumble your way to some place you have never ever been before without your GPS, right?
Okay, so why would you start a migration without a detailed map of all your data, and plans for where it ends up?
- You need to know what type of data you are migrating: VMDK (virtual machine disk file)/VHDx (Hyper-V virtual hard disk), databases, storage arrays (data storage systems for block-based storage, file-based storage, or object storage), files or other?
- What are the dependencies and relationships to other systems, workflows, and operations?
- Can I use replication tools to make data available in multiple places, at least for a transitional period?
- Verify all network and application policies are set to point users to the correct data locations.
Many times, the recovery mission fails, and only then do they realize that something was wrong with either the back-up itself or with the recovery system.
These are just a few items that may be addressed with your Roadmap. There are many possible details depending on your situation and needs.
This is information that you probably gathered in the planning and strategy stage of your migration. However, it is worth reviewing and ensuring you have identified and planned for all aspects of the migration, including contingencies for when things do not go as expected. The roadmap serves to keep you focused and allows you to track the migration progress.
Knowing where your data is, what its purpose is, and what type of migration you’re doing will make the whole process much easier. It will leave less room for error too.
Overall, find the tool that works best for you to generate a comprehensive view of your files, including all the links contained within them. (Hint: https://LinkTek.com/Free-LinkReporter/)
Mistake #4: Lack of Hardware, Software, and/or Network Preparation
“Huh? We have a lot of this stuff (particularly hardware) around here!”
Yes, but you are doing a data migration now. The rules change a bit. Consider the following.
Firewalls, busy networks and bandwidth can all impact data flow. Keep the pipeline open for a smooth migration.
Hardware:
Make sure that you have the allotted resources for your migration. When you’re doing a migration, you’re usually moving from an outdated or smaller resource to a larger one. Nonetheless, a good “rule of thumb” is to have at least 20% more capacity than what you think you will need.
Secondly, be sure that the hardware you’re migrating to has been tested and passed a benchmark test. Make sure that you’ve thoroughly tested it before the migration, so that you don’t finish up your migration, pat yourself on the back, and then have something break five minutes later.
Software:
During the migration itself, be doubly sure that you’ve shut down any processes or applications on the software you’re using. Having other processes running in the background while you’re trying to do a migration can cause interruptions or complications in your migration.
Network:
Last, but not least, you want to be absolutely sure you’re working with an open pipe. Make sure your firewall settings aren’t choking you from doing what you need to but aren’t leaving you vulnerable either.
Also, if you’re going to have users on your network while you’re doing your migration, you want to be sure to throttle your own use, so that you’re not slowing down your users.
If you know you’re going to have users on at the time of your migration, plan for it. You can’t kick them out and lock the door (tempting as that may be), so you’ll have to be sure you’re allocating the proper amounts of bandwidth for both you and them, so as to not slow down any work for either of you.
Mistake #5: Not Having a Rollback Plan
Developing a rollback plan is one of the best data migration techniques and can help you save time and energy after the move. You may run into a few challenges while relocating data sets. Right now, you might be thinking you can simply go back and fix everything later. While that’s usually true, it can be a messy and inefficient way to work.
That’s why you need a rollback plan. Rather than waiting until after you’ve moved all the data to address issues, you’ll develop backups and a rollback plan to apply fixes along the way. For instance, one way to implement a rollback plan is to set checkpoints at specific stages of the migration so that you can run assessments on the data. If any transitions need to be rolled back, you can make the backups and deployments necessary to move forward.
A good “rule of thumb” is to have at least 20% more capacity than what you think you will need.
Not having a rollback plan will add extra work to your schedule and cause otherwise preventable disruptions after deployment. By adding one to your checklist, you can make sure you’re equipped to address issues right away and create a better final result.
Summary
Some these points might seem obvious, in hindsight. Yet these are the things that have bitten many an IT pro in the you-know-what. You should pay extra attention to these areas when planning a data migration, no matter how big or small. Making sure you plan for all these things will keep you vigilant, and help make the other aspects of your migration run smoothly (“smoothly” being a relative term).
If you’re interested in learning more about data migration tools and techniques that can help you move in the right direction, contact us today.
Feel free to share this article on your social media: