Once-manageable servers have been steadily taking over premises, occupying useful space and racking up ever-bigger energy bills. So, what happens when you get to this point? Colocation, or moving into a data centre, might be the right way to go.
>See also: Barriers to public cloud adoption are fading, but server peril remains
Large colocation data centres are much more efficient at delivering power to servers and cooling them thanks to their economies of scale. Because they buy power at wholesale rates, they can pass these savings on too. They also eliminate all the costs of maintaining internal server room equipment such as UPSs, generators and air con units, as this is included in the price.
Finally, factory and office space, especially in London, is at a premium – by moving into a data centre you can free up that space for more productive activities or desk space.
Getting colocation right
Colocation is not only a smart decision, but it’s also a critical one. You may, and quite rightly, have concerns around downtime, security and application performance, as well as the nuts and bolts of what the process actually entails. For migration to go as smoothly and safely as possible, there are several factors that need to be considered.
You really need to do your background and research. Kickstarting your data centre migration blindfolded is a big no-no. Take the time to consider how relocating your critical apps, services, and data will affect your business during the migration process, and what you can do to mitigate any risks or temporary disadvantages.
Server downtime is a key consideration. How you approach this depends on the nature of your business. If it simply can’t tolerate any server downtime, you need to safeguard operations with a robust disaster recovery and backup initiative. You could also setup a temporary private or hybrid cloud to keep critical processes running during the migration.
>See also: Top tips for migrating to the cloud
Likewise, if your system-critical apps are moving, consider running a trial migration to ensure ongoing software compatibility (and reducing chances of further downtime). A good data centre provider will help you with this and ensure the process goes to plan.
Network configuration is also a factor to take on board. Organisations must decide what needs to be done to ensure that older apps will retain their functionality in cases of incompatibility.
Decisions will need to be made on a case by case basis, as some apps may or may not encounter configuration issues in the move from your local area network. It’s better to be safe than sorry, so be sure to investigate the effect migration may have on your mission-critical apps.
This might not be one that comes to mind immediately, but network latency is important, too. Colocation means accessing your data centre through a dedicated, high-speed connection. Latency (time lag on the network) shouldn’t be an issue after the migration, but it’s important to consider instances where it may occur during migration.
As servers are often migrated in batches, apps sharing a local area connection must now work harder to communicate. To mitigate potential latency issues, identify which apps work together and when, and plan your migration timeline around keeping them apart as briefly as possible.
>See also: Cloud adoption: A CIO’s case of how it will happen in a business
A successful data centre migration means knowing everything about your apps inside out. But what about those ancient apps which have been chugging along happily for years? The documentation is nowhere to be found, and nobody remembers who installed or built them to begin with. By using network tracing tools in the months running up to your migration, you can relearn everything you need to know about the intricacies of your legacy apps.
If businesses carry out the necessary preparation for migration, the actual process should be pretty straightforward, and migration should be seamless. When it’s time to move the data, you need to decide what definitely needs to move. You need to see whether, for example, you’re still operating hardware and software through contracts long overdue for termination or, for instance, still using old equipment which no longer serves a critical purpose. Importantly, you need to ask yourself if every server needs to go over, or could you virtualise some to share space, and rationalise down the number of physical boxes?
Go through everything, and consider its purpose and role in your business going forward. You may even find some equipment and contracts are more vital than expected, so it pays to be thorough. This is also an ideal time to revisit your migration timeline, and consider whether you need to setup a temporary private or hybrid cloud to avoid downtime during migration.
Furthermore, it pays to have an up-to-date record of your entire data environment – review your existing logs, and make note of any changes to the inventory. Next, identify existing workloads, software, and scheduled backups so you know exactly what will and won’t operate during migration – and for the ultimate peace of mind, run that all-important disaster recovery test.
>See also: 4 considerations for Windows optimisation and migration
You’ll also want to inform your service contractors of your plans, and point them towards your new data centre for any licensing and contractual amends. And don’t forget to note your equipment warranty information and serial numbers to avoid any nasty shocks following physical relocation.
With your equipment safely managed and maintained 24/7/365, your business is free to get back to what it does best. And should you continue to grow, scaling your data centre is easier than ever, eliminating the need for the temporary and costly solutions of old. As for those redundant processes and apps you uncovered during planning – it’s time to put those freed resources to good use and make way for a greener, more efficient future.
Click here for more advice on data centre migration
Sourced by Jack Bedell-Pearce, managing director, 4D Data Centres
