Advice for data centers looking to change operating systems
When you swap out a data center's primary OS, first, consider the purpose of your replacement OS and its optimal workloads. Then, you can create an OS migration plan.
Data centers switch primary OSes for a variety of reasons, including improving scalability and elasticity or modernizing legacy infrastructure. However, leading a data center OS migration can present a complex challenge for data center administrators, with many complicated processes and strategies.
Plan a data center migration to a new OS with a solid direction. Be proactive, recognize the obstacles you might face and plan a way around them in advance.
The OS in the data center
Which OS a data center uses can vary, but the majority of platforms are based on or have compatibility with Linux. Linux is well known for its incredible flexibility and versatility, largely thanks to its open source nature and highly active global community. Because anyone can use Linux freely, developers around the world have built custom configurations suited to almost any purpose.
Linux's modular nature also makes it a natural fit for the cloud -- easy to scale and match the pace of potentially rapid data center growth. Some of the biggest cloud platforms in the world are based on hardened versions of Linux, including AWS, Google Cloud Platform and Microsoft Azure.
Many of today's existing data center OSes are compatible with Linux, but each OS often has a specific purpose. For example, Kubernetes provides a way to configure Docker containers into clusters of interacting services. It automatically accounts for resource density replication and service grouping and intelligently schedules these factors. Photon, on the other hand, operates as a minimal Linux container host with a focus on quick booting on VMware platforms.
When considering whether to switch a data center's primary OS, first, research open source OS options. This can provide an idea of what platforms exist, which ones are popular and what specific functions they serve.
The OS migration
Certain OSes handle scaling better than others, so over time, a data center might require a system that's better suited to performance and agility or flexibility. As complexity grows, consider an OS that can consolidate controls for easier management.
When migrating to a new primary OS, design a thought-out OS migration plan, including the following steps:
- Take a detailed inventory. A newly updated, comprehensive audit of the data center helps an organization understand everything connected to and reliant on the OS. Since data centers must remain up and running 24/7, the goal of an audit and plan should be to minimize disruption as much as possible. Make it a point to list out all hardware mapping, software applications, storage layers and network configurations. This can help prioritize activating workflows after the migration.
- Develop a project management plan. Have a plan in place that explicitly outlines what actions you must take and when. This helps avoid disruption and provides clear next steps at every stage of the migration. Tackle the simpler aspects of the migration upfront, and break down the complexity into easy-to-follow processes.
- Begin the execution phase. As you transition from planning to execution, confirm you have the correct version of the new OS, complete with the latest upgrades and patches to maintain security. Even one mistake during this phase can open up vulnerabilities for hackers who seek an easy way in, so take appropriate time to configure everything appropriately. Test your applications repeatedly, checking access across servers and databases. Execute in grouped phases to minimize disruption, which you should designate in your planning phase.
- Tweak and optimize. Once everything has been tested and the new OS goes live, still plan to conduct periodic reviews of the OS and how it supports the applications running on it. Optimize management controls through careful tweaking, and regularly evaluate performance and configuration.