As they say, time is money. The sooner your VMware Cloud™ environment is operational, the sooner you can realize value from it. Here are the three key steps to get your environment up and running and ready for consumption.
Step 1: Create Your Target SDDC and Connect It with Your Existing Environment
Creating your first SDDC is an easy step overall, but doing a proper design of its properties and correctly configuring security aspects are crucial activities needed to avoid rework or reconfigurations. This is especially true for business-critical scenarios. Based on the number of hosts, storage requirements, and SLAs you want to determine:
- The type of host to use with specific CPU, memory, and storage sizing
- The number of clusters you need and if they will be stretched within the same region or split across different regions
- The range of IPs (CIDR) to be used for management
- The type of connectivity required from on-premises to the cloud
- The backup and restore strategy
- The disaster recovery strategy
You need to design and implement network connectivity. Having a dedicated network connection that provides high bandwidth and low latency is recommended, but if you need a cost-effective solution you can use route-based or policy-based VPNs.
Deploy a mobility platform that will allow you to move workloads on-demand between your environments. To minimize the changes on the network, use a solution that can quickly extend network segments (L2) and optimize bandwidth consumption during workload migrations.
You’ll also need to design security with distributed firewalls. Moving to the cloud doesn’t mean just moving workloads. You need to look at the network and security components as well. Existing physical firewalls cannot be moved into the cloud. A common strategy is to leverage the corresponding virtual appliance version of those firewalls. An alternative strategy is to embrace the native capabilities of VMware Cloud to build security and group policies, firewall rules, and implement micro-segmentations.
Step 2: Migrate Workloads
Successful cloud migration starts with a migration strategy. Your approach should be based on your business goals and priorities. Understand why you need to do the migration and what you want to achieve. Identify any deadlines or time constraints, cost constraints, and application constraints. Then develop your high-level migration plan with specific execution tasks that identifies who will do what and when.
A key aspect of your migration strategy is building the migration team. Select champions in critical functions including network and security, operations, infrastructure, and applications to drive the migration.
One of the critical activities you need to perform is discovering and analyzing the environments. Gather information from stakeholders on the applications and workloads in scope. This includes collecting information through interviews and/or with third-party tools on physical workloads, virtual workloads, applications, storage, and core infrastructure services. Dedicate time to verify the information is up to date and discover any gaps. Analyze the capacity, performance, network connectivity/throughput, and packet flow between applications in the current environment. Assess any business continuity/disaster recovery, compliance, and security requirements needed for the target environment.
As you migrate, remember: less is better. Avoid making too many big changes during the migration activities. For example, if it’s not strictly required (e.g., networking overlap) don’t change IP addresses of workloads to be migrated. That will simplify the identification of conflicts and drastically reduce potential unknown issues.
Using this information, map application dependencies for each individual application in scope. Then build a map of application dependencies between applications. This will allow you to avoid potential conflicts such as sequence, bandwidth, or data security concerns. For example, if there is a group of strictly related VMs (e.g., apps and their databases) put them in a single group and migrate them together.
From here, create a detailed migration plan by mapping the source infrastructure (server, network, and storage) to the target infrastructure. Prepare and configure in advance external and core services at the target, such as load balancers, active directory, file servers, public IP addresses, and DNS names. Create a testing schedule and contingency plans including a rollback strategy if the migration event execution encounters any problems.
Bundle, schedule, and execute the migrations in waves. This will make the overall migration more manageable. Create a runbook that describes the source VM, defines properties for the target VM and networks, and provides the procedures used to carry out the migration.
Workload migration can be a daunting process. Start small with a few simple applications to define the process. Continue to migrate small groups of workloads to refine the process. As you go, keep track of the lessons learned in a centralized repository, and eventually, you’ll get to migrating multiple parallel streams of complex workloads.
Step 3: Implement Operations and Management
Effectively operating and managing your new VMware Cloud environment may require updating existing operational processes, roles, and skills. For example, you may need to focus less on how to upgrade the infrastructure and more on monitoring applications, extending existing automation to your new cloud SDDC, and keeping budget and costs under control.
To have a clear view of how your workloads are performing in your new cloud SDDC, define and implement metrics and monitoring capabilities even before moving them to the cloud. Implement dashboards to quickly see changes in real-time. With this insight, you will be able to troubleshoot and remediate issues before they grow in severity or breadth.
Consider processes such as event management to detect and investigate exception conditions before they lead to service outages, and capacity and performance management to help you measure and analyze resource usage and determine production capacity as demand changes. Create policies, standard operating procedures, and work instructions in collaboration with your operations team.
You may need to assign new roles to your staff, like Organization Owners, Cloud Administrator, NSX Administrator, and NSX Cloud Auditor. Determine the responsibilities and skills required for each role. Map your current staff to those new roles, identify skills gaps, and create individual training plans to ensure that everyone has the appropriate knowledge and insights to fulfill their new responsibilities.
The Bottom Line
VMware Cloud is often the preferred choice as it enables consistent infrastructure and operations across on-premises and new cloud environments. The faster you get your environment productive the faster you can drive real, measurable value. To get there, you need to create your SDDC and integrate it with your existing environment, migrate the appropriate applications and related supporting services to the cloud, and implement the right operating model to manage and operate your environment at scale. While on the surface these three steps may seem straightforward, there are many important considerations that can impact the speed and overall success of the implementation.