Data Center and cloud migration programs require unique preparation, resources, and processes due to the complexity and the cross-functional nature of the effort. Here are six cloud and data center migration strategies to leverage that will set your course in the right direction.
It is critical to leverage multiple technical sources but have one single source as a starting
point. For example, begin from a ping sweep to identify all devices that are live in an
environment and then leverage Active Directory, vCenter, and other server inventories to further
identify devices and their status. Physical inventories can also be useful for verification of
non-virtual devices no longer in use, never brought online, or only brought online occasionally.
Additional system reports from Azure, AWS, or Google Cloud discovery, database
inventories, migration tools,
firewalls, storage environments, existing CMDB’s should be exploited as well.
Once a baseline inventory has been established, application to server and dependency mapping
should be performed utilizing dependency discovery solutions, cloud dependency reporting,
application portfolios, SME interviews, and URL reporting tools. Inventories should be re-run
periodically to account for changes (decommissioning, installs, etc.) during the migration
project.
It is so important to maintain a single data source with limited change access and a process in
place for data management. The accuracy of your data is one of the most essential pieces to your
migration effort, and it will constantly change. Without the proper oversight and resource(s)
managing the data, you are bound to have erroneous information. This can cause minor rework or
worse yet result in the wrong server or application migration causing an unwanted outage.
Assigning a lead resource with the right organizational and technical skills to manage the data
is key. They must understand the data, so they know when it is out of place or likely
inaccurate. They must also systematically adhere to the process. I have heard application owners
frequently note that they do not know anything about a server and do not own the application on
it, to later confirm it is their application, and it is either crucial to a few users or is ready
for archive and retirement. Removing an application or application owner’s name from the
inventory without thorough checks would have resulted in the loss of important data.
Although data center migrations are generally sponsored by Infrastructure and require plenty of
infrastructure resources, the most challenging component of any migration is the
application. An IP address change could have a major impact if an application has a hard-coded
IP address. Applications with a lot of interdependencies must be moved together if minor latency
can significantly impact performance. Service account or privileged accounts, firewall rules,
load balancing, mail relay, DNS updates, domains, time zones, antivirus, or security
applications, patching software, SSL certs, backup environments, FTP jobs, etc. could all change
as part of a migration and must be prepared for by the application team with the support of
Infrastructure.
Engage the application teams early and get their leadership's commitment to timelines. Align
with them on the migration strategy and ensure they understand their involvement.
Keep them
updated on progress and as you execute migrations, ensure careful coordination, and communicate
continually.
As mentioned, communication with the application team early and continually throughout the
project is key. In addition, communication across all Infrastructure teams and resources
assigned to execute the changes must be a priority. Everyone must understand the steps to be
taken, the order and time they will be executed, and the owners of each task. This is the
case for prerequisite tasks as well as activities during the migration windows.
I highly recommend routine status meetings to review progress made to date, upcoming
activities, and migrations, and move group planning meetings for each migration iteration.
In addition, in large organizations, holding separate Infrastructure and Application team
meetings can reduce confusion. Program Management and Technical leads should bridge the
communication between the organizations.
During migration windows, Migration Program leads should coordinate conference calls and
other communication to ensure all the teams involved and impacted are aligned. Frequent
status updates should be sent out, so all the teams and executive stakeholders are kept
current on the progress. On-call preparations should be made for weekend and overnight work
in case teams run into any issues.
Transitioning environments in phases minimizes the risk of extended downtime and failure.
Migrating low-risk, non-production environments prior to critical production
applications is highly recommended. In some cases, applications may not have a
non-production environment, however, a test environment can be spun up temporarily to
run through the migration. This helps provide the migration team with critical knowledge
and experience before migrating the production application.
Key application dependencies must be migrated together, but it is also important that
the migration workload is manageable for the resources impacted. If an application owner
has too many applications migrating in each iteration, they will not be able to complete
all their necessary work within the migration window. The same is true of Infrastructure
resources. Move Groups must be balanced based on application environments, application
dependencies, resource availability, and key business dates.
Changing too many aspects of an environment at one time is a recipe for disaster.
Minimize the number of changes completed during the migration period, and make changes in
phases where possible. Making significant upgrades while migrating an environment
increases complexity and the chance for failure. Although it can take more time, the
safer option is to make the changes in phases. Upgrade the environment in place, and
then migrate it or vice versa. When moving to the cloud, older applications may need to
be upgraded prior to the transition.
One of the most important components to a successful data center or cloud migration,
or any project for that
matter, is executive sponsorship. Often the head of Infrastructure or the CIO sponsors
migration efforts. The application team’s leadership must be also onboard with the
priority of the initiative. Without the necessary sponsorship, migration projects can
drag out increasing the cost and inefficiently use project resources.
Executive leadership needs to send clear guidance across all teams, including
Infrastructure and Application teams, regarding the priority of the program and the
support required. It is critical that the executive sponsor attend routine executive
status briefings to stay informed of progress and any issues, even if they do not
require intervention. Executive sponsors should be consulted and approve key
communications, so they can re-enforce the message with their peers and senior
leadership. Executive Sponsors should be copied on major milestones met by the teams to
congratulate them on the progress toward the program goals. Through continued visible
engagement, the Executive Sponsor highlights the criticality of the program, empowers
program leadership, and encourages the team to stay the course through the effort.
Data Center and cloud migrations are complex, and CTM Technology Group has the technical expertise to guide you through the process. Employing these strategies and more, CTM will ensure your migration program is successful. See our Data Center Migration Services and our Cloud Migration Services for details on our practices. Check out our Success Stories for data center migration case studies.
Contact us for a free consultation
Data Center Migrations are highly visible, strategic programs that must be implemented without negatively impacting essential business services. CTM has the expertise to lead your team through the initiative and minimize your risks.
Migrating to the Azure, AWS, or Google Cloud? CTM Technology Group can help create a strategy and a roadmap to migrate your legacy environments to the cloud.
CTM Technology Group successfully executed two data center migrations two months ahead of schedule. We migrated 750 application instances and 900 servers in 20 weeks.