By Jay
Botelho, Director of Engineering at LiveAction
Enterprise networks
are in a constant state of flux. Businesses are always looking to incorporate
new technologies and meet changing operational requirements. Unfortunately, network
change initiatives can be costly and time-consuming without a proactive
approach to ensuring the right data is available to drive them.
Among the most
common network change projects today are cloud migrations, SD-WAN rollouts and 802.11ax
upgrades. Data should be your guide when removing, upgrading or replacing any
IT infrastructure and managing a transition to such technologies. You must
monitor key performance metrics for all network elements involved before,
during and after every network change operation. Failing to do so will elevate
your risks of failed deployments, as well as obfuscated performance issues,
poor user experiences and more, throughout rollout and beyond.
Given COVID-19’s
impact on how and where people leverage network resources, cloud migrations are
particularly timely network change initiatives we can examine to understand
importance of a data-driven approach. Let’s
explore the consequences of a cloud migration project without the necessary
data involved, how data impacts each step in the process and the visibility you
need to succeed.
A cloud migration without foundational
data can cause many complex issues. These can range from poor connectivity and
higher latency, to even security issues. For example, after migrating several
key business applications users might experience increased latency. But is it
truly worse than before? Is it unacceptable? And are the migrated applications
really the cause? Or could it be due to increased VPN connections and bandwidth
as more users working remotely attempt to access the new cloud services? Without
solid data from before the migration, these are difficult questions to answer.
And they need to be answered quickly because a perceived degradation in
performance will encourage employees to circumvent established processes
requiring VPN usage to access key cloud-based applications such as Salesforce,
WebEx or Zoom. This would change the workflow before a clear diagnosis can be
made, and make things less secure by reducing your visibility into user activity
and any suspicious anomalies. Data is the key to getting in front of just about
every cloud migration issue.
Basing your migration on data throughout
the initiative can deliver end-to-end visibility from on-premises environments
into the public cloud, and help ensure a successful rollout. From “Day 0”
planning and “Day 1” deployment to “Day 2” ongoing monitoring and optimization data
is king when it comes to cloud migrations. Why?
First, establishing a baseline across
your existing IT infrastructure upfront can pay big dividends throughout your
cloud migration. To start, measure key data and metrics to define what’s “normal” for network performance levels, application performance trends, and
behaviors across users, devices, key services and more. Leverage all this
information to map out existing bandwidth usage and throughput patterns, SLA
requirements and quality of service (QoS) policies. Without collecting and
understanding these data upfront, you’ll lack the context and specifics you
need to be able to truly determine, tune and control how your new cloud
deployment is functioning. It’s also critical that you have the solutions in
place to ensure that the visibility and data you’re able to access
pre-deployment carries over across the cloud migration.
Next, successful cloud deployments
will rely heavily on the data you’ve collected pre-rollout. The
migration itself will be a true test of how well your team has planned the
initiative and if you’ve established the historical baselines needed to effectively
measure and manage post-migration. You’ll need to quickly identify and resolve
any network or application performance issues such as poor connectivity, high
latency, unforeseen capacity limitations, degraded user experiences and more,
as well as verify the SLAs and QoS policies you established during the planning
process.
Whether you’re migrating limited
portions of your system such as a few specific databases or servers, or an
entire application stack or data center, you need deep, end-to-end visibility
from on-prem into the public cloud, and into VPC traffic and the cloud services
running through it. Most cloud monitoring tools are burdensome to manage
alongside existing monitoring products and can’t provide a comprehensive view
of network or application issues that extend across the hybrid environment.
This goes for both monitoring dashboards from cloud providers themselves as
well as specialized point solutions. That’s why it’s critical to leverage
advanced monitoring solutions capable of capturing network traffic that
traverse the public cloud and converting it into flow data for in-depth 360-degree
performance analytics and visualization, all using the same integrated solution.
Without this level of detail, you’ll lack a complete understanding of traffic
behavior, application usage and performance within your new cloud
infrastructure, and be unable to verify the new implementation is working as
planned.
Finally, all that data impacts what
level of visibility you’ll have when it comes to ongoing cloud monitoring and
optimization. This is all about continual improvement. Your goal is to continuously
monitor the deployment, proactively identify and resolve issues before they
happen, and optimize performance to meet your business’ needs. If you’ve
established the baselines and visibility required to identify network and
application performance issues across your cloud workloads, you should be able
to take advantage of automated alerting that proactively “predicts” potential
issues – or at least the warning signs that those issues might arise – so you
can mitigate them before they impact the business. For instance, if you have multiple
automated alerts of application latency exceeding your defined threshold, and
these reports span a wide range of users, you can proactively test the latency from your
location and quickly determine if the source is in the cloud. And if so, assuming you have a solution in place that can also capture
network data from the cloud, you can quickly isolate the issue and the source,
and take corrective action.
Successful cloud migrations require end-to-end
visibility that extends from the network into the cloud. This is just one
example of the many network change initiatives many organizations are focusing
on today. But the single most important element to the success of any network
change project is data. To better support and enable your business in 2021 and
beyond, be sure to collect and analyze in-depth data you can use when planning,
deploying and optimizing new IT initiatives.
##
About the Author
Jay Botelho is the Director of Engineering at LiveAction.
Leave a Reply