By Jubil Mathew, Technical Marketing Engineer at LiveAction
Enterprise networks are more complex and sophisticated than ever
before. Managed wide area network (WAN) services, such as MPLS, continue to
play a significant role in enterprise networks, but the internet has become a feasible
WAN connectivity option. It's actually essential in the case of direct cloud
connectivity. As a result, many enterprises adopt hybrid networks, a
combination of managed WAN services and internet to address various WAN
requirements. Although the internet is less reliable than managed WAN services,
it can provide affordable, high-performance connectivity. Most enterprises
experience improved performance and user experiences when relying on the
internet for WAN connectivity.
That said, the hybrid nature of modern IT infrastructure makes it
more difficult than ever to manage. Without end-to-end visibility, it's nearly
impossible for network administrators to do so effectively. Relying on the
internet for network operations can cause headaches. In fact, new
EMA research shows that today's top hybrid WAN challenges include maintaining
application performance levels, inconsistent quality across service providers, varying
performance across geographies, poor visibility and more. Let's take a closer
look at the metrics that matter most in modern WANOps:
- Active Monitoring Data - Synthetic monitoring allows you
to assess enterprise applications by simulating user or synthetic traffic. This
enables you to proactively calculate critical issues applications traversing
the network might come across. This type of active monitoring can help eliminate
network availability issues, functionality problems and poor application
performance, improving overall user experiences.
- Endpoint Transaction Data - Understanding endpoint usage,
types of user applications, operating systems, security rules and more is
critical when it comes to managing overall enterprise network usage and user
experiences. Endpoint transaction data can help you monitor for and understand
endpoint security and performance issues, manage security risks, and administer
software updates.
- Routing Data - Tracking network routing
changes, WAN failures and alerts based on routing data helps you understand enterprise
network behavior over time and take remedial action when necessary. This
information can help you issue needed service alerts and make backup design
changes to ensure smooth network functionality.
- Network Flow Data - Using flow records from network
devices to analyze traffic flows, bandwidth volume and usage, and application
performance is essential when it comes to understanding critical information
such as traffic destination, type and volume across the enterprise. Most
administrators make network management decisions with these valuable insights
using reports generated from the traffic statistic exports of NetFlow-enabled
devices.
- Cloud Provider Flow Logs - Similar to analyzing network
traffic from on-premises devices with the help of NetFlow or IPFix records, you
need to understand and analyze traffic flow to and from cloud providers to
understand and manage performance levels across cloud-hosted applications and
services.
- Packet Data - You can store and use packet
capture from network segments to inspect and baseline enterprise application
behavior, analyze traffic patterns and network usage, and troubleshoot network
or application performance problems.
- Device Metrics - SNMP data gives you access to
true device-level information, such as CPU and memory utilization, interface
bandwidth utilization and hardware-level indicators like fan and power supply status,
and more. You can use API-level information to integrate automation with
network monitoring tools for service and event management.
Modern WANOps Requires Meaningful Visibility
According to EMA, network administrators see visibility into
end-to-end loss, latency, and jitter across internet paths and internet and ISP
outage reports as the most valuable WAN insights. DNS availability and
resolution time are quite valuable as well. Network administrators also
reported that today, hop-by-hop loss, latency, and jitter are less valuable
than end-to-end measurements, while CDN metrics and BGP routing changes are of
least importance. Network administrators prefer to have end-to-end visibility
for enterprise traffic to better understand overall user experience and traffic
patterns. End-to-end visibility means establishing insights at the desktop or
virtual machine level, across WAN-Edge devices, and throughout the service
provider core and data center devices.
When asked what to identify the top root causes of WAN issues
today, 30% of respondents listed application errors and performance, while 30%
cited network providers (ISP or MPLS providers) and 28% listed end-user error
or client device failure. You and your team must be able to collect and analyze
all the disparate data types listed above to identify and resolve these root
causes. For example, if your users complain about poor voice or video service
quality, you won be able to identify the source of the degradation without visibility
into traffic or application-level monitoring. In this case, you're left
wondering if it's a device or OS-level issue, a service provider problem, or
something else. Visibility into end-to-end flow path for a specific
conversation can show different performance metrics, DSCP markings and if there
are any quality-of-service issues. End-user OS issues can also be a cause of
bad application performance, so it important to understand end-to-end visibility.
Visibility gaps across any of the above data types and metrics can allow
performance issues to persist, and bring productivity and business operations
to a halt.
Monitoring the Metrics That Matter
Most organizations use between four and 10 unique tools to monitor
and troubleshoot their networks, from the data center to the cloud and from the
LAN to the WAN. This can produce unintended productivity challenges from a
workflow standpoint (resulting in further blind spots and performance issues),
but it's also incredibly expensive in terms of licensing, support, specialized
training, etc. Fortunately, some advanced network
monitoring solutions offer consolidated functionality, enabling NetOps teams to
see into the dark corners of each domain and better manage, optimize, and
troubleshoot their hybrid networks.
Regardless of whether you're using multiple point solutions or a
unified network management approach, you must be able to monitor the
aforementioned metrics to establish the level of visibility today's hybrid WANs
demand.
##
About the Author
Jubil Mathew is a
technical marketing engineer at LiveAction. In his current role, he supports
multi-vendor integrations with LiveAction's solutions and educates
organizations on advancements in the company's cloud deployment and analytics
platform and SD-WAN capabilities. Jubil has expertise in cloud monitoring
deployments, SD-WAN, and network performance management and diagnostics, and
was previously a part of Cisco's IWAN solutions architecture and design team.
Learn more about LiveAction here.