network performance monitoring

network performance monitoring
SOLUTION BRIEF
NETWORK PERFORMANCE MONITORING
REAL-TIME VIEW OF END-USER EXPERIENCE
This solution brief describes how Netrounds can be used for performance
management (PM) and fault management (FM) in a telecom operator’s
network. These are fundamental activities of Operations Managers
overseeing the network from a NOC and responsible for assuring network
and service health. In this role, your responsibilities typically include
obtaining an adequate view of network and service quality as perceived
by end-users, promptly sharing actionable insights with engineering and
support teams whenever problems arise.
NETROUNDS FEATURES
•
quality over time and go
back to find the root cause
of problems
•
As you have probably noticed, the preferred traditional method of
passively observing counters in network nodes is not sufficient to achieve
this. Not only is the time resolution of this data often too low, but the
correlation with actual user experience is too weak.
With Netrounds Test Agents, on the other hand, you can actively test
your network and services like an end-user, gathering fine-grained realtime data that faithfully reflects end-user experience.
The image below depicts an authentic Netrounds deployment in a Tier 1
telecom operator’s network. Full-mesh monitoring is conducted among
25 Test Agents installed in transport/core and distribution network nodes.
In full-mesh monitoring, all Test Agents send traffic to all others; between
each pair of Test Agents, UDP or TCP streams are sent concurrently for
each of four quality-of-service classes to be tested. Any errors that occur,
and that violate the configured thresholds, are reported in real time to
the NOC so that its staff can take quick action – ideally before end-users
notice and contact the help desk.
Historical data – Monitor
Alarms – Work proactively,
be notified of negative
trends and solve problems
before end-users notice
•
Collaboration – Share Test
Agents and data within
your own organization or
with vendors or clients
•
Versatile Test Agents –
Monitor both networklayer performance and
individual services with
the same Test Agents
SOLUTION BRIEF
To supplement the full-mesh monitoring, a further 300 Test Agents are
deployed across less central locations in the network, mainly in distribution
and access nodes. These peripheral Test Agents are located on spokes
going out from each of the 25 Test Agents next to the core nodes.
You can enhance your testing coverage further by having Test Agents
interact with existing network elements, using the Y.1731 and TWAMP
protocols (on Layer 2 and Layer 3 respectively), which are built into many of
today’s switches and routers. In this manner even more pathways in the
network can be tested, and you can leverage previous infrastructure
investments by making additional use of switches and routers already
deployed.
Measurements
collected
Data storage and presentation
Packet reordering
All collected data is consolidated in the Netrounds Control Center, which is
either hosted as a cloud service or deployed on-site at your premises.
Packet duplication
In the user interface, data is presented in a zoomable real-time view and
can also be distilled into on-demand or scheduled periodic reports. In
addition, data can be extracted through an API and correlated with data
from other sources in your OSS environment, such as alarm logs or events.
Service-specific testing
The same Test Agents used for the network-layer performance
measurements described here can also be employed to concurrently gauge
the performance of specific services such as VoIP, IPTV, multicast UDP, or
other business-critical services. This is in stark contrast to most competing
active measurement probes, which are tailored to a specific service so that
you need to place several of them in parallel in your network, driving
unnecessary cost. The versatility and affordability of Netrounds Test Agents
make them uniquely cost-efficient.
Targeted tests
Continuous monitoring can be complemented by periodic or one-off tests.
You can have such targeted testing triggered by alarms: for example,
whenever the jitter exceeds the SLA “Bad” threshold, you can run a test
checking QoS scheduler behavior with and without excessive load. One
reason for severe jitter is that priority headers have been corrupted.
Garbled priority headers will distort QoS classification and cause the traffic
to be incorrectly prioritized. A diagnostic test automatically helps expose
the root cause of problems such as these.
It needs to be kept in mind that such targeted tests are more intrusive and
may affect end-users’ experience of the network, whereas continuous
monitoring is normally unnoticeable to end-users.
HOW TO FIND OUT MORE ABOUT NETROUNDS
Call our local sales office and speak with a solution advisor to schedule a
personal demo, get answers to any questions you might have, or ask for a
quote: +1-617-671-0057 (North America); +46-70-592-5345 (Europe/World).
You can also contact us by email at sales@netrounds.com.
Netrounds Solutions AB
Storgatan 7, 972 38, LULEÅ, SWEDEN
+46 920 42 00 15
info@netrounds.com
www.netrounds.com
Packet loss ratio (PLR)
Packet delay (PD)
Packet delay variation
(PDV; also known as jitter)
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising