View impacted VPN clusters to understand what apps are
experiencing performance issues.
Understanding what is causing degraded performance
for your apps and services is integral to ensuring the user experience
is not impacted. Understanding why your VPN clusters are impacted
and app traffic failed over to different links helps in fine tuning
your SD-WAN configuration.
Filter
the VPN clusters based on your preferred metric from the
Site
drop-down
and select time frame. In this example, we are viewing
All
Sites
containing impacted VPN clusters in the last 12
hours.
In the Sites column, select the impacted hub or branch
firewall to view the impacted apps and the corresponding link performance.
In the App Performance section, click an app to view
detailed Traffic Characteristic information about the app traffic
such as the internet service(s) and links used:
Review the pie chart to understand the breakdown
of app traffic across the your internet services.
Review the linegraph to understand how many bytes of data
were transferred over each internet service over time.
Review the Links Used section to understand which links the
app traffic used and to understand how many of the bytes were impacted
out of the total bytes in the selected time frame.
Investigate which health metric caused the app to swap
links.
The dotted line indicates the seven day average for the
health metric.
In the Links Used section of the Traffic
Characteristics tab, click an ethernet Link to view detailed Link
Characteristics (latency, jitter, and packet loss) over the time
frame specified in Step 2 to investigate
what health metric caused the app to swap links. In this example,
we are viewing ethernet 1/1 and can see that the percentage of packets
lost regularly exceeded the seven day average threshold for the
app and can conclude that this is the reason the app traffic failed
over to the next best link.
In the
Traffic Characteristics
tab,
select another link to view the Link Characteristics. In this example,
we are viewing ethernet 1/4 and can see that after the app traffic
failed over, ethernet 1/4 experienced jitter for the link that exceeded
the seven day average threshold. This forced the app traffic to
fail over back to ethernet 1/1.
Since both links had health metrics that were exceeded,
the app traffic had no healthy link to fail over to resulting in
the VPN cluster becoming impacted.
After you have identified why the app traffic is impacted,
consider the following to resolve the issue:
Consider adding additional links to the Traffic Distribution
Profile. By adding additional links for app traffic to fail
over to, you help ensure that the app traffic and user experience
are not impacted by links with degraded health.
Reconfigure the health thresholds in your Path Quality Profile.
It may be that the health thresholds are too strict, resulting in
unnecessary link fail over. For example, if you have an app that
can experience up to 18% packet loss before user experience is impacted,
having a 10% packet loss threshold would result in the app failing
over to a different link without a need to.
Consult your internet service providor (ISP) to determine
if there are impacts to your network outside of your control that
they can resolve.