Recommendation Details Modal
Last updated
Last updated
To access the Recommendation Details modal dialog, click a recommendation from the WAN Insights > Recommendations summary screen.
The detail information about the recommendation includes the following items:
Projected quality shows the path quality that you can expect if you make the recommended changes. The projected path quality is shown along with the amount of improvement.
Capacity Check shows, for the recommended path change, how much capacity is already consumed, the additional traffic that would be added, and if that additional will be supported. The capacity check looks at peak traffic when making this assessment, so even at peak traffic volumes we know the recommendation is still applicable.
Use the Request to Apply button to initiate the suggested path changes in vManage as described in Applying WAN Insights Recommendations.
Recommendation Status is described on Life of a Recommendation.
The Request to Apply button is only visible for tenants running a compatible vManage version as listed in Supported SD-WAN Software Versions, and the user must have administrator permissions on that tenant.
The Site Quality Comparison timeline shows hourly user experience for the last 30 days.
The timeline graph shows user experience over the last 24 hours for the site:
The left-hand Y axis labeled Mean Quality (%) shows an aggregated quality score.
The right-hand Y axis Hourly Active Users indicates the hourly peak user counts over a 24-hour period, shown using the vertical gray bars.
Swimlane below lets you select any time period over the last 30 days.
Cross hatching indicates the last 7 days of historical data or the evaluation period that was last used to create or re-evaluate this recommendation.
Recommended Path Quality and Current Path Quality lines are shown for time periods with traffic. If there’s a gap, it means that no user application traffic was detected in that hour.
The recommendation summary displays in a banner at the top.
The swimlane just beneath the timeline shows the last 30 days of quality data, with Evaluation start indicating the starting point of the evaluation period.
A similar graph appears when an endpoint pair is selected, that shows quality related to that particular pair of endpoints.
The color of the cross hatching on the timeline shows when the recommendation was generated.
Once generated, recommendations are continuously re-evaluated based on the last 7 days of historical data. A color-coded cross hatching appears on the timeline:
Green indicates the evaluation period (last 7 days).
Red indicates that this quality data is no longer used for calculation.
How to use this chart:
Policy Verification: Confirm that your configured SD-WAN network policy is actively in effect by verifying the traffic distribution across various interfaces.
Performance Monitoring: Identify and monitor traffic volume routed through poorly performing interfaces, enabling timely corrective actions.
Quality and Bandwidth Correlation: Empower your network teams to correlate quality drops on SD-WAN interfaces with throughput data. This is the first step in establishing bandwidth saturation as the root cause of performance issues.
The Throughput & Quality Analysis timeline immediately beneath Site Quality Comparison shows overall throughput and which circuits are carrying more traffic currently, along with the quality for the selected circuit. You can investigate whether any path quality changes coincide with changes in throughput, for example whether a quality drop was potentially caused by a surge in application traffic.
Choose a circuit from the drop-down list.
Use the slider to show ingress or egress. Note that the volume of traffic can be different for egress or ingress, even for the same circuit.
Optionally, use a second slider to show the current application as well as total throughput. You can see how much of this throughput is consumed by the application to which this recommendation applies.
The throughput measures shown are based on total volume of application traffic, not the number of users or user sessions. The shaded area is the throughput, and the line is the quality of that circuit.
See Estimating Throughput for information on how throughput data is derived.
Throughput data is only available for cEdge routers. See Enabling WAN Insights for more information on supported SD-WAN software versions.
Hovering over the timeline displays a measure of circuit quality at that moment in time, which is shown as Interface Quality on the throughput graph.
The quality shown per circuit is obtained using the probes as described on Understanding Quality. These probes work differently depending on whether the application traffic type is peer-to-peer (P2P) or direct internet access (DIA).
For P2P traffic (between the sites in the SD-WAN fabric), the bidirectional forwarding probe shows round-trip time. Therefore, congestion on either ingress or egress may impact quality. For this type of network traffic, the recommendation only affects egress traffic (how it’s leaving the site) since the remote site has its own routing policies that control how to route back the returning traffic.
For DIA traffic, traffic egressing the router over one interface will return over the same interface. Therefore, for this type of traffic the recommendation affects ingress as well as egress.
Quality is shown for both types of network traffic, so that customers can analyze ingress and egress traffic patterns, and investigate whether quality drops are correlated with traffic spikes.
Use the section called Quality Improvement per Endpoint-Pair to see which endpoint pair would most benefit from a change in path.
This screen shows all endpoint pairs, until you select one pair and then it’ll expand to show the Quality of Service (QoS) metrics for just that one path. There are two views, a radial view and a stacked or sankey view.
For all endpoint pairs:
Source routers are in gray. If there is more than one, the gray portions will be divided.
Destination routers are shown adjacent to the source routers. They are proportionally different based on the number of active users at this site. A larger bar means more users are impacted, and thus that particular endpoint-pair might be worth investigating further.
The percent improvement is shown as a bar extending out from each destination router. So, at a glance you can tell which endpoint pairs (paths) have the most potential for improvement, and also which ones are impacting the greatest number of users.
Note that a recommendation for a Hub site may have a negative impact on some of the Spoke sites that are connected to this Hub. For most of these Spokes, the recommendation will improve the quality, while for some Spoke sites, the quality will stay neutral. However the recommendation may decrease the current quality for a few other Spoke sites. This is a trade-off of site-level recommendations. Overall, the recommendation will still improve quality for the majority of sites and users.
Here’s a comparison of the two views, showing the same data.
