cisco-sdwan-application-aware-routing-deploy-guide
cisco-sdwan-application-aware-routing-deploy-guide
Cisco Public
July, 2020
Introduction
This guide is intended to provide design and deployment guidance to deploy Application-Aware Routing on the
Cisco SD-WAN solution providing Service Level Agreement (SLA) based routing for business-critical
The ability to consider the path characteristic in path selection offers a number of advantages to the Cisco SD-
WAN solution:
● In normal network operation, the path taken by application traffic through the network can be optimized
by directing it to WAN links that support the required levels of packet loss, latency, and jitter defined in an
application’s SLA.
● In the face of network brownouts or soft failures, performance degradation can be minimized. The
tracking of network and path conditions by application-aware routing in real time can quickly reveal
performance issues, and it automatically activates strategies that redirect business critical data traffic to
the best available paths that meets the SLA. As the network recovers from the brownout or soft failure
conditions, application-aware routing automatically readjusts the data traffic paths.
● Network costs can be reduced because data traffic can be more efficiently load-balanced.
● Application performance can be increased without the need for WAN upgrades.
Although this deployment guide is about Application Aware Routing. It is presumed that
● Cisco SD-WAN Controllers (vManage, vBond, and vSmart) are already deployed with valid certificates.
● Cisco WAN Edge device is onboarded and have established control connections to Cisco SD-WAN
controllers and data tunnels to other WAN Edge devices across all available transports.
● Cisco SD-WAN WAN Edge and vSmart controller have configuration – feature templates defined, and
device template associated and are in vManage mode.
For more information on SD-WAN controller design and deployment, please refer to the Cisco SD-WAN Design
guide, Cisco SD-WAN End-to-End Deployment guide and the Cisco WAN Edge Onboarding Deployment guide.
The Define section provides a high-level overview of the SD-WAN architecture and components and
Application-Aware Routing components.
The Design section provides detailed discussion on the design considerations and prerequisites needed to
deploy Application-Aware Routing.
The Deploy section discusses step-by-step procedures to configure Application-Aware Routing policies in the
Cisco SD-WAN network. It walks through the best practices and gotchas to consider during the process.
The Operate section briefly discusses how to monitor and troubleshoot the common issues.
Refer to Appendix A for details on the platform and software versions used to build this document.
Audience
Define
About the solution
The Cisco SD-WAN solution is an enterprise-grade SD-WAN architecture overlay that enables digital and cloud
transformation for enterprise. The solution fully integrates routing, security, centralized policy and orchestration
into large-scale networks and addresses the problems and challenges of common WAN deployments.
The Cisco SD-WAN solution is comprised of separate orchestration, management, control and data plane.
● Orchestration plane assists in securely onboarding the SD-WAN WAN Edge routers into the SD-WAN
overlay. The vBond controller, or orchestrator, authenticates and authorizes the SD-WAN components
onto the network. The vBond orchestrator takes an added responsibility to distribute the list of vSmart and
vManage controller information to the WAN Edge routers.
● Management plane is responsible for central configuration and monitoring. The vManage controller is the
centralized network management system that provides a single pane of glass GUI interface to easily
deploy, configure, monitor and troubleshoot all Cisco SD-WAN components in the network.
● Control plane builds and maintains the network topology and make decisions on the traffic flows. The
vSmart controller disseminates control plane information between WAN Edge devices, implements control
plane policies and distributes data plane policies to network devices for enforcement.
● Data plane is responsible for forwarding packets based on decisions from the control plane. WAN Edge
physical or virtual devices provide secure data-plane connectivity between the sites in the same SD-WAN
overlay network. WAN Edge devices are responsible for establishing secure connections for traffic
forwarding, for security, encryption, Quality of Service (QoS) enforcement and more.
Upon securely onboarding the WAN Edge device in the Cisco SD-WAN overlay network, the WAN Edge device
establishes secure control connections with all the controllers (vBond, vManage, vSmart) where it receives
configuration, policies and routing information. The WAN Edge device would then establish secure IPsec tunnels
with other WAN Edges, which is part of the same SD-WAN overlay network, to forward data-traffic.
The Cisco SD-WAN solution leverages NETCONF to provision the WAN Edge devices with the associated
template configuration in the vManage and Overlay Management Protocol (OMP) to convey the control-plane
information such as route-prefixes, next-hop routes, crypto keys and policy information between the vSmart
controllers and the WAN Edge devices. By default, with no policies defined, the SD-WAN overlay network would
form full-mesh topology, allowing each WAN Edge device to establish a secure IPsec connection to other WAN
Edge device.
It is important to note that the WAN Edge device establishes a permanent persistent connection to the vSmart
controller over each available WAN transport and a single permanent persistence connection to the vManage
over only one WAN transport.
The BFD operates in echo mode where BFD messages generated by the WAN Edge device are reflected
(echoed) back by the remote WAN Edge device. Each BFD packet is time-stamped by the originating WAN
Edge device to determine the round-trip latency and jitter. Path loss or tunnel liveness is determined based on
the lost BFD packets.
The WAN Edge device collects the packet loss, latency and jitter for every BFD probe (default BFD Hello packets
are sent 1 sec on every WAN transport) and is preserved for every poll interval (the default poll interval value is
10 minutes). The network path liveliness, by default, is calculated for a period of 6 poll intervals for better
accuracy and to dampen any intermittent reclassification (flapping) of the tunnel. At the seventh poll interval, the
earliest polling data is discarded to accommodate the latest information.
BFD packets being forwarded at regular intervals enables the SD-WAN overlay network to not only detect any
blackout situations but also detect path characteristics such loss, latency, jitter, path-MTU that can then be
leveraged by other SD-WAN protocols to make dynamic decision and provide the best quality of experience for
business-critical applications.
The WAN Edge devices determines the WAN transports path characteristics - loss, latency and jitter from the
previously discussed BFD probes sent across each tunnel between the WAN Edge devices.
● Packet loss is calculated on the WAN Edge device on a per tunnel basis and is measured as percentage 0
through 100 percent.
● Jitter is calculated on the WAN Edge device on a per tunnel basis and is a measurement of millisecond, 0
through 1000 milliseconds.
● Latency is calculated on the WAN Edge device on a per tunnel basic and is a measurement of millisecond,
0 through 1000 milliseconds.
Note: Calculated Packet Loss, Jitter and Latency are average values for the poll intervals and are round-trip
measurements on each tunnel interface on the WAN Edge device.
Application-Aware Routing leverages the calculated network path characteristics values as a measurement and
compares with the desired SLA for the application, to dynamically optimize the data path selection for the traffic
on the WAN Edge device.
Policies
The Cisco SD-WAN solution separates the control plane functionality managed by vSmart controller with the
data plane functionality operated by the WAN Edge devices. Similarly, the Cisco SD-WAN policy architecture
separates the control policies that influence the routing information shared between vSmart controllers and the
WAN Edge devices, with the data policies that influences the data traffic between the WAN Edge devices.
By default, with no policy associated to the any SD-WAN components in the solution. In this scenario:
● WAN Edge devices advertise routes to vSmart controllers through the OMP protocol.
● The vSmart controller advertises the full routing information that is learnt from all WAN Edge devices to all
other WAN Edge devices through the OMP protocol.
● SD-WAN WAN Edge devices establish data plane connections to all other WAN Edge devices forming a
full-mesh topology.
● The centralized control policy is applied to the routing information that is stored in the vSmart controller
and enforced before the routes are advertised to the WAN Edge devices allowing for customizing routing
decisions and determining the routing paths through the overlay network.
The centralized control policy configuration remains on the vSmart controller and is never pushed to the
WAN Edge devices.
● The centralized data policy is applied to the data traffic flow for the specified VPNs in the overlay network.
These policies permit or restrict access based on 6-tuples (source-ip, destination-ip, source-port,
destination-port, protocol, dscp values) or based on VPN memberships allowing for customizing routing
decision and determining routing paths at a local site.
Depending on the policies, the policy is provisioned on vSmart, pushed (via OMP protocol) and enforced
on the WAN Edge devices.
● Localized control policy is applied to the routing information that is stored on the WAN Edge device
influencing the routing behavior on the device at local site level.
● Localized data policy is applied on the interface of the WAN Edge device. The data policy is leveraged to
match traffic and defines QoS, policing, mirroring at the interface level of the WAN Edge device.
Application-Aware Routing
Advanced features set such as Application-Aware Routing provide network administrators the necessary
flexibility to pin certain business-critical application to a specific WAN transport on the device. Actively probe
the network and path characteristics and dynamically re-route the application traffic, in real-time, on the WAN
transport links that meets the specified application SLA requirements.
Cisco SD-WAN WAN Edge device supports up to 8 TLOCs, allowing a single device to be connected to 8
different WAN transports. Each WAN Edge device in the SD-WAN environment advertises its local routes to the
vSmart controller using OMP protocol. The vSmart controller computes the best path selection algorithm for the
entire SD-WAN environment and applies any configured centralized control policy before advertises the route-
selection to the WAN Edge devices.
The WAN Edge device installs the received OMP route in its forwarding table. For destination prefixes with
multiple best paths, the device natively performs Layer-3 ECMP (Equal Cost Multiple Path) load-balancing
across 4 reachable next-hop TLOCs. The number of paths installed on the WAN Edge device can be increased
to 16 as long as the next-hop is reachable.
SLAs for each of the transport tunnels are calculated periodically using BFD probes by the WAN Edge device
and is made available for advanced features like Application-Aware Routing to leverage and provide
deterministic experience for business-critical applications.
Application-Aware Routing allows the network administrator to evaluate the network path characteristics for the
selected business critical applications and set a preferred path as long as the SLAs are satisfied and a backup
preferred path. The backup preferred path is chosen when no available WAN transport(s) meets the specified
SLA.
Application-Aware Routing policy is configured in vManage as a centralized data policy that maps the service-
side application(s) to specific SLA requirements. The centralized policies provisioned in vSmart controller is
pushed to relevant WAN Edge devices for enforcement. The defined policy consists of match-action pairs,
where the match statement defines the application-list or the type of traffic to match, and the action statement
defines the SLA action the WAN Edge devices must enforce for the specified traffic.
Prefix Network prefix that can be matched for source or destination addresses
In addition to above, network administrator can also match traffic based on:
Cloud SaaS Application List Leverage the pre-defined Cloud-SaaS Application list
DNS Application List DNS application list is mainly used when the network deployment needs
split DNS lookup for certain application or application-lists.
DSCP Pre-configured traffic with DSCP values, through QoS policy on the
service-side traffic, can be leveraged.
PLP Pre-configured traffic part of the Packet Loss Priority (PLP) queue,
configured part of QoS policy, can be matched.
Cisco IOS-XE SD-WAN devices can incorporate SD-AVC along with the NBAR2 DPI engine providing the
capability to inspect and classify the flows on the first packet. Once the applications are classified, policies can
leverage this information to match on the application and/or application-list and influence the traffic.
The vManage centralized policy wizard provides network administrator options to define custom Service Level
Agreement (SLA) or leverage the pre-defined SLA’s as shown below:
Transactional-Data 5 50 100
Voice-And-Video 2 45 100
Tech tip
Before defining custom SLA values, monitor the current WAN transport SLA statistics across the SD-WAN environment by
navigating to vManage > Main Dashboard > Application-Aware Routing widget.
By default, the Loss, Latency and Jitter values are calculated for every WAN transport available on the WAN Edge device
and are made available for a period of up to 7 days. This historical data can be used as a baseline to evaluate and define
custom SLA values to better fit the environment and the application requirements.
Careful consideration must to taken when choosing the SLA values. Choosing a more aggressive values might
cause undesired result with too many SLA violations, while choosing a more relaxed values might not yield the
desired result for the enterprise.
Administrators can define any number of custom SLA’s but can associate only 4 SLA’s to the Application-Aware
Routing policy. Please check the Release Notes for the corresponding software version on the number of SLA
supported for the release.
Application-Aware Routing Policy – Policy maps the classified traffic to the WAN transport based on
the defined SLA requirement.
The Application-Aware Routing policy defined in vManage binds the selected application/traffic list with the
SLA. For all matched data-traffic traversing from LAN/Service side to remote site through the device WAN
transports, the AAR policy defines the
● Preferred Color – the selected data traffic is pinned to the chosen WAN transport(s) as long as the
transport(s) meets the specified SLA.
Note that when multiple WAN transports are selected and transports satisfying the SLA requirements, the
WAN Edge performs ECMP load balancing across the tunnels for the selected data traffic.
● Strict – if enabled, the selected data traffic would be dropped if any of the WAN transport(s) doesn’t meet
the specified SLA.
● Backup SLA Preferred Color – the selected data traffic is pinned to the chosen WAN transport(s) only
when no transport(s) meets the specified SLA and Strict option is not enabled.
It is important to note that the traffic flow that is switched to backup WAN transport will not be switched
back to preferred WAN transport incase the transport meets the SLA.
● Log – if enabled, a syslog message is generated first time a packet flow is logged and every 5 minutes
thereafter, as long as the flow is active.
For all other data traffic that doesn’t match the selected application/traffic list, the traffic would be load-
balanced across all the available WAN transport links available on the WAN Edge device.
Along with defining the preferred dynamic path selection, AAR policy provides the network administrators with
flexibility of defining Site-list’s and VPN-list’s where the policy needs to be enforced across the SD-WAN
overlay network.
Following the WAN Edge onboarding – authenticating and joining the overlay network, the device would
establish secure data-tunnel to other WAN Edge devices part of the same SD-WAN overlay network. Upon
establishing the data-tunnel, the WAN Edge device would:
● probe each WAN transport tunnel with BFD Hello packets at every 1 second interval. Below table lists the
BFD default values.
List Description Default Values
Hello interval Interval at which BFD Hello Interval are send across the WAN 1000 msec
transport
(configurable)
BFD Multiplier Value defines number of BFD Hello Packet Intervals the device 7
waits before declaring that tunnel has failed.
(configurable)
Note: The BFD Hello Interval and BFD Multiplier can be changed per tunnel or across all the WAN transports on
the WAN Edge device.
● collect and preserve the packet loss, latency and jitter for every BFD poll and calculate the network path
characteristics at each poll interval. The default poll interval value is 10 minutes, implies 600 BFD hello
packets are considered for each poll interval calculation.
Poll Interval = poll-interval in secs * BFD Hello packet interval
● collect 6 poll intervals and average these values to calculate network path liveliness for better accuracy
and to dampen any intermittent reclassification (flapping) of the tunnel. At the seventh poll interval, the
earliest polling data is discarded to accommodate the latest information.
Below table lists the default values for the Application Aware Routing polling data:
List Description Default Values
Poll Interval Interval at which WAN Edge calculates the average loss, latency and 10 minutes
jitter for each WAN transport.
(configurable)
Note: The Poll Interval and Multiplier values can be modified on the WAN Edge device effecting all the
tunnel/WAN transport(s) associated with the device.
Careful consideration must be taken before changing the default values as the BFD packets gets high priority
treatment on the WAN Edge device. The BFD packets are marked with high priority DSCP 48 marking. By
default, the control traffic and the BFD packets are mapped to Queue 0 on the device and is configured for low-
latency queuing, strict-priority traffic queue for control and delay-sensitive traffic. Packets in this queue is
Being aggressive by lowering the BFD values would impact the WAN Edge device performance and being
conservative by increasing the BFD values would delay the calculations of the network path resulting in
undesired dynamic path selection for the business-critical applications.
Tech tip
For optimal device performance, Cisco recommends not to change the default BFD values as this might impact WAN Edge
performance.
In situation where the default values need to be modified to meet the business requirements, such as to meet
the SLA requirements for highly susceptible applications or to reduce bandwidth consumption on the tunnel or
to reduce the WAN charges at the remote site. Additional caution and tests must be done before deploying in
production environment.
It is always recommended to have a consistent BFD values configured across WAN Edge devices in the same
site for the same WAN transport type.
Below table shows examples on the modified values and how quickly the path characteristics are calculated:
BFD Hello Interval 1000 msec (Default) 1000 msec (Default) 1000 msec (Default)
Path characteristics 120 sec intervals using 30 sec intervals using 120 sec intervals using
calculated every last 12 minutes poll data last 1minute poll data 10 minutes poll data
Policies
Policies are an important part of the Cisco SD-WAN solution and are used to influence the overlay topology and
to influence the flow of data traffic across the WAN Edge devices in the overlay network. Policies are applied
either at control plane or data plane level, configured in the vManage and enforced either on vSmart controllers
or on the WAN Edge devices.
Application Aware Routing policy is part of the centralized policy associated to the vSmart controller. The
vSmart controller would distribute the policy through the OMP protocol to the appropriate WAN Edge devices in
the overlay network that are part of the selected Site Lists, VPN lists and enforced on the WAN Edge devices for
the selected traffic traversing from service-side (LAN network) to the WAN-transport (WAN network).
The WAN Edge device can also be associated with other localized policy. Similar to the centralized policy, only
one localized policy can be applied to WAN Edge device. The localized policy can contain several different
policies such as prefix-lists, access control list policy, route policy, community-lists, QoS etc.
With multiple policies being configured and enforced on the WAN Edge device, it is important to understand the
order of precedence when the packet is moving from service-side to WAN transport-side.
It is possible for a centralized data policy to overwrite the actions of a centralized application-aware routing
policy. Careful consideration must be kept in mind when defining multiple policies for the network as mutually
exclusive policies can influence the traffic traversing the WAN Edge device.
Procedure 1. Verify the WAN Edge device is successfully onboarded in the SD-WAN overlay network.
Step 1. In vManage, navigate to Dashboard > Main Dashboard, make sure WAN Edge devices are
successfully onboarded.
Step 2. In vManage, navigate to Monitor > Network, select the device from the WAN-Edge list.
select System Status from the left panel to view the device status.
Procedure 4. Verify the BFD sessions are established between WAN Edge devices.
Step 1. In vManage, navigate to Monitor > Network, select the device from the WAN-Edge list and select Real
Time option from the left panel. Search for BFD sessions option in the Device Options search bar.
Following the WAN Edge onboarding, the device establishes secure IPSec data tunnels with other WAN Edge
devices and periodically probes the secure tunnels, with BFD Hello packets, to measure the transport tunnel
liveliness and path characteristics.
Below table lists the default values for the BFD polling data defined globally for all tunnel on the WAN Edge
device:
Hello interval Interval at which BFD Hello Interval are send across the WAN 1000 msec
transport
(configurable)
BFD Multiplier Value defines number of BFD Hello Packet Intervals the device 7
waits before declaring that tunnel has failed.
(configurable)
and the default values used to calculate the network path characteristics:
Poll Interval Interval at which WAN Edge calculates the average loss, latency 10 minutes
and jitter for each WAN transport.
(configurable)
The above default value can be changed either effecting all the tunnels on the WAN Edge device or on each
tunnel basis on the WAN Edge device. Careful consideration must be taken when changing the BFD values as
this may cause adverse effect on the performance of the WAN Edge device.
Below procedure walks through steps on the procedure to change the default BFD values.
Step 1. Create BFD template for the WAN Edge device(s) in the SD-WAN network.
In vManage, navigate to Configuration > Templates > Feature and click Add Template. Select all the
appropriate devices deployed in the SD-WAN network from the Select Devices list and choose BFD template
from the Basic Information category.
Alternatively, to influence values on the tunnel/color basis, select the New Color from the Color section, choose
the appropriate Color from the drop-down menu option and modify the Hello Interval and Multiplier values.
Click Add in the Color section and Save at the bottom of the page to save the BFD feature template.
In vManage, navigate to vManage > Configuration > Templates > Device, select the devices from the device
list and click the three dots (…), located at the end of each table row and choose the Edit option from the drop-
down options
Step 4. Click Next, Configure Devices to configure the WAN Edge(s) with custom BFD template.
Step 2. Select Application from the left panel and click New Application List
name the custom application list and select either Application or Application Family option.
Choose appropriate application family category option from the drop-down option and Click Add.
Input the Data Prefix List Name and select either IPv4 / IPv6 from the Internet Protocol option and add the
network prefix that needs to be matched and select Add
To view the tunnel characteristics across the SD-WAN infrastructure. Navigate to vManage > Dashboard >
Main Dashboard > Transport Health widget.
Step 2. Expand the widget by clicking the square icon in the top right corner of the Transport Health widget
and select the Type option to view the chart By Loss, By Latency or by Jitter values and view the transport
health over the maximum of past 7 days.
Input the SLA Class List Name and Loss, Latency, Jitter value requirement for the application family and click
Add.
Tech tip
Note that any number of SLA Class can be created, but only 4 SLA class can be associated to the Application Aware
Routing policy. Please refer to the Software Release Notes for the latest supported number of SLAs for the version in use.
Input the VPN List Name and Add VPN. click Add.
Step 2. Click Add Policy > Create New to create a new Application Aware Routing Policy.
Step 3. select the traffic by matching on the options available in the Match tab.
The match criteria matches the data traffic originating from the service side. Below table lists the different
options available to match on:
DNS Application List Pre-defined / custom-defined list DNS application list is used to split DNS lookup per the
selected application-lists.
PLP High / Low options Pre-configured traffic part of the Packet Loss Priority
(PLP) queue, configured part of Policer QoS section
policy, can be matched. By default, packets have a PLP
value of low. To set the PLP value to high, apply
a policer that includes the exceed remark option.
Protocol Protocol number Traffic with defined protocol number can be matched
Source Data Prefix custom-defined data prefix Pre-defined custom data-prefix of the traffic can be
matched on
Source Port Port number Data traffic with defined port number can be matched.
Destination Data Prefix custom-defined data prefix Pre-defined custom data-prefix of the traffic can be
matched on
Destination Port Port number Data traffic with defined port number can be matched.
For this guide, lets pick the previously created custom Application Family list.
Note that multiple match statements can be configured within the same sequence to select more specific traffic,
as shown below
Backup SLA Preferred Color choose predefined color(s) that traffic is forwarded if the SLA is not met
Log If enabled, syslog message is generated first time a packet flow is logged and
every 5 minutes thereafter, as long as the flow is active. ‘show log’ can be
leveraged to view the log.
SLA Class List Select the custom/pre-defined SLA Class List and choose preferred color(s)
that traffic gets forwarded as long as the specified SLA is satisfied.
App Route policy actions statement allows us to define Preferred Color list as long as the SLA class is satisfied
and action if SLA is breached, either to use the Backup SLA Preferred Color or drop the traffic by enabling the
Strict option.
Step 5. Multiple App-Route rule can be defined part of the same policy, each rule is recognized by different
sequence number, each rule containing a match-action pair defining the preferred treatment for the classified
traffic. The data-traffic matching a sequence rule (executed from low to high sequence number), executes the
appropriate action and exists the policy.
If no policy matches the traffic, Default Action rule is applied for the traffic. The WAN Edge behavior for Default
Action is to perform load-balance the traffic across all available WAN transports
Modify the default behavior to redirect the data traffic to a preferred WAN transport that meets the specified
SLA class list.
select the SLA Class List and select the appropriate SLA Class from the drop-down menu. Click Save Match
and Actions option.
Tech tip
If no WAN transports satisfies the selected SLA class in the default action, the WAN Edge device will load-balance the data
traffic across all the available links.
Step 8. Create any additional Application-Aware Routing policies as shown in previous steps if necessary.
Step 9. Final step in configuring the Application Aware Routing Policy is to choose the Site List and VPN list for
the policy to be associated with
In the Centralized Policy wizard, select Next to navigate to Apply Policies to Sites and VPNS and select
Application-Aware Routing tab.
Input Policy Name and Policy Description and under the previously created App-Aware policy section, click
the New Site List and VPN List option.
Configuring the Application Aware Routing policy does not push the policy to vSmart controller. Final step is to
configure the vSmart controller and enforce the policy by activating the policy.
Pop-up window will ask for confirmation to push the configuration to all vSmart controllers for enforcement.
Click Activate
Deployments with an existing active centralized policy can add the Application-Aware Routing policy to the
existing policy. This process walks through procedure needed to append the Application-Aware routing policy
Creating Application Aware Routing policies consists of defining the three core components:
Step 2. Select Application from the left panel and click New Application List
Select the Application or Application Family and choose appropriate option from the drop-down option.
Provide an Application List Name and Click Add.
To define the data prefix. In vManage, navigate to Configuration > Policies > Centralized Policy. Click Custom
Options from the top right menu options and select Lists from the Centralized Policy section.
select the Data Prefix option from the List type on the left side panel, and click New Data Prefix List
Input the Data Prefix List Name and select either IPv4 / IPv6 from the Internet Protocol option and add the
prefix that needs to be matched and select Add
Step 2. select the SLA Class option from the List type on the left side panel, and click New SLA Class List
Step 3. Create additional SLA Class as shown in the previous step, if needed.
Tech tip
Any number of SLA Class can be created, but only 4 SLA class can be associated to the Application Aware Policy. Please
refer to the corresponding version Release Notes for the latest supports number of SLAs.
Step 2. Select Site from the left panel and create Site list by clicking New Site List
Input the Site List Name and Add Site. click Add
Step 2. Select VPN from the left panel and create VPN list by clicking New VPN List
Input the VPN List Name and Add VPN. click Add.
This procedure walks through the steps needed to create Application-Aware Routing policy. The policy binds
the previously created traffic class to the specified WAN Edge device transport tunnel that satisfies the selected
SLA class requirements.
Step 1. In vManage, navigate to Configuration > Policies > Centralized Policy. Click Custom Options and
select Traffic Policy from the Centralized Policy section.
Step 2. Select the Application Aware Routing tab and select Add Policy > Create New
Step 4. Select the match statement options to match the application/traffic set by clicking on the options
available.
Data traffic originating from the service side can be classified and matched. Below table points on to different
possible options to match on:
Cloud SaaS Application List Pre-defined list Leverage the pre-defined Cloud-Saas Application
list
DNS Application List Pre-defined / custom-defined list DNS application list is used to split DNS lookup per
the selected application-lists.
PLP High / Low options Pre-configured traffic part of the Packet Loss Priority
(PLP) queue, configured part of Policer QoS section
policy, can be matched. By default, packets have a
PLP value of low. To set the PLP value to high, apply
a policer that includes the exceed remark option.
Source Data Prefix custom-defined data prefix Pre-defined custom data-prefix of the traffic can be
matched on
Source Port Port number Data traffic with defined port number can be
matched.
Destination Data Prefix custom-defined data prefix Pre-defined custom data-prefix of the traffic can be
matched on
Destination Port Port number Data traffic with defined port number can be matched.
In this guide, we would pick the previously created custom Application Family list.
Multiple match statements can be configured within the same sequence to select more specific traffic, as
shown below:
Backup SLA Preferred Color choose predefined color(s) that traffic is forwarded if the SLA is not met
Log If enabled, syslog message is generated first time a packet flow is logged and
every 5 minutes thereafter, as long as the flow is active. ‘show log’ can be
leveraged to view the log.
SLA Class List Select the custom/pre-defined SLA Class List and choose preferred color(s)
that traffic gets forwarded as long as the specified SLA is satisfied.
Select SLA class List, Backup SLA Preferred Color option from the Actions. Select the previously created SLA
Class, Preferred Color and Backup SLA Preferred Color from the drop-down menu.
Multiple App-Route rule can be defined part of the same policy, each rule recognized by sequence number,
each rule containing a match-action pair defining the preferred treatment for the classified traffic. When the
data-traffic matches a rule (executed from low to high sequence number), the appropriate action is applied for
the classified traffic. If no policy matches the traffic, Default Action rule is applied for the traffic.
WAN Edge behavior for Default Action is to perform load-balance the traffic across all available WAN transports.
Step 6. (optional) Changing the Default Action behavior.
Modify the default behavior to redirect data traffic to the WAN transports that meets the selected SLA Class List
in the Actions tab as shown below.
Click the Default Action from the Sequence Type section and click Edit option.
Tech tip
If none of the WAN transports satisfies the selected SLA class in the default action, the WAN Edge device will load-balance
the data traffic across all the available links.
Step 2. Input the Policy Name and Description for the new policy and click Copy
Step 3. Confirm that the newly created Policy is inactive (column Activated: false status) and then select the
three dots (...) against the policy and choose Edit to add App-Aware routing policy.
Select the policy created from the drop-down menu and click Import.
Step 6. Select the Policy Application and select the Application-Aware Routing tab.
Configuring and associating the Application Aware Routing policy does not push the policy to vSmart controller
or enforce the policy in the SD-WAN environment. Final step is to activate the policy, which provisions the
vSmart controller and enforces the policy.
In vManage, navigate to Configuration > Policies > Centralized Policy. Choose the newly configured
centralized policy from the list, on the far right-side select the three dots (…) and select Activate option from
the menu.
the new policy configuration is pushed to vSmart and the configuration is activated for enforcement.
Operate
With Application-Aware routing deployed and activated, this section covers steps to manage, monitor and
troubleshoot the various component in the SD-WAN environment using vManage GUI.
To confirm the centralized policy that contains Application-Aware Routing policy is activated, navigate to
vManage > Configuration > Policies > Centralized Policy and the Activated section should be true.
To view the Application-Aware Routing Policy, navigate to vManage > Configuration > Policies. Click the
Custom Options on the top right corner and select the Centralized Policy > Traffic Policy.
Click the three dots (…) to the right of each table row against the policy and select View from the options
Step 3. To verify the Application-Aware routing policy configuration on the vSmart controller
Step 4. View the Application-Aware routing policy configured on the WAN Edge device.
To view the Application-Aware Routing Policy configured on the WAN Edge device, navigate to vManage >
Tools > SSH Terminal. Select the WAN Edge device and issue show sdwan policy from-vsmart on IOS-XE
SD-WAN platform
To view the SLA configured, navigate to vManage, Tools > SSH Terminal. Select the WAN Edge device and
issue sh sdwan app-route sla-class on IOS-XE SD-WAN platform sh app-route sla-class on vipteal platform.
To view the tunnel characteristics across the SD-WAN infrastructure. Navigate to vManage > Dashboard >
Main Dashboard > Transport Health widget.
Step 3. Expand the widget by clicking the square icon in the top right corner of the Transport Health widget
and select the Type option to view the chart By Loss, By Latency or by Jitter values and view the transport
health over the maximum of past 7 days.
To view the tunnel characteristics on the WAN Edge device, navigate to vManage, Monitor > Network > WAN -
Edge. select the WAN Edge device.
Step 2. click WAN > TLOC from the left panel options for the device.
Select the Chart options and choose Loss percentage or Latency/Jitter option. By default, the value is shown
for 24hours, but can be changed.
To view the tunnel characteristics on the WAN Edge device, navigate to vManage, Monitor > Network > WAN -
Edge. click on the WAN Edge device.
Step 2. click WAN > Tunnel from the left panel options for the device.
Select the Tunnel Endpoints from the list and view the path characteristics – Jitter, Loss, Latency for all
transports (mpls, public-internet).
The chart can be changed to view the Latency/Jitter path characteristics by selecting the appropriate option
form the Chart Options option as shown below.
Application Aware statistics can be monitored for each WAN transport in the SD-WAN environment from
vManage GUI.
Step 1. Monitor the Application-Aware Routing across the SD-WAN environment.
To view the tunnel characteristics across the SD-WAN infrastructure, navigate to vManage >Dashboard > Main
Dashboard > Application-Aware Routing widget.
Step 2. expand the widget by clicking the square icon in the top right corner of the Application-Aware Routing
widget and select the Chart options to view the Loss Percentage, Latency, Jitter as far as past 7 days on the
WAN transports.
To view the tunnel characteristics across the SD-WAN infrastructure, navigate to vManage > Dashboard > Main
Dashboard > Application-Aware Routing widget.
Expand the widget by clicking the square icon in the top right corner of the Application-Aware Routing widget
and select the Chart options to view the Loss Percentage, Latency, Jitter. Search the WAN Edge device in the
search options and select the transports to view in the chart
To view the App-route statistics, navigate to vManage > Tools > SSH Terminal. Select the WAN Edge device
and issue show sdwan app-route stats on IOS-XE SD-WAN platform and show app-route stats on vipteal
platform.
Application-Aware routing specific events related to SLA changes, BFD events, and App-route events can be
monitored across the SD-WAN environment from vManage GUI.
Step 2. click Events from the left panel options for the device.
Step 3. Click Filter > Filter By > Component. choose BFD, App-route from the drop-down options and click
Search.
vManage has a very useful tool for network administrators to simulate traffic on the Service side on the Viptela
platform and view the traffic path taken on the WAN Edge to confirm the desired policy enforcement.
Step 1. Simulate the traffic flow from service side to remote branch service site with App-Aware routing
enabled.
In vManage, navigate to Monitor > Network. Select the WAN Edge device and click the Troubleshooting option
from the left-side panel. Click Traffic section > Simulate Flows