ExpressRoute monitoring, metrics, and alerts

This article helps you understand ExpressRoute monitoring, metrics, and alerts using Azure Monitor. Azure Monitor is one stop shop for all metrics, alerting, diagnostic logs across all of Azure.

Note

Using Classic Metrics is not recommended.

ExpressRoute metrics

To view Metrics, go to the Azure Monitor page and select Metrics. To view ExpressRoute metrics, filter by Resource Type ExpressRoute circuits. To view Global Reach metrics, filter by Resource Type ExpressRoute circuits and select an ExpressRoute circuit resource that has Global Reach enabled. To view ExpressRoute Direct metrics, filter Resource Type by ExpressRoute Ports.

Once a metric is selected, the default aggregation is applied. Optionally, you can apply splitting, which shows the metric with different dimensions.

Important

When viewing ExpressRoute metrics in the Azure portal, select a time granularity of 5 minutes or greater for best possible results.

Screenshot of time granularity options.

Aggregation Types:

Metrics explorer supports sum, maximum, minimum, average and count as aggregation types. You should use the recommended Aggregation type when reviewing the insights for each ExpressRoute metric.

  • Sum: The sum of all values captured during the aggregation interval.
  • Count: The number of measurements captured during the aggregation interval.
  • Average: The average of the metric values captured during the aggregation interval.
  • Min: The smallest value captured during the aggregation interval.
  • Max: The largest value captured during the aggregation interval.

ExpressRoute circuit

Metric Category Unit Aggregation Type Description Dimensions Exportable via Diagnostic Settings?
ARP Availability Availability Percent Average ARP Availability from MSEE towards all peers. Peering Type, Peer Yes
BGP Availability Availability Percent Average BGP Availability from MSEE towards all peers. Peering Type, Peer Yes
BitsInPerSecond Traffic BitsPerSecond Average Bits ingressing Azure per second Peering Type Yes
BitsOutPerSecond Traffic BitsPerSecond Average Bits egressing Azure per second Peering Type Yes
DroppedInBitsPerSecond Traffic BitsPerSecond Average Ingress bits of data dropped per second Peering Type Yes
DroppedOutBitsPerSecond Traffic BitPerSecond Average Egress bits of data dropped per second Peering Type Yes

ExpressRoute gateways

Metric Category Unit Aggregation Type Description Dimensions Exportable via Diagnostic Settings?
Bits received per second Performance BitsPerSecond Average Total bits received on ExpressRoute gateway per second roleInstance Yes
CPU utilization Performance Count Average CPU Utilization of the ExpressRoute Gateway roleInstance Yes
Packets per second Performance CountPerSecond Average Total Packets received on ExpressRoute Gateway per second roleInstance Yes
Count of routes advertised to peer Availability Count Maximum Count Of Routes Advertised To Peer by ExpressRouteGateway roleInstance Yes
Count of routes learned from peer Availability Count Maximum Count Of Routes Learned From Peer by ExpressRouteGateway roleInstance Yes
Frequency of routes changed Availability Count Total Frequency of Routes change in ExpressRoute Gateway roleInstance Yes
Number of VMs in virtual network Availability Count Maximum Estimated number of VMs in the virtual network No Dimensions Yes
Active flows Scalability Count Average Number of active flows on ExpressRoute Gateway roleInstance Yes
Max flows created per second Scalability FlowsPerSecond Maximum Maximum number of flows created per second on ExpressRoute Gateway roleInstance, direction Yes

ExpressRoute Gateway connections

Metric Category Unit Aggregation Type Description Dimensions Exportable via Diagnostic Settings?
BitsInPerSecond Traffic BitsPerSecond Average Bits ingressing Azure per second through ExpressRoute gateway ConnectionName Yes
BitsOutPerSecond Traffic BitsPerSecond Average Bits egressing Azure per second through ExpressRoute gateway ConnectionName Yes

ExpressRoute Direct

Metric Category Unit Aggregation Type Description Dimensions Exportable via Diagnostic Settings?
BitsInPerSecond Traffic BitsPerSecond Average Bits ingressing Azure per second Link Yes
BitsOutPerSecond Traffic BitsPerSecond Average Bits egressing Azure per second Link Yes
DroppedInBitsPerSecond Traffic BitsPerSecond Average Ingress bits of data dropped per second Link Yes
DroppedOutBitsPerSecond Traffic BitPerSecond Average Egress bits of data dropped per second Link Yes
AdminState Physical Connectivity Count Average Admin state of the port Link Yes
LineProtocol Physical Connectivity Count Average Line protocol status of the port Link Yes
RxLightLevel Physical Connectivity Count Average Rx Light level in dBm Link, Lane Yes
TxLightLevel Physical Connectivity Count Average Tx light level in dBm Link, Lane Yes

Circuits metrics

Bits In and Out - Metrics across all peerings

Aggregation type: Avg

You can view metrics across all peerings on a given ExpressRoute circuit.

circuit metrics

Bits In and Out - Metrics per peering

Aggregation type: Avg

You can view metrics for private, public, and Microsoft peering in bits/second.

metrics per peering

BGP Availability - Split by Peer

Aggregation type: Avg

You can view near to real-time availability of BGP (Layer-3 connectivity) across peerings and peers (Primary and Secondary ExpressRoute routers). This dashboard shows the Primary BGP session status is up for private peering and the Second BGP session status is down for private peering.

BGP availability per peer

Note

During maintenance between the Azure edge and core network, BGP availability will appear down even if the BGP session between the customer edge and Azure edge remains up. For information about maintenance between the Azure edge and core network, make sure to have your maintenance alerts turned on and configured.

ARP Availability - Split by Peering

Aggregation type: Avg

You can view near to real-time availability of ARP (Layer-2 connectivity) across peerings and peers (Primary and Secondary ExpressRoute routers). This dashboard shows the Private Peering ARP session status is up across both peers, but down for Microsoft peering for both peers. The default aggregation (Average) was utilized across both peers.

ARP availability per peer

ExpressRoute Direct Metrics

Aggregation type: Avg

You can view the Admin state for each link of the ExpressRoute Direct port pair. The Admin state represents if the physical port is on or off. This state is required to pass traffic across the ExpressRoute Direct connection.

ER Direct admin state

Aggregation type: Avg

You can view the bits in per second across both links of the ExpressRoute Direct port pair. Monitor this dashboard to compare inbound bandwidth for both links.

ER Direct bits in per second

Aggregation type: Avg

You can also view the bits out per second across both links of the ExpressRoute Direct port pair. Monitor this dashboard to compare outbound bandwidth for both links.

ER Direct bits out per second

Aggregation type: Avg

You can view the line protocol across each link of the ExpressRoute Direct port pair. The Line Protocol indicates if the physical link is up and running over ExpressRoute Direct. Monitor this dashboard and set alerts to know when the physical connection goes down.

ER Direct line protocol

Aggregation type: Avg

You can view the Rx light level (the light level that the ExpressRoute Direct port is receiving) for each port. Healthy Rx light levels generally fall within a range of -10 dBm to 0 dBm. Set alerts to be notified if the Rx light level falls outside of the healthy range.

ER Direct line Rx Light Level

Note

ExpressRoute Direct connectivity is hosted across different device platforms. Some ExpressRoute Direct connections will support a split view for Rx light levels by lane. However, this is not supported on all deployments.

Aggregation type: Avg

You can view the Tx light level (the light level that the ExpressRoute Direct port is transmitting) for each port. Healthy Tx light levels generally fall within a range of -10 dBm to 0 dBm. Set alerts to be notified if the Tx light level falls outside of the healthy range.

ER Direct line Tx Light Level

Note

ExpressRoute Direct connectivity is hosted across different device platforms. Some ExpressRoute Direct connections will support a split view for Tx light levels by lane. However, this is not supported on all deployments.

ExpressRoute Virtual Network Gateway Metrics

Aggregation type: Avg

When you deploy an ExpressRoute gateway, Azure manages the compute and functions of your gateway. There are six gateway metrics available to you to better understand the performance of your gateway:

  • Bits received per second
  • CPU Utilization
  • Packets per seconds
  • Count of routes advertised to peers
  • Count of routes learned from peers
  • Frequency of routes changed
  • Number of VMs in the virtual network
  • Active flows
  • Max flows created per second

We highly recommended you set alerts for each of these metrics so that you're aware of when your gateway could be seeing performance issues.

Bits received per second - Split by instance

Aggregation type: Avg

This metric captures inbound bandwidth utilization on the ExpressRoute virtual network gateway instances. Set an alert for how frequent the bandwidth utilization exceeds a certain threshold. If you need more bandwidth, increase the size of the ExpressRoute virtual network gateway.

Screenshot of inbound bit per second - split metrics.

CPU Utilization - Split by instance

Aggregation type: Avg

You can view the CPU utilization of each gateway instance. The CPU utilization might spike briefly during routine host maintenance but prolong high CPU utilization could indicate your gateway is reaching a performance bottleneck. Increasing the size of the ExpressRoute gateway might resolve this issue. Set an alert for how frequent the CPU utilization exceeds a certain threshold.

Screenshot of CPU utilization - split metrics.

Packets Per Second - Split by instance

Aggregation type: Avg

This metric captures the number of inbound packets traversing the ExpressRoute gateway. You should expect to see a consistent stream of data here if your gateway is receiving traffic from your on-premises network. Set an alert for when the number of packets per second drops below a threshold indicating that your gateway is no longer receiving traffic.

Screenshot of packets per second - split metrics.

Count of Routes Advertised to Peer - Split by instance

Aggregation type: Max

This metric shows the number of routes the ExpressRoute gateway is advertising to the circuit. The address spaces might include virtual networks that are connected using virtual network peering and uses remote ExpressRoute gateway. You should expect the number of routes to remain consistent unless there are frequent changes to the virtual network address spaces. Set an alert for when the number of advertised routes drop below the threshold for the number of virtual network address spaces you're aware of.

Screenshot of count of routes advertised to peer.

Count of routes learned from peer - Split by instance

Aggregation type: Max

This metric shows the number of routes the ExpressRoute gateway is learning from peers connected to the ExpressRoute circuit. These routes can be either from another virtual network connected to the same circuit or learned from on-premises. Set an alert for when the number of learned routes drop below a certain threshold. This metric can indicate either the gateway is seeing a performance problem or remote peers are no longer advertising routes to the ExpressRoute circuit.

Screenshot of count of routes learned from peer.

Frequency of routes change - Split by instance

Aggregation type: Sum

This metric shows the frequency of routes being learned from or advertised to remote peers. You should first investigate your on-premises devices to understand why the network is changing so frequently. A high frequency in routes change could indicate a performance problem on the ExpressRoute gateway where scaling the gateway SKU up might resolve the problem. Set an alert for a frequency threshold to be aware of when your ExpressRoute gateway is seeing abnormal route changes.

Screenshot of frequency of routes changed metric.

Number of VMs in the virtual network

Aggregation type: Max

This metric shows the number of virtual machines that are using the ExpressRoute gateway. The number of virtual machines might include VMs from peered virtual networks that use the same ExpressRoute gateway. Set an alert for this metric if the number of VMs goes above a certain threshold that could affect the gateway performance.

Screenshot of number of virtual machines in the virtual network metric.

Note

To maintain reliability of the service, Azure often performs platform or OS maintenance on the gateway service. During this time, this metric may fluctuate and report inaccurately.

Active flows

Aggregation type: Avg

Split by: Gateway Instance

This metric displays a count of the total number of active flows on the ExpressRoute Gateway. Only inbound traffic from on-premises is captured for active flows. Through split at instance level, you can see active flow count per gateway instance. For more information, see understand network flow limits.

Screenshot of number of active flows per second metrics dashboard.

Max flows created per second

Aggregation type: Max

Split by: Gateway Instance and Direction (Inbound/Outbound)

This metric displays the maximum number of flows created per second on the ExpressRoute Gateway. Through split at instance level and direction, you can see max flow creation rate per gateway instance and inbound/outbound direction respectively. For more information, see understand network flow limits.

Screenshot of the maximum number of flows created per second metrics dashboard.

ExpressRoute gateway connections in bits/seconds

Aggregation type: Avg

This metric shows the bits per second for ingress and egress to Azure through the ExpressRoute gateway. You can split this metric further to see specific connections to the ExpressRoute circuit.

Screenshot of gateway connection bandwidth usage metric.

Alerts for ExpressRoute gateway connections

  1. To configure alerts, navigate to Azure Monitor, then select Alerts.

    Screenshot of the alerts option from the monitor overview page.

  2. Select + Create > Alert rule and select the ExpressRoute gateway connection resource. Select Next: Condition > to configure the signal.

    Screenshot of the selecting ExpressRoute virtual network gateway from the select a resource page.

  3. On the Select a signal page, select a metric, resource health, or activity log that you want to be alerted. Depending on the signal you select, you might need to enter additional information such as a threshold value. You can also combine multiple signals into a single alert. Select Next: Actions > to define who and how they get notify.

    Screenshot of list of signals that can be alerted for ExpressRoute gateways.

  4. Select + Select action groups to choose an existing action group you previously created or select + Create action group to define a new one. In the action group, you determine how notifications get sent and who receives them.

    Screenshot of add action groups page.

  5. Select Review + create and then Create to deploy the alert into your subscription.

Alerts based on each peering

After you select a metric, certain metric allow you to set up dimensions based on peering or a specific peer (virtual networks).

Screenshot of an alert rule based on ExpressRoute peering setup.

Configure alerts for activity logs on circuits

When selecting signals to be alerted on, you can select Activity Log signal type.

Screenshot of activity log signals from the select a signal page.

More metrics in Log Analytics

You can also view ExpressRoute metrics by going to your ExpressRoute circuit resource and selecting the Logs tab. For any metrics you query, the output contains the following columns.

Column Type Description
TimeGrain string PT1M (metric values are pushed every minute)
Count real Usually is 2 (each MSEE pushes a single metric value every minute)
Minimum real The minimum of the two metric values pushed by the two MSEEs
Maximum real The maximum of the two metric values pushed by the two MSEEs
Average real Equal to (Minimum + Maximum)/2
Total real Sum of the two metric values from both MSEEs (the main value to focus on for the metric queried)

Next steps

Set up your ExpressRoute connection.