Virtual nodes
Important
End of support notice: On September 30, 2026, AWS will discontinue support for AWS App Mesh. After September 30, 2026, you will no longer be able to access the AWS App Mesh console or AWS App Mesh resources. For more information, visit this blog post Migrating from AWS App Mesh to Amazon ECS Service Connect
A virtual node acts as a logical pointer to a particular task group, such as an Amazon ECS service or a Kubernetes deployment. When you create a virtual node, you must specify a service discovery method for your task group. Any inbound traffic that your virtual node expects is specified as a listener. Any virtual service that a virtual node sends outbound traffic to is specified as a backend.
The response metadata for your new virtual node contains the Amazon Resource Name (ARN) that is associated with the
virtual node. Set this value as the APPMESH_RESOURCE_ARN
environment variable for your task
group's Envoy proxy container in your Amazon ECS task definition or Kubernetes pod spec. For example, the value
could be
arn:aws:appmesh:
.
This is then mapped to the us-west-2
:111122223333
:mesh/myMesh
/virtualNode/myVirtualNode
node.id
and node.cluster
Envoy parameters. You must be
using 1.15.0
or later of the Envoy image when setting this variable. For more information
about App Mesh Envoy variables, see Envoy image.
Note
By default, App Mesh uses the name of the resource you specified in APPMESH_RESOURCE_ARN
when Envoy is referring to itself in metrics and traces. You can override this behavior by setting the
APPMESH_RESOURCE_CLUSTER
environment variable with your own name.
Creating a virtual node
- AWS Management Console
-
To create a virtual node using the AWS Management Console
-
Open the App Mesh console at https://console.aws.amazon.com/appmesh/
. -
Choose the mesh that you want to create the virtual node in. All of the meshes that you own and that have been shared with you are listed.
-
Choose Virtual nodes in the left navigation.
-
Choose Create virtual node and then specify settings for your virtual node.
-
For Virtual node name, enter a name for your virtual node.
-
For Service discovery method, choose one of the following options:
-
DNS – Specify the DNS hostname of the actual service that the virtual node represents. The Envoy proxy is deployed in an Amazon VPC. The proxy sends name resolution requests to the DNS server that is configured for the VPC. If the hostname resolves, the DNS server returns one or more IP addresses. For more information about VPC DNS settings, see Using DNS with your VPC. For DNS response type (optional), specify the types of endpoints returned by the DNS resolver. Load Balancer means that the DNS resolver returns a loadbalanced set of endpoints. Endpoints means that the DNS resolver is returning all the endpoints. By default, the response type is assumed to be Load Balancer.
Note
If you use Route53, you'll need to use Load Balancer.
-
AWS Cloud Map – Specify an existing Service name and HTTP Namespace. Optionally, you can also specify attributes that App Mesh can query AWS Cloud Map for by selecting Add row and specifying a Key and Value. Only instances that match all of the specified key/value pairs will be returned. To use AWS Cloud Map, your account must have the
AWSServiceRoleForAppMesh
service-linked role. For more information about AWS Cloud Map, see the AWS Cloud Map Developer Guide. -
None – Select if your virtual node doesn't expect any inbound traffic.
-
-
IP version preference
Control which IP version should be used for traffic within the mesh by toggling on Override default IP version behavior. By default, App Mesh uses a variety of IP versions.
Note
Setting the IP preference on the virtual node only overrides the IP preference set for the mesh on this specific node.
Default
-
Envoy's DNS resolver prefers
IPv6
and falls back toIPv4
. -
We use the
IPv4
address returned by AWS Cloud Map if available and falls back to using theIPv6
address. -
The endpoint created for the local app uses an
IPv4
address. -
The Envoy listeners bind to all
IPv4
addresses.
-
IPv6 preferred
-
Envoy's DNS resolver prefers
IPv6
and falls back toIPv4
. -
The
IPv6
address returned by AWS Cloud Map is used if available and falls back to using theIPv4
address -
The endpoint created for the local app uses an
IPv6
address. -
The Envoy listeners bind to all
IPv4
andIPv6
addresses.
-
IPv4 preferred
-
Envoy's DNS resolver prefers
IPv4
and falls back toIPv6
. -
We use the
IPv4
address returned by AWS Cloud Map if available and falls back to using theIPv6
address. -
The endpoint created for the local app uses an
IPv4
address. -
The Envoy listeners bind to all
IPv4
andIPv6
addresses.
-
IPv6 only
-
Envoy's DNS resolver only uses
IPv6
. -
Only the
IPv6
address returned by AWS Cloud Map is used. If AWS Cloud Map returns anIPv4
address, no IP addresses are used and empty results are returned to the Envoy. -
The endpoint created for the local app uses an
IPv6
address. -
The Envoy listeners bind to all
IPv4
andIPv6
addresses.
-
IPv4 only
-
Envoy's DNS resolver only uses
IPv4
. -
Only the
IPv4
address returned by AWS Cloud Map is used. If AWS Cloud Map returns anIPv6
address, no IP addresses are used and empty results are returned to the Envoy. -
The endpoint created for the local app uses an
IPv4
address. -
The Envoy listeners bind to all
IPv4
andIPv6
addresses.
-
(Optional) Client policy defaults – Configure default requirements when communicating to backend virtual services.
Note
-
If you want to enable Transport Layer Security (TLS) for an existing virtual node, then we recommend that you create a new virtual node, which represents the same service as the existing virtual node, on which to enable TLS. Then gradually shift traffic to the new virtual node using a virtual router and route. For more information about creating a route and adjusting weights for the transition, see Routes. If you update an existing, traffic-serving virtual node with TLS, there is a chance that the downstream client Envoy proxies will receive TLS validation context before the Envoy proxy for the virtual node that you have updated receives the certificate. This can cause TLS negotiation errors on the downstream Envoy proxies.
-
Proxy authorization must be enabled for the Envoy proxy deployed with the application represented by the backend service's virtual nodes. We recommend that when you enable proxy authorization, you restrict access to only the virtual nodes that this virtual node is communicating with.
-
(Optional) Select Enforce TLS if you want to require the virtual node to communicate with all backends using Transport Layer Security (TLS).
-
(Optional) If you only want to require the use of TLS for one or more specific ports, then enter a number in Ports. To add additional ports, select Add port. If you don't specify any ports, TLS is enforced for all ports.
-
For Validation method, select one of the following options. The certificate that you specify must already exist and meet specific requirements. For more information, see Certificate requirements.
-
AWS Private Certificate Authority hosting – Select one or more existing Certificates. For a complete, end-to-end walk through of deploying a mesh with a sample application using encryption with an ACM certificate, see Configuring TLS with AWS Certificate Manager
on GitHub. -
Envoy Secret Discovery Service (SDS) hosting – Enter the name of the secret Envoy will fetch using the Secret Discovery Service.
-
Local file hosting – Specify the path to the Certificate chain file on the file system where Envoy is deployed. For a complete, end-to-end walk through of deploying a mesh with a sample application using encryption with local files, see Configuring TLS with File Provided TLS Certificates
on GitHub.
-
-
(Optional) Enter a Subject Alternative Name. To add additional SANs, select Add SAN. SANs must be FQDN or URI formatted.
-
(Optional) Select Provide client certificate and one of the options below to provide a client certificate when a server requests it and enable mutual TLS authentication . To learn more about mutual TLS, see the App Mesh Mutual TLS Authentication docs.
-
Envoy Secret Discovery Service (SDS) hosting – Enter the name of the secret Envoy will fetch using the Secret Discovery Service.
-
Local file hosting – Specify the path to the Certificate chain file, as well as the Private key, on the file system where Envoy is deployed.
-
-
(Optional) Service backends – Specify the App Mesh virtual service that the virtual node will communicate with.
-
Enter an App Mesh virtual service name or full Amazon Resource Name (ARN) for the virtual service that your virtual node communicates with.
-
(Optional) If you want to set unique TLS settings for a backend, select TLS settings and then select Override defaults.
-
(Optional) Select Enforce TLS if you want to require the virtual node to communicate with all backends using TLS.
-
(Optional) If you only want to require the use of TLS for one or more specific ports, then enter a number in Ports. To add additional ports, select Add port. If you don't specify any ports, TLS is enforced for all ports.
-
For Validation method, select one of the following options. The certificate that you specify must already exist and meet specific requirements. For more information, see Certificate requirements.
-
AWS Private Certificate Authority hosting – Select one or more existing Certificates.
-
Envoy Secret Discovery Service (SDS) hosting – Enter the name of the secret Envoy will fetch using the Secret Discovery Service.
-
Local file hosting – Specify the path to the Certificate chain file on the file system where Envoy is deployed.
-
-
(Optional) Enter a Subject Alternative Name. To add additional SANs, select Add SAN. SANs must be FQDN or URI formatted.
-
(Optional) Select Provide client certificate and one of the options below to provide a client certificate when a server requests it and enable mutual TLS authentication. To learn more about mutual TLS, see the App Mesh Mutual TLS Authentication docs.
-
Envoy Secret Discovery Service (SDS) hosting – Enter the name of the secret Envoy will fetch using the Secret Discovery Service.
-
Local file hosting – Specify the path to the Certificate chain file, as well as the Private key, on the file system where Envoy is deployed.
-
-
To add additional backends, select Add backend.
-
-
(Optional) Logging
To configure logging, enter the HTTP access logs path that you want Envoy to use. We recommend the
/dev/stdout
path so that you can use Docker log drivers to export your Envoy logs to a service such as Amazon CloudWatch Logs.Note
Logs must still be ingested by an agent in your application and sent to a destination. This file path only instructs Envoy where to send the logs.
-
Listener configuration
Listeners support
HTTP
,HTTP/2
,GRPC
, andTCP
protocols.HTTPS
is not supported.-
If your virtual node expects inbound traffic, specify a Port and Protocol for the Listener. The http listener permits connection transition to websockets. You can click Add Listener to add multiple listeners. The Remove button will remove that listener.
(Optional) Enable connection pool
Connection pooling limits the number of connections that an Envoy can concurrently establish with the local application cluster. It is intended to protect your local application from being overwhelmed with connections and lets you adjust traffic shaping for the needs of your applications.
You can configure destination-side connection pool settings for a virtual node listener. App Mesh sets the client-side connection pool settings to infinite by default, simplifying mesh configuration.
Note
The connectionPool and portMapping protocols must be the same. If your listener protocol is tcp, specify maxConnections only. If your listener protocol is grpc or http2, specify maxRequests only. If your listener protocol is http, you can specify both maxConnections and maxPendingRequests.
-
For Maximum connections, specify the maximum number of outbound connections.
-
(Optional) For Maximum pending requests, specify the number of overflowing requests after Maximum connections that an Envoy will queue. The default value is
2147483647
.
-
(Optional) Enable outlier detection
Outlier detection applied at the client Envoy allows clients to take near-immediate action on connections with observed known bad failures. It is a form of a circuit breaker implementation that tracks the health status of individual hosts in the upstream service.
Outlier detection dynamically determines whether endpoints in an upstream cluster are performing unlike the others and removes them from the healthy load balancing set.
Note
To effectively configure outlier detection for a server Virtual Node, the service discovery method of that Virtual Node can be either AWS Cloud Map or DNS with the response type field set to
ENDPOINTS
. If you use DNS service discovery method with response type asLOADBALANCER
, the Envoy proxy would only elect a single IP address for routing to the upstream service. This nullifies the outlier detection behavior of ejecting an unhealthy host from a set of hosts. Refer to the Service discovery method section for more details on the Envoy proxy's behavior in relation to the service discovery type.-
For Server errors, specify the number of consecutive 5xx errors required for ejection.
-
For Outlier detection interval, specify the time interval and unit between ejection sweep analysis.
-
For Base ejection duration, specify the base amount of time and unit for which a host is ejected.
-
For Ejection percentage, specify the maximum percentage of hosts in the load balancing pool that can be ejected.
-
(Optional) Enable health check – Configure settings for a health check policy.
A health check policy is optional, but if you specify any values for a health policy, then you must specify values for Healthy threshold, Health check interval, Health check protocol, Timeout period, and Unhealthy threshold.
-
For Health check protocol, choose a protocol. If you select grpc, then your service must conform to the GRPC Health Checking Protocol
. -
For Health check port, specify the port that the health check should run on.
-
For Healthy threshold, specify the number of consecutive successful health checks that must occur before declaring the listener healthy.
-
For Health check interval, specify the time period in milliseconds between each health check execution.
-
For Path, specify the destination path for the health check request. This value is only used if the Health check protocol is
http
orhttp2
. The value is ignored for other protocols. -
For Timeout period, specify the amount of time to wait when receiving a response from the health check, in milliseconds.
-
For Unhealthy threshold, specify the number of consecutive failed health checks that must occur before declaring the listener unhealthy.
-
(Optional) Enable TLS termination – Configure how other virtual nodes communicate with this virtual node using TLS.
-
For Mode, select the mode you want TLS to be configured for on the listener.
-
For Certificate method, select one of the following options. The certificate must meet specific requirements. For more information, see Certificate requirements.
-
AWS Certificate Manager hosting – Select an existing Certificate.
-
Envoy Secret Discovery Service (SDS) hosting – Enter the name of the secret Envoy will fetch using the Secret Discovery Service.
-
Local file hosting – Specify the path to the Certificate chain file, as well as the Private key, on the file system where the Envoy proxy is deployed.
-
-
(Optional) Select Require client certificates and one of the options below to enable mutual TLS authentication when a client provides a certificate. To learn more about mutual TLS, see the App Mesh Mutual TLS Authentication docs.
-
Envoy Secret Discovery Service (SDS) hosting – Enter the name of the secret Envoy will fetch using the Secret Discovery Service.
-
Local file hosting – Specify the path to the Certificate chain file on the file system where Envoy is deployed.
-
-
(Optional) Enter a Subject Alternative Name. To add additional SANs, select Add SAN. SANs must be FQDN or URI formatted.
-
(Optional) Timeouts
Note
If you specify a timeout greater than the default, make sure to set up a virtual router and a route with a timeout greater than the default. However, if you decrease the timeout to a value that is lower than the default, it's optional to update the timeouts at Route. For more information, see Routes.
-
Request timeout – You can specify a request timeout if you selected grpc, http, or http2 for the listener's Protocol. The default is 15 seconds. A value of
0
disables the timeout. -
Idle duration – You can specify an idle duration for any listener protocol. The default is 300 seconds.
-
-
-
Choose Create virtual node to finish.
-
- AWS CLI
-
To create a virtual node using the AWS CLI.
Create a virtual node that uses DNS for service discovery using the following command and an input JSON file (replace the
red
values with your own):-
aws appmesh create-virtual-node \ --cli-input-json file://
create-virtual-node-dns
.json -
Contents of example create-virtual-node-dns.json:
{ "meshName": "meshName", "spec": { "listeners": [ { "portMapping": { "port": 80, "protocol": "http" } } ], "serviceDiscovery": { "dns": { "hostname": "serviceBv1.svc.cluster.local" } } }, "virtualNodeName": "nodeName" }
-
Example output:
{ "virtualNode": { "meshName": "meshName", "metadata": { "arn": "arn:aws:appmesh:
us-west-2
:210987654321:mesh/meshName/virtualNode/nodeName", "createdAt": "2022-04-06T09:12:24.348000-05:00", "lastUpdatedAt": "2022-04-06T09:12:24.348000-05:00", "meshOwner": "123456789012", "resourceOwner": "210987654321", "uid": "a1b2c3d4-5678-90ab-cdef-11111EXAMPLE", "version": 1 }, "spec": { "listeners": [ { "portMapping": { "port": 80, "protocol": "http" } } ], "serviceDiscovery": { "dns": { "hostname": "serviceBv1.svc.cluster.local" } } }, "status": { "status": "ACTIVE" }, "virtualNodeName": "nodeName" } }
For more information on creating a virtual node with the AWS CLI for App Mesh, see the create-virtual-node command in the AWS CLI reference.
-
Deleting a virtual node
Note
You can't delete a virtual node if it is specified as a target in any route or as a provider in any virtual service.
- AWS Management Console
-
To delete a virtual node using the AWS Management Console
-
Open the App Mesh console at https://console.aws.amazon.com/appmesh/
. -
Choose the mesh that you want to delete a virtual node from. All of the meshes that you own and that have been shared with you are listed.
-
Choose Virtual nodes in the left navigation.
-
In the Virtual Nodes table, choose the virtual node that you want to delete and select Delete. To delete a virtual node, your account ID must be listed in either the Mesh owner or the Resource owner columns of the virtual node.
-
In the confirmation box, type
delete
and then select Delete.
-
- AWS CLI
-
To delete a virtual node using the AWS CLI
-
Use the following command to delete your virtual node (replace the
red
values with your own):aws appmesh delete-virtual-node \ --mesh-name
meshName
\ --virtual-node-namenodeName
-
Example output:
{ "virtualNode": { "meshName": "meshName", "metadata": { "arn": "arn:aws:appmesh:
us-west-2
:210987654321:mesh/meshName/virtualNode/nodeName", "createdAt": "2022-04-06T09:12:24.348000-05:00", "lastUpdatedAt": "2022-04-07T11:03:48.120000-05:00", "meshOwner": "123456789012", "resourceOwner": "210987654321", "uid": "a1b2c3d4-5678-90ab-cdef-11111EXAMPLE", "version": 2 }, "spec": { "backends": [], "listeners": [ { "portMapping": { "port": 80, "protocol": "http" } } ], "serviceDiscovery": { "dns": { "hostname": "serviceBv1.svc.cluster.local" } } }, "status": { "status": "DELETED" }, "virtualNodeName": "nodeName" } }
For more information on deleting a virtual node with the AWS CLI for App Mesh, see the delete-virtual-node command in the AWS CLI reference.
-