MCPRoute
aigateway.envoyproxy.io / v1beta1
apiVersion: aigateway.envoyproxy.io/v1beta1
kind: MCPRoute
metadata:
name: example
apiVersion
string
APIVersion defines the versioned schema of this representation of an object.
Servers should convert recognized schemas to the latest internal value, and
may reject unrecognized values.
More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#resources
kind
string
Kind is a string value representing the REST resource this object represents.
Servers may infer this from the endpoint the client submits requests to.
Cannot be updated.
In CamelCase.
More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds
metadata
object
spec object
Spec defines the details of the MCPRoute.
backendRefs []object required
BackendRefs is a list of backend references to the MCP servers.
These MCP servers will be aggregated and exposed as a single MCP endpoint to the clients.
From the client's perspective, they only need to configure a single MCP server URL, e.g. "https://api.example.com/mcp",
and the Envoy AI Gateway will route the requests to the appropriate MCP server based on the requests.
All names must be unique within this list to avoid potential tools, resources, etc. name collisions.
Also, cross-namespace references are not supported. In other words, the backend MCP servers must be in the
same namespace as the MCPRoute.
minItems:
1maxItems:
256forwardHeaders []object
ForwardHeaders specifies HTTP headers to extract from the incoming client request
and forward to this backend MCP server.
This enables per-user authentication passthrough (e.g., personal access tokens)
without requiring OAuth configuration.
Each entry specifies a header name to extract and an optional rename for the backend.
maxItems:
32
backendHeader
string
BackendHeader is the header name to use when forwarding to the backend.
If not specified, the original header name is used.
minLength:
1
name
string required
Name is the header name to extract from the incoming client request.
minLength:
1
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
path
string
Path is the HTTP endpoint path of the backend MCP server.
If not specified, the default is "/mcp".
maxLength:
1024
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535securityPolicy object
SecurityPolicy is the security policy to apply to this MCP server.
apiKey object
APIKey is a mechanism to access a backend. The API key will be injected into the request headers.
header
string
Header is the HTTP header to inject the API key into. If not specified,
defaults to "Authorization".
When the header is "Authorization", the injected header value will be
prefixed with "Bearer ".
Either one of Header or QueryParam can be specified to inject the API key.
minLength:
1
inline
string
Inline contains the API key as an inline string.
queryParam
string
QueryParam is the HTTP query parameter to inject the API key into.
For example, if QueryParam is set to "api_key", and the API key is "mysecretkey", the request URL will be modified to include
"?api_key=mysecretkey".
Either one of Header or QueryParam can be specified to inject the API key.
Note: Embedding credentials in URLs (including query parameters) is generally not recommended because URLs can be exposed in logs
and intermediary systems; prefer header-based injection when possible.
minLength:
1secretRef object
secretRef is the Kubernetes secret which contains the API keys.
The key of the secret should be "apiKey".
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is kind of the referent. For example "Secret".
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the referenced object. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63toolSelector object
ToolSelector filters the tools exposed by this MCP server.
Supports exact matches and RE2-compatible regular expressions for both include and exclude patterns.
If not specified, all tools from the MCP server are exposed.
exclude
[]string
Exclude is a list of tool names to exclude. The specified tools will not be available.
Exclude rules take precedence over include rules.
maxItems:
32
excludeRegex
[]string
ExcludeRegex is a list of RE2-compatible regular expressions that, when matched, exclude the tool.
Tools matching these patterns will not be available. Exclude rules take precedence over include rules.
maxItems:
32
include
[]string
Include is a list of tool names to include. Only the specified tools will be available.
maxItems:
32
includeRegex
[]string
IncludeRegex is a list of RE2-compatible regular expressions that, when matched, include the tool.
Only tools matching these patterns will be available.
maxItems:
32headers []object
Headers are HTTP headers that must match for this route to be selected.
Multiple match values are ANDed together, meaning, a request must match all the specified headers to select the route.
maxItems:
16
name
string required
Name is the name of the HTTP Header to be matched. Name matching MUST be
case-insensitive. (See https://tools.ietf.org/html/rfc7230#section-3.2).
If multiple entries specify equivalent header names, only the first
entry with an equivalent name MUST be considered for a match. Subsequent
entries with an equivalent header name MUST be ignored. Due to the
case-insensitivity of header names, "foo" and "Foo" are considered
equivalent.
When a header is repeated in an HTTP request, it is
implementation-specific behavior as to how this is represented.
Generally, proxies should follow the guidance from the RFC:
https://www.rfc-editor.org/rfc/rfc7230.html#section-3.2.2 regarding
processing a repeated header, with special handling for "Set-Cookie".
pattern:
^[A-Za-z0-9!#$%&'*+\-.^_\x60|~]+$minLength:
1maxLength:
256
type
string
Type specifies how to match against the value of the header.
Support: Core (Exact)
Support: Implementation-specific (RegularExpression)
Since RegularExpression HeaderMatchType has implementation-specific
conformance, implementations can support POSIX, PCRE or any other dialects
of regular expressions. Please read the implementation's documentation to
determine the supported dialect.
enum:
Exact, RegularExpression
value
string required
Value is the value of HTTP Header to be matched.
<gateway:experimental:description>
Must consist of printable US-ASCII characters, optionally separated
by single tabs or spaces. See: https://tools.ietf.org/html/rfc7230#section-3.2
</gateway:experimental:description>
<gateway:experimental:validation:Pattern=`^[!-~]+([\t ]?[!-~]+)*$`>
minLength:
1maxLength:
4096parentRefs []object required
ParentRefs are the names of the Gateway resources this MCPRoute is being attached to.
Cross namespace references are not supported. In other words, the Gateway resources must be in the
same namespace as the MCPRoute. Currently, each reference's Kind must be Gateway.
minItems:
1maxItems:
16
group
string
Group is the group of the referent.
When unspecified, "gateway.networking.k8s.io" is inferred.
To set the core API group (such as for a "Service" kind referent),
Group must be explicitly set to "" (empty string).
Support: Core
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is kind of the referent.
There are two kinds of parent resources with "Core" support:
* Gateway (Gateway conformance profile)
* Service (Mesh conformance profile, ClusterIP Services only)
Support for other resources is Implementation-Specific.
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
Support: Core
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the referent. When unspecified, this refers
to the local namespace of the Route.
Note that there are specific rules for ParentRefs which cross namespace
boundaries. Cross-namespace references are only valid if they are explicitly
allowed by something in the namespace they are referring to. For example:
Gateway has the AllowedRoutes field, and ReferenceGrant provides a
generic way to enable any other kind of cross-namespace reference.
<gateway:experimental:description>
ParentRefs from a Route to a Service in the same namespace are "producer"
routes, which apply default routing rules to inbound connections from
any namespace to the Service.
ParentRefs from a Route to a Service in a different namespace are
"consumer" routes, and these routing rules are only applied to outbound
connections originating from the same namespace as the Route, for which
the intended destination of the connections are a Service targeted as a
ParentRef of the Route.
</gateway:experimental:description>
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port is the network port this Route targets. It can be interpreted
differently based on the type of parent resource.
When the parent resource is a Gateway, this targets all listeners
listening on the specified port that also support this kind of Route(and
select this Route). It's not recommended to set `Port` unless the
networking behaviors specified in a Route must apply to a specific port
as opposed to a listener(s) whose port(s) may be changed. When both Port
and SectionName are specified, the name and port of the selected listener
must match both specified values.
<gateway:experimental:description>
When the parent resource is a Service, this targets a specific port in the
Service spec. When both Port (experimental) and SectionName are specified,
the name and port of the selected port must match both specified values.
</gateway:experimental:description>
Implementations MAY choose to support other parent resources.
Implementations supporting other types of parent resources MUST clearly
document how/if Port is interpreted.
For the purpose of status, an attachment is considered successful as
long as the parent resource accepts it partially. For example, Gateway
listeners can restrict which Routes can attach to them by Route kind,
namespace, or hostname. If 1 of 2 Gateway listeners accept attachment
from the referencing Route, the Route MUST be considered successfully
attached. If no Gateway listeners accept attachment from this Route,
the Route MUST be considered detached from the Gateway.
Support: Extended
format:
int32minimum:
1maximum:
65535
sectionName
string
SectionName is the name of a section within the target resource. In the
following resources, SectionName is interpreted as the following:
* Gateway: Listener name. When both Port (experimental) and SectionName
are specified, the name and port of the selected listener must match
both specified values.
* Service: Port name. When both Port (experimental) and SectionName
are specified, the name and port of the selected listener must match
both specified values.
Implementations MAY choose to support attaching Routes to other resources.
If that is the case, they MUST clearly document how SectionName is
interpreted.
When unspecified (empty string), this will reference the entire resource.
For the purpose of status, an attachment is considered successful if at
least one section in the parent resource accepts it. For example, Gateway
listeners can restrict which Routes can attach to them by Route kind,
namespace, or hostname. If 1 of 2 Gateway listeners accept attachment from
the referencing Route, the Route MUST be considered successfully
attached. If no Gateway listeners accept attachment from this Route, the
Route MUST be considered detached from the Gateway.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$minLength:
1maxLength:
253
path
string
Path is the HTTP endpoint path that serves MCP requests over the Streamable HTTP transport.
If not specified, the default is "/mcp".
maxLength:
1024securityPolicy object
SecurityPolicy defines the security policy for this MCPRoute.
apiKeyAuth object
APIKeyAuth defines the configuration for the API Key Authentication.
credentialRefs []object required
CredentialRefs is the Kubernetes secret which contains the API keys.
This is an Opaque secret.
Each API key is stored in the key representing the client id.
If the secrets have a key for a duplicated client, the first one will be used.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is kind of the referent. For example "Secret".
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the referenced object. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63extractFrom []object required
ExtractFrom is where to fetch the key from the coming request.
The value from the first source that has a key will be used.
cookies
[]string
Cookies is the names of the cookie to fetch the key from.
If multiple cookies are specified, envoy will look for the api key in the order of the list.
This field is optional, but only one of headers, params or cookies is supposed to be specified.
headers
[]string
Headers is the names of the header to fetch the key from.
If multiple headers are specified, envoy will look for the api key in the order of the list.
This field is optional, but only one of headers, params or cookies is supposed to be specified.
params
[]string
Params is the names of the query parameter to fetch the key from.
If multiple params are specified, envoy will look for the api key in the order of the list.
This field is optional, but only one of headers, params or cookies is supposed to be specified.
forwardClientIDHeader
string
ForwardClientIDHeader is the name of the header to forward the client identity to the backend
service. The header will be added to the request with the client id as the value.
sanitize
boolean
Sanitize indicates whether to remove the API key from the request before forwarding it to the backend service.
authorization object
Authorization defines the configuration for the MCP spec compatible authorization.
defaultAction
string
DefaultAction is the action to take when no rules match. If unspecified, defaults to Deny.
enum:
Allow, Denyrules []object
Rules defines a list of authorization rules.
These rules are evaluated in order, the first matching rule will be applied,
and the rest will be skipped.
If no rules are defined, the default action will be applied to all requests.
maxItems:
32
action
string
Action is the authorization decision for matching requests. If unspecified, defaults to Allow.
enum:
Allow, Deny
cel
string
CEL specifies a Common Expression Language (CEL) expression evaluated for this rule.
The expression must return a boolean; evaluation errors or non-boolean results
are treated as "no match".
Example CEL expressions:
* `request.method == "POST"`
* `request.headers["x-custom-header"] == "AllowedValue"`
* `request.mcp.tool in ["toolA", "toolB"]`
Available attributes in the CEL expression:
* request.method: HTTP method such as GET or POST. Type: string.
* request.headers: map of headers with lowercased keys, first value only. Type: map[string]string.
* request.headers_all: map of headers with lowercased keys, all values. Type: map[string][]string.
* request.path: request path such as /mcp. Type: string.
* request.auth.jwt.claims: JWT claims when a bearer JWT is present. Type: map[string]any.
* request.auth.jwt.scopes: JWT scopes when a bearer JWT is present. Type: []string.
* request.mcp.method: MCP method such as tools/list or tools/call. Type: string.
* request.mcp.backend: upstream backend name (for example, "kiwi" or "github"). Type: string.
* request.mcp.tool: tool name without backend prefix (for example, "list_issues"). Type: string.
* request.mcp.params: parameters of the MCP method, including keys like "_meta" and "arguments". Type: object.
Note: The CEL expression support is experimental, and the attributes
available to the expression may change in future releases.
maxLength:
4096source object
Source defines the authorization source for this rule.
If not specified, the rule will match all sources.
jwt object required
JWT defines the JWT scopes required for this rule to match.
claims []object
Claims defines the list of JWT claims required for the rule. Each claim must exist on the token
and have at least one of the expected values. Use to enforce tenant or subject-based access.
maxItems:
16
name
string required
Name is the name of the claim.
If it is a nested claim, use a dot (.) separated string as the name to
represent the full path to the claim.
For example, if the claim is in the "department" field in the "organization" field,
the name should be "organization.department".
minLength:
1maxLength:
253
valueType
string
ValueType is the type of the claim value.
Only String and StringArray types are supported for now.
enum:
String, StringArray
values
[]string required
Values are the values that the claim must match.
If the claim is a string type, the specified value must match exactly.
If the claim is a string array type, the specified value must match one of the values in the array.
If multiple values are specified, one of the values must match for the rule to match.
minItems:
1maxItems:
128
scopes
[]string
Scopes defines the list of JWT scopes required for the rule.
If multiple scopes are specified, all scopes must be present in the JWT for the rule to match.
maxItems:
16target object
Target defines the authorization target for this rule.
If not specified, the rule will match all targets.
tools []object required
Tools defines the list of tools this rule applies to.
minItems:
1maxItems:
16
backend
string required
Backend is the name of the backend this tool belongs to.
tool
string required
Tool is the name of the tool.
extAuth object
ExtAuth defines the configuration for External Authorization.
bodyToExtAuth object
BodyToExtAuth defines the Body to Ext Auth configuration.
maxRequestBytes
integer required
MaxRequestBytes is the maximum size of a message body that the filter will hold in memory.
Envoy will return HTTP 413 and will not initiate the authorization process when buffer
reaches the number set in this field.
Note that this setting will have precedence over failOpen mode.
format:
int32minimum:
1contextExtensions []object
ContextExtensions are analogous to http_request.headers, however these
contents will not be sent to the upstream server. This provides an
extension mechanism for sending additional information to the auth server
without modifying the proto definition. It maps to the internal opaque
context in the filter chain.
name
string required
Name of the context extension.
type
string required
Type is the type of method to use to read the ContextExtension value.
Valid values are Value and ValueRef, default is Value.
enum:
Value, ValueRef
value
string
Value of the context extension.
valueRef object
ValueRef for the context extension's value.
group
string required
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
key
string required
The key to select.
kind
string required
Kind is kind of the referent. For example "HTTPRoute" or "Service".
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
failOpen
boolean
FailOpen is a switch used to control the behavior when a response from the External Authorization service cannot be obtained.
If FailOpen is set to true, the system allows the traffic to pass through.
Otherwise, if it is set to false or not set (defaulting to false),
the system blocks the traffic and returns a HTTP 5xx error, reflecting a fail-closed approach.
This setting determines whether to prioritize accessibility over strict security in case of authorization service failure.
If set to true, the External Authorization will also be bypassed if its configuration is invalid.
grpc object
GRPC defines the gRPC External Authorization service.
Either GRPCService or HTTPService must be specified,
and only one of them can be provided.
backendRef object
BackendRef references a Kubernetes object that represents the
backend server to which the authorization request will be sent.
Deprecated: Use BackendRefs instead.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535backendRefs []object
BackendRefs references a Kubernetes object that represents the
backend server to which the authorization request will be sent.
maxItems:
16
fallback
boolean
Fallback indicates whether the backend is designated as a fallback.
Multiple fallback backends can be configured.
It is highly recommended to configure active or passive health checks to ensure that failover can be detected
when the active backends become unhealthy and to automatically readjust once the primary backends are healthy again.
The overprovisioning factor is set to 1.4, meaning the fallback backends will only start receiving traffic when
the health of the active backends falls below 72%.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535
weight
integer
Weight specifies the proportion of requests forwarded to the referenced
backend. This is computed as weight/(sum of all weights in this
BackendRefs list). For non-zero values, there may be some epsilon from
the exact proportion defined here depending on the precision an
implementation supports. Weight is not a percentage and the sum of
weights does not need to equal 100.
If only one backend is specified and it has a weight greater than 0, 100%
of the traffic is forwarded to that backend. If weight is set to 0, no
traffic should be forwarded for this entry. If unspecified, weight
defaults to 1.
Support for this field varies based on the context where used.
format:
int32minimum:
0maximum:
1e+06backendSettings object
BackendSettings holds configuration for managing the connection
to the backend.
circuitBreaker object
Circuit Breaker settings for the upstream connections and requests.
If not set, circuit breakers will be enabled with the default thresholds
maxConnections
integer
The maximum number of connections that Envoy will establish to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxParallelRequests
integer
The maximum number of parallel requests that Envoy will make to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxParallelRetries
integer
The maximum number of parallel retries that Envoy will make to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxPendingRequests
integer
The maximum number of pending requests that Envoy will queue to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxRequestsPerConnection
integer
The maximum number of requests that Envoy will make over a single connection to the referenced backend defined within a xRoute rule.
Default: unlimited.
format:
int64minimum:
0maximum:
4.294967295e+09perEndpoint object
PerEndpoint defines Circuit Breakers that will apply per-endpoint for an upstream cluster
maxConnections
integer
MaxConnections configures the maximum number of connections that Envoy will establish per-endpoint to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09retryBudget object
RetryBudget provides settings for retry budget, which limits the number of retries in a given percentage.
RetryBudget take precedence over maxParallelRetries.
minRetryConcurrency
integer
MinRetryConcurrency specifies the minimum retry concurrency allowed for the retry budget.
For example, a budget of 20% with a minimum retry concurrency of 3
will allow 5 active retries while there are 25 active requests.
If there are 2 active requests, there are still 3 active retries
allowed because of the minimum retry concurrency.
Defaults to 3.
format:
int32percent object required
Percent specifies the limit on concurrent retries as a percentage [0, 100] of
the sum of active requests and active pending requests.
denominator
integer
format:
int32minimum:
1
numerator
integer required
format:
int32minimum:
0connection object
Connection includes backend connection settings.
bufferLimit
string | integer
BufferLimit Soft limit on size of the cluster’s connections read and write buffers.
BufferLimit applies to connection streaming (maybe non-streaming) channel between processes, it's in user space.
If unspecified, an implementation defined default is applied (32768 bytes).
For example, 20Mi, 1Gi, 256Ki etc.
Note: that when the suffix is not provided, the value is interpreted as bytes.
preconnect object
Preconnect configures proactive upstream connections to reduce latency by establishing
connections before they’re needed and avoiding connection establishment overhead.
If unset, Envoy will fetch connections as needed to serve in-flight requests.
perEndpointPercent
integer
PerEndpointPercent configures how many additional connections to maintain per
upstream endpoint, useful for high-QPS or latency sensitive services. Expressed as a
percentage of the connections required by active streams
(e.g. 100 = preconnect disabled, 105 = 1.05x connections per-endpoint, 200 = 2.00×).
Allowed value range is between 100-300. When both PerEndpointPercent and
PredictivePercent are set, Envoy ensures both are satisfied (max of the two).
format:
int32minimum:
100maximum:
300
predictivePercent
integer
PredictivePercent configures how many additional connections to maintain
across the cluster by anticipating which upstream endpoint the load balancer
will select next, useful for low-QPS services. Relies on deterministic
loadbalancing and is only supported with Random or RoundRobin.
Expressed as a percentage of the connections required by active streams
(e.g. 100 = 1.0 (no preconnect), 105 = 1.05× connections across the cluster, 200 = 2.00×).
Minimum allowed value is 100. When both PerEndpointPercent and PredictivePercent are
set Envoy ensures both are satisfied per host (max of the two).
format:
int32minimum:
100
socketBufferLimit
string | integer
SocketBufferLimit provides configuration for the maximum buffer size in bytes for each socket
to backend.
SocketBufferLimit applies to socket streaming channel between TCP/IP stacks, it's in kernel space.
For example, 20Mi, 1Gi, 256Ki etc.
Note that when the suffix is not provided, the value is interpreted as bytes.
dns object
DNS includes dns resolution settings.
dnsRefreshRate
string
DNSRefreshRate specifies the rate at which DNS records should be refreshed.
Defaults to 30 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
lookupFamily
string
LookupFamily determines how Envoy would resolve DNS for Routes where the backend is specified as a fully qualified domain name (FQDN).
If set, this configuration overrides other defaults.
enum:
IPv4, IPv6, IPv4Preferred, IPv6Preferred, IPv4AndIPv6
respectDnsTtl
boolean
RespectDNSTTL indicates whether the DNS Time-To-Live (TTL) should be respected.
If the value is set to true, the DNS refresh rate will be set to the resource record’s TTL.
Defaults to true.
healthCheck object
HealthCheck allows gateway to perform active health checking on backends.
active object
Active health check configuration
grpc object
GRPC defines the configuration of the GRPC health checker.
It's optional, and can only be used if the specified type is GRPC.
service
string
Service to send in the health check request.
If this is not specified, then the health check request applies to the entire
server and not to a specific service.
healthyThreshold
integer
HealthyThreshold defines the number of healthy health checks required before a backend host is marked healthy.
format:
int32minimum:
1http object
HTTP defines the configuration of http health checker.
It's required while the health checker type is HTTP.
expectedResponse object
ExpectedResponse defines a list of HTTP expected responses to match.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
expectedStatuses
[]integer
ExpectedStatuses defines a list of HTTP response statuses considered healthy.
Defaults to 200 only
hostname
string
Hostname defines the HTTP Host header used for active HTTP health checks.
Host selection uses this order: this field, the associated Backend endpoint
hostname if available, then the effective Route hostname.
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$minLength:
1maxLength:
253
method
string
Method defines the HTTP method used for health checking.
Defaults to GET
path
string required
Path defines the HTTP path that will be requested during health checking.
minLength:
1maxLength:
1024
retriableStatuses
[]integer
RetriableStatuses defines a list of HTTP response statuses considered retriable.
Responses matching these statuses count towards the unhealthy threshold but
do not result in the host being considered immediately unhealthy.
The expected statuses take precedence for any range overlaps with this field.
initialJitter
string
InitialJitter defines the maximum time Envoy will wait before the first health check.
Envoy will randomly select a value between 0 and the initial jitter value.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
Interval defines the time between active health checks.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$overrides object
Overrides defines the configuration of the overriding health check settings for all endpoints
in the backend cluster. This allows customization of port and other settings that may differ
from the main service configuration.
port
integer
Port overrides the health check port.
If not set, the endpoint's serving port is used for health checks.
This is useful when health checks are served on a different port than
the main service port (e.g., port 443 for service, port 9090 for health checks).
format:
int32minimum:
1maximum:
65535tcp object
TCP defines the configuration of tcp health checker.
It's required while the health checker type is TCP.
receive object
Receive defines the expected response payload.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
send object
Send defines the request payload.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
timeout
string
Timeout defines the time to wait for a health check response.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
type
string required
Type defines the type of health checker.
unhealthyThreshold
integer
UnhealthyThreshold defines the number of unhealthy health checks required before a backend host is marked unhealthy.
Without RetriableStatuses configured, any health check failure results in the host being immediately
considered unhealthy. When RetriableStatuses is set, health checks returning those statuses are retried
up to this threshold before the host is marked unhealthy.
format:
int32minimum:
1
panicThreshold
integer
When number of unhealthy endpoints for a backend reaches this threshold
Envoy will disregard health status and balance across all endpoints.
It's designed to prevent a situation in which host failures cascade throughout the cluster
as load increases. If not set, the default value is 50%. To disable panic mode, set value to `0`.
format:
int32minimum:
0maximum:
100passive object
Passive passive check configuration
alwaysEjectOneEndpoint
boolean
AlwaysEjectOneEndpoint defines whether at least one host should be ejected,
regardless of MaxEjectionPercent.
baseEjectionTime
string
BaseEjectionTime defines the base duration for which a host will be ejected on consecutive failures.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
consecutive5XxErrors
integer
Consecutive5xxErrors sets the number of consecutive 5xx errors triggering ejection.
format:
int32
consecutiveGatewayErrors
integer
ConsecutiveGatewayErrors sets the number of consecutive gateway errors triggering ejection.
format:
int32
consecutiveLocalOriginFailures
integer
ConsecutiveLocalOriginFailures sets the number of consecutive local origin failures triggering ejection.
Parameter takes effect only when split_external_local_origin_errors is set to true.
format:
int32
failurePercentageThreshold
integer
FailurePercentageThreshold sets the failure percentage threshold for outlier detection.
If the failure percentage of a given host is greater than or equal to this value, it will be ejected.
Defaults to 85.
format:
int32minimum:
0maximum:
100
interval
string
Interval defines the time between passive health checks.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxEjectionPercent
integer
MaxEjectionPercent sets the maximum percentage of hosts in a cluster that can be ejected.
format:
int32minimum:
0maximum:
100
splitExternalLocalOriginErrors
boolean
SplitExternalLocalOriginErrors enables splitting of errors between external and local origin.
http2 object
HTTP2 provides HTTP/2 configuration for backend connections.
connectionKeepalive object
ConnectionKeepalive configures HTTP/2 connection keepalive using PING frames.
idleInterval
string
IdleInterval specifies how long a connection must be idle before a PING is sent.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
Interval specifies how often to send HTTP/2 PING frames to keep the connection alive.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
intervalJitter
integer
IntervalJitter specifies a random jitter percentage added to each interval.
Defaults to 15% if not specified.
format:
int32minimum:
0maximum:
100
timeout
string
Timeout specifies how long to wait for a PING response before considering the connection dead.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
initialConnectionWindowSize
string | integer
InitialConnectionWindowSize sets the initial window size for HTTP/2 connections.
If not set, the default value is 1 MiB.
initialStreamWindowSize
string | integer
InitialStreamWindowSize sets the initial window size for HTTP/2 streams.
If not set, the default value is 64 KiB(64*1024).
maxConcurrentStreams
integer
MaxConcurrentStreams sets the maximum number of concurrent streams allowed per connection.
If not set, the default value is 100.
format:
int32minimum:
1maximum:
2.147483647e+09
onInvalidMessage
string
OnInvalidMessage determines if Envoy will terminate the connection or just the offending stream in the event of HTTP messaging error
It's recommended for L2 Envoy deployments to set this value to TerminateStream.
https://www.envoyproxy.io/docs/envoy/latest/configuration/best_practices/level_two
Default: TerminateConnection
loadBalancer object
LoadBalancer policy to apply when routing traffic from the gateway to
the backend endpoints. Defaults to `LeastRequest`.
backendUtilization object
BackendUtilization defines the configuration when the load balancer type is
set to BackendUtilization.
blackoutPeriod
string
A given endpoint must report load metrics continuously for at least this long before the endpoint weight will be used.
Default is 10s.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
errorUtilizationPenaltyPercent
integer
ErrorUtilizationPenaltyPercent adjusts endpoint weights based on the error rate (eps/qps).
This is expressed as a percentage-based integer where 100 represents 1.0, 150 represents 1.5, etc.
For example:
- 100 => 1.0x
- 120 => 1.2x
- 200 => 2.0x
Must be non-negative.
format:
int32minimum:
0
keepResponseHeaders
boolean
KeepResponseHeaders keeps the ORCA load report headers/trailers before sending the response to the client.
Defaults to false.
metricNamesForComputingUtilization
[]string
Metric names used to compute utilization if application_utilization is not set.
For map fields in ORCA proto, use the form "<map_field>.<key>", e.g., "named_metrics.foo".
weightExpirationPeriod
string
If a given endpoint has not reported load metrics in this long, stop using the reported weight. Defaults to 3m.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
weightUpdatePeriod
string
How often endpoint weights are recalculated. Values less than 100ms are capped at 100ms. Default 1s.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$consistentHash object
ConsistentHash defines the configuration when the load balancer type is
set to ConsistentHash
cookie object
Cookie configures the cookie hash policy when the consistent hash type is set to Cookie.
attributes
object
Additional Attributes to set for the generated cookie.
name
string required
Name of the cookie to hash.
If this cookie does not exist in the request, Envoy will generate a cookie and set
the TTL on the response back to the client based on Layer 4
attributes of the backend endpoint, to ensure that these future requests
go to the same backend endpoint. Make sure to set the TTL field for this case.
ttl
string
TTL of the generated cookie if the cookie is not present. This value sets the
Max-Age attribute value.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$header object
Header configures the header hash policy when the consistent hash type is set to Header.
Deprecated: use Headers instead
name
string required
Name of the header to hash.
headers []object
Headers configures the header hash policy for each header, when the consistent hash type is set to Headers.
name
string required
Name of the header to hash.
queryParams []object
QueryParams configures the query parameter hash policy when the consistent hash type is set to QueryParams.
name
string required
Name of the query param to hash.
tableSize
integer
The table size for consistent hashing, must be prime number limited to 5000011.
format:
int64minimum:
2maximum:
5.000011e+06
type
string required
ConsistentHashType defines the type of input to hash on. Valid Type values are
"SourceIP",
"Header",
"Headers",
"Cookie".
"QueryParams".
enum:
SourceIP, Header, Headers, Cookie, QueryParamsdynamicModule object
DynamicModule defines the configuration when the load balancer type is
set to DynamicModule. The referenced module must be registered in the
EnvoyProxy resource's dynamicModules allowlist.
config
object
Config is optional configuration for the module's load balancer
implementation. This is serialized and passed to the module's
initialization function.
lbPolicyName
string required
LBPolicyName identifies a specific load balancer implementation within
the dynamic module. A single shared library can contain multiple LB
policy implementations. This value is passed to the module's
initialization function to select the appropriate implementation.
minLength:
1maxLength:
253
name
string required
Name references a dynamic module registered in the EnvoyProxy resource's
dynamicModules list. The referenced module must exist in the registry;
otherwise, the policy will be rejected.
pattern:
^[a-z0-9]([a-z0-9.-]*[a-z0-9])?$minLength:
1maxLength:
253endpointOverride object
EndpointOverride defines the configuration for endpoint override.
When specified, the load balancer will attempt to route requests to endpoints
based on the override information extracted from request headers or metadata.
If the override endpoints are not available, the configured load balancer policy will be used as fallback.
extractFrom []object required
ExtractFrom defines the sources to extract endpoint override information from.
minItems:
1maxItems:
10
header
string
Header defines the header to get the override endpoint addresses.
The header value must specify at least one endpoint in `IP:Port` format or multiple endpoints in `IP:Port,IP:Port,...` format.
For example `10.0.0.5:8080` or `[2600:4040:5204::1574:24ae]:80`.
The IPv6 address is enclosed in square brackets.
slowStart object
SlowStart defines the configuration related to the slow start load balancer policy.
If set, during slow start window, traffic sent to the newly added hosts will gradually increase.
Supported for RoundRobin, LeastRequest, and BackendUtilization load balancers.
window
string required
Window defines the duration of the warm up period for newly added host.
During slow start window, traffic sent to the newly added hosts will gradually increase.
Currently only supports linear growth of traffic. For additional details,
see https://www.envoyproxy.io/docs/envoy/latest/api-v3/config/cluster/v3/cluster.proto#config-cluster-v3-cluster-slowstartconfig
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
type
string required
Type decides the type of Load Balancer policy.
Valid LoadBalancerType values are
"ConsistentHash",
"LeastRequest",
"Random",
"RoundRobin",
"BackendUtilization",
"DynamicModule".
enum:
ConsistentHash, LeastRequest, Random, RoundRobin, BackendUtilization, DynamicModulezoneAware object
ZoneAware defines the configuration related to the distribution of requests between locality zones.
preferLocal object
PreferLocalZone configures zone-aware routing to prefer sending traffic to the local locality zone.
force object
ForceLocalZone defines override configuration for forcing all traffic to stay within the local zone instead of the default behavior
which maintains equal distribution among upstream endpoints while sending as much traffic as possible locally.
minEndpointsInZoneThreshold
integer
MinEndpointsInZoneThreshold is the minimum number of upstream endpoints in the local zone required to honor the forceLocalZone
override. This is useful for protecting zones with fewer endpoints.
format:
int32
minEndpointsThreshold
integer
MinEndpointsThreshold is the minimum number of total upstream endpoints across all zones required to enable zone-aware routing.
format:
int64
percentageEnabled
integer
Configures percentage of requests that will be considered for zone aware routing if zone aware routing is configured. If not specified, Envoy defaults to 100%.
format:
int32minimum:
0maximum:
100weightedZones []object
WeightedZones configures weight-based traffic distribution across locality zones.
Traffic is distributed proportionally based on the sum of all zone weights.
weight
integer required
Weight defines the weight for this locality.
Higher values receive more traffic. The actual traffic distribution
is proportional to this value relative to other localities.
format:
int32
zone
string required
Zone specifies the topology zone this weight applies to.
The value should match the topology.kubernetes.io/zone label
of the nodes where endpoints are running.
Zones not listed in the configuration receive a default weight of 1.
proxyProtocol object
ProxyProtocol enables the Proxy Protocol when communicating with the backend.
version
string required
Version of ProxyProtocol
Valid ProxyProtocolVersion values are
"V1"
"V2"
enum:
V1, V2retry object
Retry provides more advanced usage, allowing users to customize the number of retries, retry fallback strategy, and retry triggering conditions.
If not set, retry will be disabled.
numAttemptsPerPriority
integer
NumAttemptsPerPriority defines the number of requests (initial attempt + retries)
that should be sent to the same priority before switching to a different one.
If not specified or set to 0, all requests are sent to the highest priority that is healthy.
format:
int32
numRetries
integer
NumRetries is the number of retries to be attempted. Defaults to 2.
format:
int32minimum:
0perRetry object
PerRetry is the retry policy to be applied per retry attempt.
backOff object
Backoff is the backoff policy to be applied per retry attempt. gateway uses a fully jittered exponential
back-off algorithm for retries. For additional details,
see https://www.envoyproxy.io/docs/envoy/latest/configuration/http/http_filters/router_filter#config-http-filters-router-x-envoy-max-retries
baseInterval
string
BaseInterval is the base interval between retries.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxInterval
string
MaxInterval is the maximum interval between retries. This parameter is optional, but must be greater than or equal to the base_interval if set.
The default is 10 times the base_interval
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
timeout
string
Timeout is the timeout per retry attempt.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$retryOn object
RetryOn specifies the retry trigger condition.
If not specified, the default is to retry on connect-failure,refused-stream,unavailable,cancelled,retriable-status-codes(503).
httpStatusCodes
[]integer
HttpStatusCodes specifies the http status codes to be retried.
The retriable-status-codes trigger must also be configured for these status codes to trigger a retry.
triggers
[]string
Triggers specifies the retry trigger condition(Http/Grpc).
tcpKeepalive object
TcpKeepalive settings associated with the upstream client connection.
Disabled by default.
idleTime
string
The duration a connection needs to be idle before keep-alive
probes start being sent.
The duration format is
Defaults to `7200s`.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
The duration between keep-alive probes.
Defaults to `75s`.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
probes
integer
The total number of unacknowledged probes to send before deciding
the connection is dead.
Defaults to 9.
format:
int32timeout object
Timeout settings for the backend connections.
http object
Timeout settings for HTTP.
connectionIdleTimeout
string
The idle timeout for an HTTP connection. Idle time is defined as a period in which there are no active requests in the connection.
Default: 1 hour.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxConnectionDuration
string
The maximum duration of an HTTP connection.
Default: unlimited.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxStreamDuration
string
MaxStreamDuration is the maximum duration for a stream to complete. This timeout measures the time
from when the request is sent until the response stream is fully consumed and does not apply to
non-streaming requests.
When set to "0s", no max duration is applied and streams can run indefinitely.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
requestTimeout
string
RequestTimeout is the time until which entire response is received from the upstream.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
streamIdleTimeout
string
The stream idle timeout defines the amount of time a stream can exist without any upstream or downstream activity.
If not specified, StreamIdleTimeout is inherited from the listener-level setting, which can be configured via ClientTrafficPolicy.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$tcp object
Timeout settings for TCP.
connectTimeout
string
The timeout for network connection establishment, including TCP and TLS handshakes.
Default: 10 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
headersToExtAuth
[]string
HeadersToExtAuth defines the client request headers that will be included
in the request to the external authorization service.
Note: If not specified, the default behavior for gRPC and HTTP external
authorization services is different due to backward compatibility reasons.
All headers will be included in the check request to a gRPC authorization server.
Only the following headers will be included in the check request to an HTTP
authorization server: Host, Method, Path, Content-Length, and Authorization.
And these headers will always be included to the check request to an HTTP
authorization server by default, no matter whether they are specified
in HeadersToExtAuth or not.
http object
HTTP defines the HTTP External Authorization service.
Either GRPCService or HTTPService must be specified,
and only one of them can be provided.
backendRef object
BackendRef references a Kubernetes object that represents the
backend server to which the authorization request will be sent.
Deprecated: Use BackendRefs instead.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535backendRefs []object
BackendRefs references a Kubernetes object that represents the
backend server to which the authorization request will be sent.
maxItems:
16
fallback
boolean
Fallback indicates whether the backend is designated as a fallback.
Multiple fallback backends can be configured.
It is highly recommended to configure active or passive health checks to ensure that failover can be detected
when the active backends become unhealthy and to automatically readjust once the primary backends are healthy again.
The overprovisioning factor is set to 1.4, meaning the fallback backends will only start receiving traffic when
the health of the active backends falls below 72%.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535
weight
integer
Weight specifies the proportion of requests forwarded to the referenced
backend. This is computed as weight/(sum of all weights in this
BackendRefs list). For non-zero values, there may be some epsilon from
the exact proportion defined here depending on the precision an
implementation supports. Weight is not a percentage and the sum of
weights does not need to equal 100.
If only one backend is specified and it has a weight greater than 0, 100%
of the traffic is forwarded to that backend. If weight is set to 0, no
traffic should be forwarded for this entry. If unspecified, weight
defaults to 1.
Support for this field varies based on the context where used.
format:
int32minimum:
0maximum:
1e+06backendSettings object
BackendSettings holds configuration for managing the connection
to the backend.
circuitBreaker object
Circuit Breaker settings for the upstream connections and requests.
If not set, circuit breakers will be enabled with the default thresholds
maxConnections
integer
The maximum number of connections that Envoy will establish to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxParallelRequests
integer
The maximum number of parallel requests that Envoy will make to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxParallelRetries
integer
The maximum number of parallel retries that Envoy will make to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxPendingRequests
integer
The maximum number of pending requests that Envoy will queue to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxRequestsPerConnection
integer
The maximum number of requests that Envoy will make over a single connection to the referenced backend defined within a xRoute rule.
Default: unlimited.
format:
int64minimum:
0maximum:
4.294967295e+09perEndpoint object
PerEndpoint defines Circuit Breakers that will apply per-endpoint for an upstream cluster
maxConnections
integer
MaxConnections configures the maximum number of connections that Envoy will establish per-endpoint to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09retryBudget object
RetryBudget provides settings for retry budget, which limits the number of retries in a given percentage.
RetryBudget take precedence over maxParallelRetries.
minRetryConcurrency
integer
MinRetryConcurrency specifies the minimum retry concurrency allowed for the retry budget.
For example, a budget of 20% with a minimum retry concurrency of 3
will allow 5 active retries while there are 25 active requests.
If there are 2 active requests, there are still 3 active retries
allowed because of the minimum retry concurrency.
Defaults to 3.
format:
int32percent object required
Percent specifies the limit on concurrent retries as a percentage [0, 100] of
the sum of active requests and active pending requests.
denominator
integer
format:
int32minimum:
1
numerator
integer required
format:
int32minimum:
0connection object
Connection includes backend connection settings.
bufferLimit
string | integer
BufferLimit Soft limit on size of the cluster’s connections read and write buffers.
BufferLimit applies to connection streaming (maybe non-streaming) channel between processes, it's in user space.
If unspecified, an implementation defined default is applied (32768 bytes).
For example, 20Mi, 1Gi, 256Ki etc.
Note: that when the suffix is not provided, the value is interpreted as bytes.
preconnect object
Preconnect configures proactive upstream connections to reduce latency by establishing
connections before they’re needed and avoiding connection establishment overhead.
If unset, Envoy will fetch connections as needed to serve in-flight requests.
perEndpointPercent
integer
PerEndpointPercent configures how many additional connections to maintain per
upstream endpoint, useful for high-QPS or latency sensitive services. Expressed as a
percentage of the connections required by active streams
(e.g. 100 = preconnect disabled, 105 = 1.05x connections per-endpoint, 200 = 2.00×).
Allowed value range is between 100-300. When both PerEndpointPercent and
PredictivePercent are set, Envoy ensures both are satisfied (max of the two).
format:
int32minimum:
100maximum:
300
predictivePercent
integer
PredictivePercent configures how many additional connections to maintain
across the cluster by anticipating which upstream endpoint the load balancer
will select next, useful for low-QPS services. Relies on deterministic
loadbalancing and is only supported with Random or RoundRobin.
Expressed as a percentage of the connections required by active streams
(e.g. 100 = 1.0 (no preconnect), 105 = 1.05× connections across the cluster, 200 = 2.00×).
Minimum allowed value is 100. When both PerEndpointPercent and PredictivePercent are
set Envoy ensures both are satisfied per host (max of the two).
format:
int32minimum:
100
socketBufferLimit
string | integer
SocketBufferLimit provides configuration for the maximum buffer size in bytes for each socket
to backend.
SocketBufferLimit applies to socket streaming channel between TCP/IP stacks, it's in kernel space.
For example, 20Mi, 1Gi, 256Ki etc.
Note that when the suffix is not provided, the value is interpreted as bytes.
dns object
DNS includes dns resolution settings.
dnsRefreshRate
string
DNSRefreshRate specifies the rate at which DNS records should be refreshed.
Defaults to 30 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
lookupFamily
string
LookupFamily determines how Envoy would resolve DNS for Routes where the backend is specified as a fully qualified domain name (FQDN).
If set, this configuration overrides other defaults.
enum:
IPv4, IPv6, IPv4Preferred, IPv6Preferred, IPv4AndIPv6
respectDnsTtl
boolean
RespectDNSTTL indicates whether the DNS Time-To-Live (TTL) should be respected.
If the value is set to true, the DNS refresh rate will be set to the resource record’s TTL.
Defaults to true.
healthCheck object
HealthCheck allows gateway to perform active health checking on backends.
active object
Active health check configuration
grpc object
GRPC defines the configuration of the GRPC health checker.
It's optional, and can only be used if the specified type is GRPC.
service
string
Service to send in the health check request.
If this is not specified, then the health check request applies to the entire
server and not to a specific service.
healthyThreshold
integer
HealthyThreshold defines the number of healthy health checks required before a backend host is marked healthy.
format:
int32minimum:
1http object
HTTP defines the configuration of http health checker.
It's required while the health checker type is HTTP.
expectedResponse object
ExpectedResponse defines a list of HTTP expected responses to match.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
expectedStatuses
[]integer
ExpectedStatuses defines a list of HTTP response statuses considered healthy.
Defaults to 200 only
hostname
string
Hostname defines the HTTP Host header used for active HTTP health checks.
Host selection uses this order: this field, the associated Backend endpoint
hostname if available, then the effective Route hostname.
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$minLength:
1maxLength:
253
method
string
Method defines the HTTP method used for health checking.
Defaults to GET
path
string required
Path defines the HTTP path that will be requested during health checking.
minLength:
1maxLength:
1024
retriableStatuses
[]integer
RetriableStatuses defines a list of HTTP response statuses considered retriable.
Responses matching these statuses count towards the unhealthy threshold but
do not result in the host being considered immediately unhealthy.
The expected statuses take precedence for any range overlaps with this field.
initialJitter
string
InitialJitter defines the maximum time Envoy will wait before the first health check.
Envoy will randomly select a value between 0 and the initial jitter value.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
Interval defines the time between active health checks.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$overrides object
Overrides defines the configuration of the overriding health check settings for all endpoints
in the backend cluster. This allows customization of port and other settings that may differ
from the main service configuration.
port
integer
Port overrides the health check port.
If not set, the endpoint's serving port is used for health checks.
This is useful when health checks are served on a different port than
the main service port (e.g., port 443 for service, port 9090 for health checks).
format:
int32minimum:
1maximum:
65535tcp object
TCP defines the configuration of tcp health checker.
It's required while the health checker type is TCP.
receive object
Receive defines the expected response payload.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
send object
Send defines the request payload.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
timeout
string
Timeout defines the time to wait for a health check response.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
type
string required
Type defines the type of health checker.
unhealthyThreshold
integer
UnhealthyThreshold defines the number of unhealthy health checks required before a backend host is marked unhealthy.
Without RetriableStatuses configured, any health check failure results in the host being immediately
considered unhealthy. When RetriableStatuses is set, health checks returning those statuses are retried
up to this threshold before the host is marked unhealthy.
format:
int32minimum:
1
panicThreshold
integer
When number of unhealthy endpoints for a backend reaches this threshold
Envoy will disregard health status and balance across all endpoints.
It's designed to prevent a situation in which host failures cascade throughout the cluster
as load increases. If not set, the default value is 50%. To disable panic mode, set value to `0`.
format:
int32minimum:
0maximum:
100passive object
Passive passive check configuration
alwaysEjectOneEndpoint
boolean
AlwaysEjectOneEndpoint defines whether at least one host should be ejected,
regardless of MaxEjectionPercent.
baseEjectionTime
string
BaseEjectionTime defines the base duration for which a host will be ejected on consecutive failures.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
consecutive5XxErrors
integer
Consecutive5xxErrors sets the number of consecutive 5xx errors triggering ejection.
format:
int32
consecutiveGatewayErrors
integer
ConsecutiveGatewayErrors sets the number of consecutive gateway errors triggering ejection.
format:
int32
consecutiveLocalOriginFailures
integer
ConsecutiveLocalOriginFailures sets the number of consecutive local origin failures triggering ejection.
Parameter takes effect only when split_external_local_origin_errors is set to true.
format:
int32
failurePercentageThreshold
integer
FailurePercentageThreshold sets the failure percentage threshold for outlier detection.
If the failure percentage of a given host is greater than or equal to this value, it will be ejected.
Defaults to 85.
format:
int32minimum:
0maximum:
100
interval
string
Interval defines the time between passive health checks.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxEjectionPercent
integer
MaxEjectionPercent sets the maximum percentage of hosts in a cluster that can be ejected.
format:
int32minimum:
0maximum:
100
splitExternalLocalOriginErrors
boolean
SplitExternalLocalOriginErrors enables splitting of errors between external and local origin.
http2 object
HTTP2 provides HTTP/2 configuration for backend connections.
connectionKeepalive object
ConnectionKeepalive configures HTTP/2 connection keepalive using PING frames.
idleInterval
string
IdleInterval specifies how long a connection must be idle before a PING is sent.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
Interval specifies how often to send HTTP/2 PING frames to keep the connection alive.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
intervalJitter
integer
IntervalJitter specifies a random jitter percentage added to each interval.
Defaults to 15% if not specified.
format:
int32minimum:
0maximum:
100
timeout
string
Timeout specifies how long to wait for a PING response before considering the connection dead.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
initialConnectionWindowSize
string | integer
InitialConnectionWindowSize sets the initial window size for HTTP/2 connections.
If not set, the default value is 1 MiB.
initialStreamWindowSize
string | integer
InitialStreamWindowSize sets the initial window size for HTTP/2 streams.
If not set, the default value is 64 KiB(64*1024).
maxConcurrentStreams
integer
MaxConcurrentStreams sets the maximum number of concurrent streams allowed per connection.
If not set, the default value is 100.
format:
int32minimum:
1maximum:
2.147483647e+09
onInvalidMessage
string
OnInvalidMessage determines if Envoy will terminate the connection or just the offending stream in the event of HTTP messaging error
It's recommended for L2 Envoy deployments to set this value to TerminateStream.
https://www.envoyproxy.io/docs/envoy/latest/configuration/best_practices/level_two
Default: TerminateConnection
loadBalancer object
LoadBalancer policy to apply when routing traffic from the gateway to
the backend endpoints. Defaults to `LeastRequest`.
backendUtilization object
BackendUtilization defines the configuration when the load balancer type is
set to BackendUtilization.
blackoutPeriod
string
A given endpoint must report load metrics continuously for at least this long before the endpoint weight will be used.
Default is 10s.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
errorUtilizationPenaltyPercent
integer
ErrorUtilizationPenaltyPercent adjusts endpoint weights based on the error rate (eps/qps).
This is expressed as a percentage-based integer where 100 represents 1.0, 150 represents 1.5, etc.
For example:
- 100 => 1.0x
- 120 => 1.2x
- 200 => 2.0x
Must be non-negative.
format:
int32minimum:
0
keepResponseHeaders
boolean
KeepResponseHeaders keeps the ORCA load report headers/trailers before sending the response to the client.
Defaults to false.
metricNamesForComputingUtilization
[]string
Metric names used to compute utilization if application_utilization is not set.
For map fields in ORCA proto, use the form "<map_field>.<key>", e.g., "named_metrics.foo".
weightExpirationPeriod
string
If a given endpoint has not reported load metrics in this long, stop using the reported weight. Defaults to 3m.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
weightUpdatePeriod
string
How often endpoint weights are recalculated. Values less than 100ms are capped at 100ms. Default 1s.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$consistentHash object
ConsistentHash defines the configuration when the load balancer type is
set to ConsistentHash
cookie object
Cookie configures the cookie hash policy when the consistent hash type is set to Cookie.
attributes
object
Additional Attributes to set for the generated cookie.
name
string required
Name of the cookie to hash.
If this cookie does not exist in the request, Envoy will generate a cookie and set
the TTL on the response back to the client based on Layer 4
attributes of the backend endpoint, to ensure that these future requests
go to the same backend endpoint. Make sure to set the TTL field for this case.
ttl
string
TTL of the generated cookie if the cookie is not present. This value sets the
Max-Age attribute value.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$header object
Header configures the header hash policy when the consistent hash type is set to Header.
Deprecated: use Headers instead
name
string required
Name of the header to hash.
headers []object
Headers configures the header hash policy for each header, when the consistent hash type is set to Headers.
name
string required
Name of the header to hash.
queryParams []object
QueryParams configures the query parameter hash policy when the consistent hash type is set to QueryParams.
name
string required
Name of the query param to hash.
tableSize
integer
The table size for consistent hashing, must be prime number limited to 5000011.
format:
int64minimum:
2maximum:
5.000011e+06
type
string required
ConsistentHashType defines the type of input to hash on. Valid Type values are
"SourceIP",
"Header",
"Headers",
"Cookie".
"QueryParams".
enum:
SourceIP, Header, Headers, Cookie, QueryParamsdynamicModule object
DynamicModule defines the configuration when the load balancer type is
set to DynamicModule. The referenced module must be registered in the
EnvoyProxy resource's dynamicModules allowlist.
config
object
Config is optional configuration for the module's load balancer
implementation. This is serialized and passed to the module's
initialization function.
lbPolicyName
string required
LBPolicyName identifies a specific load balancer implementation within
the dynamic module. A single shared library can contain multiple LB
policy implementations. This value is passed to the module's
initialization function to select the appropriate implementation.
minLength:
1maxLength:
253
name
string required
Name references a dynamic module registered in the EnvoyProxy resource's
dynamicModules list. The referenced module must exist in the registry;
otherwise, the policy will be rejected.
pattern:
^[a-z0-9]([a-z0-9.-]*[a-z0-9])?$minLength:
1maxLength:
253endpointOverride object
EndpointOverride defines the configuration for endpoint override.
When specified, the load balancer will attempt to route requests to endpoints
based on the override information extracted from request headers or metadata.
If the override endpoints are not available, the configured load balancer policy will be used as fallback.
extractFrom []object required
ExtractFrom defines the sources to extract endpoint override information from.
minItems:
1maxItems:
10
header
string
Header defines the header to get the override endpoint addresses.
The header value must specify at least one endpoint in `IP:Port` format or multiple endpoints in `IP:Port,IP:Port,...` format.
For example `10.0.0.5:8080` or `[2600:4040:5204::1574:24ae]:80`.
The IPv6 address is enclosed in square brackets.
slowStart object
SlowStart defines the configuration related to the slow start load balancer policy.
If set, during slow start window, traffic sent to the newly added hosts will gradually increase.
Supported for RoundRobin, LeastRequest, and BackendUtilization load balancers.
window
string required
Window defines the duration of the warm up period for newly added host.
During slow start window, traffic sent to the newly added hosts will gradually increase.
Currently only supports linear growth of traffic. For additional details,
see https://www.envoyproxy.io/docs/envoy/latest/api-v3/config/cluster/v3/cluster.proto#config-cluster-v3-cluster-slowstartconfig
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
type
string required
Type decides the type of Load Balancer policy.
Valid LoadBalancerType values are
"ConsistentHash",
"LeastRequest",
"Random",
"RoundRobin",
"BackendUtilization",
"DynamicModule".
enum:
ConsistentHash, LeastRequest, Random, RoundRobin, BackendUtilization, DynamicModulezoneAware object
ZoneAware defines the configuration related to the distribution of requests between locality zones.
preferLocal object
PreferLocalZone configures zone-aware routing to prefer sending traffic to the local locality zone.
force object
ForceLocalZone defines override configuration for forcing all traffic to stay within the local zone instead of the default behavior
which maintains equal distribution among upstream endpoints while sending as much traffic as possible locally.
minEndpointsInZoneThreshold
integer
MinEndpointsInZoneThreshold is the minimum number of upstream endpoints in the local zone required to honor the forceLocalZone
override. This is useful for protecting zones with fewer endpoints.
format:
int32
minEndpointsThreshold
integer
MinEndpointsThreshold is the minimum number of total upstream endpoints across all zones required to enable zone-aware routing.
format:
int64
percentageEnabled
integer
Configures percentage of requests that will be considered for zone aware routing if zone aware routing is configured. If not specified, Envoy defaults to 100%.
format:
int32minimum:
0maximum:
100weightedZones []object
WeightedZones configures weight-based traffic distribution across locality zones.
Traffic is distributed proportionally based on the sum of all zone weights.
weight
integer required
Weight defines the weight for this locality.
Higher values receive more traffic. The actual traffic distribution
is proportional to this value relative to other localities.
format:
int32
zone
string required
Zone specifies the topology zone this weight applies to.
The value should match the topology.kubernetes.io/zone label
of the nodes where endpoints are running.
Zones not listed in the configuration receive a default weight of 1.
proxyProtocol object
ProxyProtocol enables the Proxy Protocol when communicating with the backend.
version
string required
Version of ProxyProtocol
Valid ProxyProtocolVersion values are
"V1"
"V2"
enum:
V1, V2retry object
Retry provides more advanced usage, allowing users to customize the number of retries, retry fallback strategy, and retry triggering conditions.
If not set, retry will be disabled.
numAttemptsPerPriority
integer
NumAttemptsPerPriority defines the number of requests (initial attempt + retries)
that should be sent to the same priority before switching to a different one.
If not specified or set to 0, all requests are sent to the highest priority that is healthy.
format:
int32
numRetries
integer
NumRetries is the number of retries to be attempted. Defaults to 2.
format:
int32minimum:
0perRetry object
PerRetry is the retry policy to be applied per retry attempt.
backOff object
Backoff is the backoff policy to be applied per retry attempt. gateway uses a fully jittered exponential
back-off algorithm for retries. For additional details,
see https://www.envoyproxy.io/docs/envoy/latest/configuration/http/http_filters/router_filter#config-http-filters-router-x-envoy-max-retries
baseInterval
string
BaseInterval is the base interval between retries.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxInterval
string
MaxInterval is the maximum interval between retries. This parameter is optional, but must be greater than or equal to the base_interval if set.
The default is 10 times the base_interval
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
timeout
string
Timeout is the timeout per retry attempt.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$retryOn object
RetryOn specifies the retry trigger condition.
If not specified, the default is to retry on connect-failure,refused-stream,unavailable,cancelled,retriable-status-codes(503).
httpStatusCodes
[]integer
HttpStatusCodes specifies the http status codes to be retried.
The retriable-status-codes trigger must also be configured for these status codes to trigger a retry.
triggers
[]string
Triggers specifies the retry trigger condition(Http/Grpc).
tcpKeepalive object
TcpKeepalive settings associated with the upstream client connection.
Disabled by default.
idleTime
string
The duration a connection needs to be idle before keep-alive
probes start being sent.
The duration format is
Defaults to `7200s`.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
The duration between keep-alive probes.
Defaults to `75s`.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
probes
integer
The total number of unacknowledged probes to send before deciding
the connection is dead.
Defaults to 9.
format:
int32timeout object
Timeout settings for the backend connections.
http object
Timeout settings for HTTP.
connectionIdleTimeout
string
The idle timeout for an HTTP connection. Idle time is defined as a period in which there are no active requests in the connection.
Default: 1 hour.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxConnectionDuration
string
The maximum duration of an HTTP connection.
Default: unlimited.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxStreamDuration
string
MaxStreamDuration is the maximum duration for a stream to complete. This timeout measures the time
from when the request is sent until the response stream is fully consumed and does not apply to
non-streaming requests.
When set to "0s", no max duration is applied and streams can run indefinitely.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
requestTimeout
string
RequestTimeout is the time until which entire response is received from the upstream.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
streamIdleTimeout
string
The stream idle timeout defines the amount of time a stream can exist without any upstream or downstream activity.
If not specified, StreamIdleTimeout is inherited from the listener-level setting, which can be configured via ClientTrafficPolicy.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$tcp object
Timeout settings for TCP.
connectTimeout
string
The timeout for network connection establishment, including TCP and TLS handshakes.
Default: 10 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
headersToBackend
[]string
HeadersToBackend are the authorization response headers that will be added
to the original client request before sending it to the backend server.
Note that coexisting headers will be overridden.
If not specified, no authorization response headers will be added to the
original client request.
path
string
Path is the path of the HTTP External Authorization service.
If path is specified, the authorization request will be sent to that path,
or else the authorization request will use the path of the original request.
Please note that the original request path will be appended to the path specified here.
For example, if the original request path is "/hello", and the path specified here is "/auth",
then the path of the authorization request will be "/auth/hello". If the path is not specified,
the path of the authorization request will be "/hello".
Only one of Path or PathOverride can be set.
pathOverride
string
PathOverride replaces the original request path in the authorization request.
If set, the path will be overridden to this value during authorization.
For example, if the original request path is "/hello", and PathOverride is set to "/auth",
then the path of the authorization request will be "/auth".
Only one of Path or PathOverride can be set.
includeRouteMetadata
boolean
IncludeRouteMetadata sends Envoy Gateway's built-in route metadata to the
external authorization service as context.
This includes Envoy Gateway's built-in metadata for the selected route in
the "envoy-gateway" metadata namespace.
The metadata is exposed under the "resources" field as a list of route
resource objects. For example:
envoy-gateway:
resources:
- kind: HTTPRoute
name: backend
namespace: default
annotations:
foo: bar
The resource object may include fields such as kind, namespace, name,
sectionName, and supported route annotations.
recomputeRoute
boolean
RecomputeRoute clears the route cache and recalculates the routing decision.
This field must be enabled if the headers added or modified by the ExtAuth are used for
route matching decisions. If the recomputation selects a new route, features targeting
the new matched route will be applied.
statusOnError
integer
Sets the HTTP status that is returned when the authorization service returns an error
or cannot be reached. Defaults to 403 Forbidden.
Only 4xx and 5xx status codes are supported.
enum:
enum: 400, 401, 402, 403, 404, 405, 406, 407, 408, 409, 410, 411, 412, 413, 414, 415, 416, 417, 421, 422, 423, 424, 426, 428, ...
400, 401, 402, 403, 404, 405, 406, 407, 408, 409, 410, 411, 412, 413, 414, 415, 416, 417, 421, 422, 423, 424, 426, 428, 429, 431, 500, 501, 502, 503, 504, 505, 506, 507, 508, 510, 511
format:
int32
timeout
string
Timeout defines the timeout for requests to the external authorization service.
If not specified, defaults to 10 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$oauth object
OAuth defines the configuration for the MCP spec compatible OAuth authentication.
audiences
[]string
Audiences is a list of JWT audiences allowed access.
It is recommended to set this field for token audience validation, as it is a security best practice to prevent token misuse.
Reference: https://modelcontextprotocol.io/specification/2025-06-18/basic/authorization#token-audience-binding-and-validation
maxItems:
32claimToHeaders []object
ClaimToHeaders specifies JWT claims to extract and forward as HTTP headers to backend MCP servers.
This enables backends to access user identity for authorization, auditing, or personalization.
Security considerations:
- Any client-provided headers matching the configured header names will be stripped to prevent forgery
- Only the specified claims are extracted; the full JWT is not forwarded to backends
- Consider using a header prefix (e.g., "X-Jwt-Claim-") to avoid conflicts with other headers
maxItems:
16
claim
string required
Claim is the JWT Claim that should be saved into the header : it can be a nested claim of type
(eg. "claim.nested.key", "sub"). The nested claim name must use dot "."
to separate the JSON name path.
header
string required
Header defines the name of the HTTP request header that the JWT Claim will be saved into.
issuer
string required
Issuer is the authorization server's issuer identity.
format:
urijwks object
JWKS defines how a JSON Web Key Sets (JWKS) can be obtained to verify the access tokens presented by the clients.
If not specified, the JWKS URI will be discovered from the OAuth 2.0 Authorization Server Metadata
as per RFC 8414 by querying the `/.well-known/oauth-authorization-server` endpoint on the Issuer.
localJWKS object
LocalJWKS defines how to get the JSON Web Key Sets (JWKS) from a local source.
inline
string
Inline contains the value as an inline string.
type
string required
Type is the type of method to use to read the body value.
Valid values are Inline and ValueRef, default is Inline.
enum:
Inline, ValueRefvalueRef object
ValueRef is a reference to a local ConfigMap that contains the JSON Web Key Sets (JWKS).
The value of key `jwks` in the ConfigMap will be used.
If the key is not found, the first value in the ConfigMap will be used.
group
string required
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string required
Kind is kind of the referent. For example "HTTPRoute" or "Service".
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253remoteJWKS object
RemoteJWKS defines how to fetch and cache JSON Web Key Sets (JWKS) from a remote
HTTP/HTTPS endpoint.
backendRef object
BackendRef references a Kubernetes object that represents the
backend server to which the authorization request will be sent.
Deprecated: Use BackendRefs instead.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535backendRefs []object
BackendRefs references a Kubernetes object that represents the
backend server to which the authorization request will be sent.
maxItems:
16
fallback
boolean
Fallback indicates whether the backend is designated as a fallback.
Multiple fallback backends can be configured.
It is highly recommended to configure active or passive health checks to ensure that failover can be detected
when the active backends become unhealthy and to automatically readjust once the primary backends are healthy again.
The overprovisioning factor is set to 1.4, meaning the fallback backends will only start receiving traffic when
the health of the active backends falls below 72%.
group
string
Group is the group of the referent. For example, "gateway.networking.k8s.io".
When unspecified or empty string, core API group is inferred.
pattern:
^$|^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$maxLength:
253
kind
string
Kind is the Kubernetes resource kind of the referent. For example
"Service".
Defaults to "Service" when not specified.
ExternalName services can refer to CNAME DNS records that may live
outside of the cluster and as such are difficult to reason about in
terms of conformance. They also may not be safe to forward to (see
CVE-2021-25740 for more information). Implementations SHOULD NOT
support ExternalName Services.
Support: Core (Services with a type other than ExternalName)
Support: Implementation-specific (Services with type ExternalName)
pattern:
^[a-zA-Z]([-a-zA-Z0-9]*[a-zA-Z0-9])?$minLength:
1maxLength:
63
name
string required
Name is the name of the referent.
minLength:
1maxLength:
253
namespace
string
Namespace is the namespace of the backend. When unspecified, the local
namespace is inferred.
Note that when a namespace different than the local namespace is specified,
a ReferenceGrant object is required in the referent namespace to allow that
namespace's owner to accept the reference. See the ReferenceGrant
documentation for details.
Support: Core
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?$minLength:
1maxLength:
63
port
integer
Port specifies the destination port number to use for this resource.
Port is required when the referent is a Kubernetes Service. In this
case, the port number is the service port number, not the target port.
For other resources, destination port might be derived from the referent
resource or this field.
format:
int32minimum:
1maximum:
65535
weight
integer
Weight specifies the proportion of requests forwarded to the referenced
backend. This is computed as weight/(sum of all weights in this
BackendRefs list). For non-zero values, there may be some epsilon from
the exact proportion defined here depending on the precision an
implementation supports. Weight is not a percentage and the sum of
weights does not need to equal 100.
If only one backend is specified and it has a weight greater than 0, 100%
of the traffic is forwarded to that backend. If weight is set to 0, no
traffic should be forwarded for this entry. If unspecified, weight
defaults to 1.
Support for this field varies based on the context where used.
format:
int32minimum:
0maximum:
1e+06backendSettings object
BackendSettings holds configuration for managing the connection
to the backend.
circuitBreaker object
Circuit Breaker settings for the upstream connections and requests.
If not set, circuit breakers will be enabled with the default thresholds
maxConnections
integer
The maximum number of connections that Envoy will establish to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxParallelRequests
integer
The maximum number of parallel requests that Envoy will make to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxParallelRetries
integer
The maximum number of parallel retries that Envoy will make to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxPendingRequests
integer
The maximum number of pending requests that Envoy will queue to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09
maxRequestsPerConnection
integer
The maximum number of requests that Envoy will make over a single connection to the referenced backend defined within a xRoute rule.
Default: unlimited.
format:
int64minimum:
0maximum:
4.294967295e+09perEndpoint object
PerEndpoint defines Circuit Breakers that will apply per-endpoint for an upstream cluster
maxConnections
integer
MaxConnections configures the maximum number of connections that Envoy will establish per-endpoint to the referenced backend defined within a xRoute rule.
format:
int64minimum:
0maximum:
4.294967295e+09retryBudget object
RetryBudget provides settings for retry budget, which limits the number of retries in a given percentage.
RetryBudget take precedence over maxParallelRetries.
minRetryConcurrency
integer
MinRetryConcurrency specifies the minimum retry concurrency allowed for the retry budget.
For example, a budget of 20% with a minimum retry concurrency of 3
will allow 5 active retries while there are 25 active requests.
If there are 2 active requests, there are still 3 active retries
allowed because of the minimum retry concurrency.
Defaults to 3.
format:
int32percent object required
Percent specifies the limit on concurrent retries as a percentage [0, 100] of
the sum of active requests and active pending requests.
denominator
integer
format:
int32minimum:
1
numerator
integer required
format:
int32minimum:
0connection object
Connection includes backend connection settings.
bufferLimit
string | integer
BufferLimit Soft limit on size of the cluster’s connections read and write buffers.
BufferLimit applies to connection streaming (maybe non-streaming) channel between processes, it's in user space.
If unspecified, an implementation defined default is applied (32768 bytes).
For example, 20Mi, 1Gi, 256Ki etc.
Note: that when the suffix is not provided, the value is interpreted as bytes.
preconnect object
Preconnect configures proactive upstream connections to reduce latency by establishing
connections before they’re needed and avoiding connection establishment overhead.
If unset, Envoy will fetch connections as needed to serve in-flight requests.
perEndpointPercent
integer
PerEndpointPercent configures how many additional connections to maintain per
upstream endpoint, useful for high-QPS or latency sensitive services. Expressed as a
percentage of the connections required by active streams
(e.g. 100 = preconnect disabled, 105 = 1.05x connections per-endpoint, 200 = 2.00×).
Allowed value range is between 100-300. When both PerEndpointPercent and
PredictivePercent are set, Envoy ensures both are satisfied (max of the two).
format:
int32minimum:
100maximum:
300
predictivePercent
integer
PredictivePercent configures how many additional connections to maintain
across the cluster by anticipating which upstream endpoint the load balancer
will select next, useful for low-QPS services. Relies on deterministic
loadbalancing and is only supported with Random or RoundRobin.
Expressed as a percentage of the connections required by active streams
(e.g. 100 = 1.0 (no preconnect), 105 = 1.05× connections across the cluster, 200 = 2.00×).
Minimum allowed value is 100. When both PerEndpointPercent and PredictivePercent are
set Envoy ensures both are satisfied per host (max of the two).
format:
int32minimum:
100
socketBufferLimit
string | integer
SocketBufferLimit provides configuration for the maximum buffer size in bytes for each socket
to backend.
SocketBufferLimit applies to socket streaming channel between TCP/IP stacks, it's in kernel space.
For example, 20Mi, 1Gi, 256Ki etc.
Note that when the suffix is not provided, the value is interpreted as bytes.
dns object
DNS includes dns resolution settings.
dnsRefreshRate
string
DNSRefreshRate specifies the rate at which DNS records should be refreshed.
Defaults to 30 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
lookupFamily
string
LookupFamily determines how Envoy would resolve DNS for Routes where the backend is specified as a fully qualified domain name (FQDN).
If set, this configuration overrides other defaults.
enum:
IPv4, IPv6, IPv4Preferred, IPv6Preferred, IPv4AndIPv6
respectDnsTtl
boolean
RespectDNSTTL indicates whether the DNS Time-To-Live (TTL) should be respected.
If the value is set to true, the DNS refresh rate will be set to the resource record’s TTL.
Defaults to true.
healthCheck object
HealthCheck allows gateway to perform active health checking on backends.
active object
Active health check configuration
grpc object
GRPC defines the configuration of the GRPC health checker.
It's optional, and can only be used if the specified type is GRPC.
service
string
Service to send in the health check request.
If this is not specified, then the health check request applies to the entire
server and not to a specific service.
healthyThreshold
integer
HealthyThreshold defines the number of healthy health checks required before a backend host is marked healthy.
format:
int32minimum:
1http object
HTTP defines the configuration of http health checker.
It's required while the health checker type is HTTP.
expectedResponse object
ExpectedResponse defines a list of HTTP expected responses to match.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
expectedStatuses
[]integer
ExpectedStatuses defines a list of HTTP response statuses considered healthy.
Defaults to 200 only
hostname
string
Hostname defines the HTTP Host header used for active HTTP health checks.
Host selection uses this order: this field, the associated Backend endpoint
hostname if available, then the effective Route hostname.
pattern:
^[a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*$minLength:
1maxLength:
253
method
string
Method defines the HTTP method used for health checking.
Defaults to GET
path
string required
Path defines the HTTP path that will be requested during health checking.
minLength:
1maxLength:
1024
retriableStatuses
[]integer
RetriableStatuses defines a list of HTTP response statuses considered retriable.
Responses matching these statuses count towards the unhealthy threshold but
do not result in the host being considered immediately unhealthy.
The expected statuses take precedence for any range overlaps with this field.
initialJitter
string
InitialJitter defines the maximum time Envoy will wait before the first health check.
Envoy will randomly select a value between 0 and the initial jitter value.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
Interval defines the time between active health checks.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$overrides object
Overrides defines the configuration of the overriding health check settings for all endpoints
in the backend cluster. This allows customization of port and other settings that may differ
from the main service configuration.
port
integer
Port overrides the health check port.
If not set, the endpoint's serving port is used for health checks.
This is useful when health checks are served on a different port than
the main service port (e.g., port 443 for service, port 9090 for health checks).
format:
int32minimum:
1maximum:
65535tcp object
TCP defines the configuration of tcp health checker.
It's required while the health checker type is TCP.
receive object
Receive defines the expected response payload.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
send object
Send defines the request payload.
binary
string
Binary payload base64 encoded.
format:
byte
text
string
Text payload in plain text.
type
string required
Type defines the type of the payload.
timeout
string
Timeout defines the time to wait for a health check response.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
type
string required
Type defines the type of health checker.
unhealthyThreshold
integer
UnhealthyThreshold defines the number of unhealthy health checks required before a backend host is marked unhealthy.
Without RetriableStatuses configured, any health check failure results in the host being immediately
considered unhealthy. When RetriableStatuses is set, health checks returning those statuses are retried
up to this threshold before the host is marked unhealthy.
format:
int32minimum:
1
panicThreshold
integer
When number of unhealthy endpoints for a backend reaches this threshold
Envoy will disregard health status and balance across all endpoints.
It's designed to prevent a situation in which host failures cascade throughout the cluster
as load increases. If not set, the default value is 50%. To disable panic mode, set value to `0`.
format:
int32minimum:
0maximum:
100passive object
Passive passive check configuration
alwaysEjectOneEndpoint
boolean
AlwaysEjectOneEndpoint defines whether at least one host should be ejected,
regardless of MaxEjectionPercent.
baseEjectionTime
string
BaseEjectionTime defines the base duration for which a host will be ejected on consecutive failures.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
consecutive5XxErrors
integer
Consecutive5xxErrors sets the number of consecutive 5xx errors triggering ejection.
format:
int32
consecutiveGatewayErrors
integer
ConsecutiveGatewayErrors sets the number of consecutive gateway errors triggering ejection.
format:
int32
consecutiveLocalOriginFailures
integer
ConsecutiveLocalOriginFailures sets the number of consecutive local origin failures triggering ejection.
Parameter takes effect only when split_external_local_origin_errors is set to true.
format:
int32
failurePercentageThreshold
integer
FailurePercentageThreshold sets the failure percentage threshold for outlier detection.
If the failure percentage of a given host is greater than or equal to this value, it will be ejected.
Defaults to 85.
format:
int32minimum:
0maximum:
100
interval
string
Interval defines the time between passive health checks.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxEjectionPercent
integer
MaxEjectionPercent sets the maximum percentage of hosts in a cluster that can be ejected.
format:
int32minimum:
0maximum:
100
splitExternalLocalOriginErrors
boolean
SplitExternalLocalOriginErrors enables splitting of errors between external and local origin.
http2 object
HTTP2 provides HTTP/2 configuration for backend connections.
connectionKeepalive object
ConnectionKeepalive configures HTTP/2 connection keepalive using PING frames.
idleInterval
string
IdleInterval specifies how long a connection must be idle before a PING is sent.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
Interval specifies how often to send HTTP/2 PING frames to keep the connection alive.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
intervalJitter
integer
IntervalJitter specifies a random jitter percentage added to each interval.
Defaults to 15% if not specified.
format:
int32minimum:
0maximum:
100
timeout
string
Timeout specifies how long to wait for a PING response before considering the connection dead.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
initialConnectionWindowSize
string | integer
InitialConnectionWindowSize sets the initial window size for HTTP/2 connections.
If not set, the default value is 1 MiB.
initialStreamWindowSize
string | integer
InitialStreamWindowSize sets the initial window size for HTTP/2 streams.
If not set, the default value is 64 KiB(64*1024).
maxConcurrentStreams
integer
MaxConcurrentStreams sets the maximum number of concurrent streams allowed per connection.
If not set, the default value is 100.
format:
int32minimum:
1maximum:
2.147483647e+09
onInvalidMessage
string
OnInvalidMessage determines if Envoy will terminate the connection or just the offending stream in the event of HTTP messaging error
It's recommended for L2 Envoy deployments to set this value to TerminateStream.
https://www.envoyproxy.io/docs/envoy/latest/configuration/best_practices/level_two
Default: TerminateConnection
loadBalancer object
LoadBalancer policy to apply when routing traffic from the gateway to
the backend endpoints. Defaults to `LeastRequest`.
backendUtilization object
BackendUtilization defines the configuration when the load balancer type is
set to BackendUtilization.
blackoutPeriod
string
A given endpoint must report load metrics continuously for at least this long before the endpoint weight will be used.
Default is 10s.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
errorUtilizationPenaltyPercent
integer
ErrorUtilizationPenaltyPercent adjusts endpoint weights based on the error rate (eps/qps).
This is expressed as a percentage-based integer where 100 represents 1.0, 150 represents 1.5, etc.
For example:
- 100 => 1.0x
- 120 => 1.2x
- 200 => 2.0x
Must be non-negative.
format:
int32minimum:
0
keepResponseHeaders
boolean
KeepResponseHeaders keeps the ORCA load report headers/trailers before sending the response to the client.
Defaults to false.
metricNamesForComputingUtilization
[]string
Metric names used to compute utilization if application_utilization is not set.
For map fields in ORCA proto, use the form "<map_field>.<key>", e.g., "named_metrics.foo".
weightExpirationPeriod
string
If a given endpoint has not reported load metrics in this long, stop using the reported weight. Defaults to 3m.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
weightUpdatePeriod
string
How often endpoint weights are recalculated. Values less than 100ms are capped at 100ms. Default 1s.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$consistentHash object
ConsistentHash defines the configuration when the load balancer type is
set to ConsistentHash
cookie object
Cookie configures the cookie hash policy when the consistent hash type is set to Cookie.
attributes
object
Additional Attributes to set for the generated cookie.
name
string required
Name of the cookie to hash.
If this cookie does not exist in the request, Envoy will generate a cookie and set
the TTL on the response back to the client based on Layer 4
attributes of the backend endpoint, to ensure that these future requests
go to the same backend endpoint. Make sure to set the TTL field for this case.
ttl
string
TTL of the generated cookie if the cookie is not present. This value sets the
Max-Age attribute value.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$header object
Header configures the header hash policy when the consistent hash type is set to Header.
Deprecated: use Headers instead
name
string required
Name of the header to hash.
headers []object
Headers configures the header hash policy for each header, when the consistent hash type is set to Headers.
name
string required
Name of the header to hash.
queryParams []object
QueryParams configures the query parameter hash policy when the consistent hash type is set to QueryParams.
name
string required
Name of the query param to hash.
tableSize
integer
The table size for consistent hashing, must be prime number limited to 5000011.
format:
int64minimum:
2maximum:
5.000011e+06
type
string required
ConsistentHashType defines the type of input to hash on. Valid Type values are
"SourceIP",
"Header",
"Headers",
"Cookie".
"QueryParams".
enum:
SourceIP, Header, Headers, Cookie, QueryParamsdynamicModule object
DynamicModule defines the configuration when the load balancer type is
set to DynamicModule. The referenced module must be registered in the
EnvoyProxy resource's dynamicModules allowlist.
config
object
Config is optional configuration for the module's load balancer
implementation. This is serialized and passed to the module's
initialization function.
lbPolicyName
string required
LBPolicyName identifies a specific load balancer implementation within
the dynamic module. A single shared library can contain multiple LB
policy implementations. This value is passed to the module's
initialization function to select the appropriate implementation.
minLength:
1maxLength:
253
name
string required
Name references a dynamic module registered in the EnvoyProxy resource's
dynamicModules list. The referenced module must exist in the registry;
otherwise, the policy will be rejected.
pattern:
^[a-z0-9]([a-z0-9.-]*[a-z0-9])?$minLength:
1maxLength:
253endpointOverride object
EndpointOverride defines the configuration for endpoint override.
When specified, the load balancer will attempt to route requests to endpoints
based on the override information extracted from request headers or metadata.
If the override endpoints are not available, the configured load balancer policy will be used as fallback.
extractFrom []object required
ExtractFrom defines the sources to extract endpoint override information from.
minItems:
1maxItems:
10
header
string
Header defines the header to get the override endpoint addresses.
The header value must specify at least one endpoint in `IP:Port` format or multiple endpoints in `IP:Port,IP:Port,...` format.
For example `10.0.0.5:8080` or `[2600:4040:5204::1574:24ae]:80`.
The IPv6 address is enclosed in square brackets.
slowStart object
SlowStart defines the configuration related to the slow start load balancer policy.
If set, during slow start window, traffic sent to the newly added hosts will gradually increase.
Supported for RoundRobin, LeastRequest, and BackendUtilization load balancers.
window
string required
Window defines the duration of the warm up period for newly added host.
During slow start window, traffic sent to the newly added hosts will gradually increase.
Currently only supports linear growth of traffic. For additional details,
see https://www.envoyproxy.io/docs/envoy/latest/api-v3/config/cluster/v3/cluster.proto#config-cluster-v3-cluster-slowstartconfig
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
type
string required
Type decides the type of Load Balancer policy.
Valid LoadBalancerType values are
"ConsistentHash",
"LeastRequest",
"Random",
"RoundRobin",
"BackendUtilization",
"DynamicModule".
enum:
ConsistentHash, LeastRequest, Random, RoundRobin, BackendUtilization, DynamicModulezoneAware object
ZoneAware defines the configuration related to the distribution of requests between locality zones.
preferLocal object
PreferLocalZone configures zone-aware routing to prefer sending traffic to the local locality zone.
force object
ForceLocalZone defines override configuration for forcing all traffic to stay within the local zone instead of the default behavior
which maintains equal distribution among upstream endpoints while sending as much traffic as possible locally.
minEndpointsInZoneThreshold
integer
MinEndpointsInZoneThreshold is the minimum number of upstream endpoints in the local zone required to honor the forceLocalZone
override. This is useful for protecting zones with fewer endpoints.
format:
int32
minEndpointsThreshold
integer
MinEndpointsThreshold is the minimum number of total upstream endpoints across all zones required to enable zone-aware routing.
format:
int64
percentageEnabled
integer
Configures percentage of requests that will be considered for zone aware routing if zone aware routing is configured. If not specified, Envoy defaults to 100%.
format:
int32minimum:
0maximum:
100weightedZones []object
WeightedZones configures weight-based traffic distribution across locality zones.
Traffic is distributed proportionally based on the sum of all zone weights.
weight
integer required
Weight defines the weight for this locality.
Higher values receive more traffic. The actual traffic distribution
is proportional to this value relative to other localities.
format:
int32
zone
string required
Zone specifies the topology zone this weight applies to.
The value should match the topology.kubernetes.io/zone label
of the nodes where endpoints are running.
Zones not listed in the configuration receive a default weight of 1.
proxyProtocol object
ProxyProtocol enables the Proxy Protocol when communicating with the backend.
version
string required
Version of ProxyProtocol
Valid ProxyProtocolVersion values are
"V1"
"V2"
enum:
V1, V2retry object
Retry provides more advanced usage, allowing users to customize the number of retries, retry fallback strategy, and retry triggering conditions.
If not set, retry will be disabled.
numAttemptsPerPriority
integer
NumAttemptsPerPriority defines the number of requests (initial attempt + retries)
that should be sent to the same priority before switching to a different one.
If not specified or set to 0, all requests are sent to the highest priority that is healthy.
format:
int32
numRetries
integer
NumRetries is the number of retries to be attempted. Defaults to 2.
format:
int32minimum:
0perRetry object
PerRetry is the retry policy to be applied per retry attempt.
backOff object
Backoff is the backoff policy to be applied per retry attempt. gateway uses a fully jittered exponential
back-off algorithm for retries. For additional details,
see https://www.envoyproxy.io/docs/envoy/latest/configuration/http/http_filters/router_filter#config-http-filters-router-x-envoy-max-retries
baseInterval
string
BaseInterval is the base interval between retries.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxInterval
string
MaxInterval is the maximum interval between retries. This parameter is optional, but must be greater than or equal to the base_interval if set.
The default is 10 times the base_interval
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
timeout
string
Timeout is the timeout per retry attempt.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$retryOn object
RetryOn specifies the retry trigger condition.
If not specified, the default is to retry on connect-failure,refused-stream,unavailable,cancelled,retriable-status-codes(503).
httpStatusCodes
[]integer
HttpStatusCodes specifies the http status codes to be retried.
The retriable-status-codes trigger must also be configured for these status codes to trigger a retry.
triggers
[]string
Triggers specifies the retry trigger condition(Http/Grpc).
tcpKeepalive object
TcpKeepalive settings associated with the upstream client connection.
Disabled by default.
idleTime
string
The duration a connection needs to be idle before keep-alive
probes start being sent.
The duration format is
Defaults to `7200s`.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
interval
string
The duration between keep-alive probes.
Defaults to `75s`.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
probes
integer
The total number of unacknowledged probes to send before deciding
the connection is dead.
Defaults to 9.
format:
int32timeout object
Timeout settings for the backend connections.
http object
Timeout settings for HTTP.
connectionIdleTimeout
string
The idle timeout for an HTTP connection. Idle time is defined as a period in which there are no active requests in the connection.
Default: 1 hour.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxConnectionDuration
string
The maximum duration of an HTTP connection.
Default: unlimited.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
maxStreamDuration
string
MaxStreamDuration is the maximum duration for a stream to complete. This timeout measures the time
from when the request is sent until the response stream is fully consumed and does not apply to
non-streaming requests.
When set to "0s", no max duration is applied and streams can run indefinitely.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
requestTimeout
string
RequestTimeout is the time until which entire response is received from the upstream.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
streamIdleTimeout
string
The stream idle timeout defines the amount of time a stream can exist without any upstream or downstream activity.
If not specified, StreamIdleTimeout is inherited from the listener-level setting, which can be configured via ClientTrafficPolicy.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$tcp object
Timeout settings for TCP.
connectTimeout
string
The timeout for network connection establishment, including TCP and TLS handshakes.
Default: 10 seconds.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
cacheDuration
string
Duration is a string value representing a duration in time. The format is as specified
in GEP-2257, a strict subset of the syntax parsed by Golang time.ParseDuration.
pattern:
^([0-9]{1,5}(h|m|s|ms)){1,4}$
uri
string required
URI is the HTTPS URI to fetch the JWKS. Envoy's system trust bundle is used to validate the server certificate.
If a custom trust bundle is needed, it can be specified in a BackendTLSConfig resource and target the BackendRefs.
minLength:
1maxLength:
253protectedResourceMetadata object required
ProtectedResourceMetadata defines the OAuth 2.0 Resource Server Metadata as per RFC 8414.
This is used to expose the metadata endpoint for mcp clients to discover the authorization servers,
supported scopes, and JWKS URI.
resource
string required
Resource is the identifier of the protected resource.
This should match the MCPRoute's URL. For example, if the MCPRoute's URL is
"https://api.example.com/mcp", the Resource should be "https://api.example.com/mcp".
format:
uri
resourceDocumentation
string
ResourceDocumentation is a URL that provides human-readable documentation for the resource.
format:
uri
resourceName
string
ResourceName is a human-readable name for the protected resource.
maxLength:
256
resourcePolicyUri
string
ResourcePolicyURI is a URL that points to the resource server's policy document.
format:
uri
resourceSigningAlgValuesSupported
[]string
ResourceSigningAlgValuesSupported is a list of JWS signing algorithms supported by the resource server.
These algorithms are used in the "alg" field of the JOSE header in signed tokens.
minItems:
1maxItems:
16
scopesSupported
[]string
ScopesSupported defines the minimal set of scopes required for the basic functionality of the MCPRoute.
It should avoid broad or overly permissive scopes to prevent clients from requesting tokens with excessive privileges.
If an operation requires additional scopes that are not present in the access token, the client will receive a
403 Forbidden response that includes the required scopes in the `scope` field of the `WWW-Authenticate` header.
This enables incremental privilege elevation through targeted `WWW-Authenticate: scope="..."` challenges when
privileged operations are first attempted.
maxItems:
32status object
Status defines the status details of the MCPRoute.
conditions []object
Conditions is the list of conditions by the reconciliation result.
Currently, at most one condition is set.
Known .status.conditions.type are: "Accepted", "NotAccepted".
lastTransitionTime
string required
lastTransitionTime is the last time the condition transitioned from one status to another.
This should be when the underlying condition changed. If that is not known, then using the time when the API field changed is acceptable.
format:
date-time
message
string required
message is a human readable message indicating details about the transition.
This may be an empty string.
maxLength:
32768
observedGeneration
integer
observedGeneration represents the .metadata.generation that the condition was set based upon.
For instance, if .metadata.generation is currently 12, but the .status.conditions[x].observedGeneration is 9, the condition is out of date
with respect to the current state of the instance.
format:
int64minimum:
0
reason
string required
reason contains a programmatic identifier indicating the reason for the condition's last transition.
Producers of specific condition types may define expected values and meanings for this field,
and whether the values are considered a guaranteed API.
The value should be a CamelCase string.
This field may not be empty.
pattern:
^[A-Za-z]([A-Za-z0-9_,:]*[A-Za-z0-9_])?$minLength:
1maxLength:
1024
status
string required
status of the condition, one of True, False, Unknown.
enum:
True, False, Unknown
type
string required
type of condition in CamelCase or in foo.example.com/CamelCase.
pattern:
^([a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*/)?(([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9])$maxLength:
316No matches. Try .spec.backendRefs for an exact path