mirror of https://github.com/hashicorp/consul
1912 lines
80 KiB
Markdown
1912 lines
80 KiB
Markdown
---
|
|
layout: docs
|
|
page_title: Service defaults configuration reference
|
|
description: ->
|
|
Use the service-defaults configuration entry to set default configurations for services, such as upstreams, protocols, and namespaces. Learn how to configure service-defaults.
|
|
---
|
|
|
|
# Service defaults configuration reference
|
|
|
|
This topic describes how to configure service defaults configuration entries. The service defaults configuration entry contains common configuration settings for service mesh services, such as upstreams and gateways. Refer to [Define service defaults](/consul/docs/services/usage/define-services#define-service-defaults) for usage information.
|
|
|
|
## Configuration model
|
|
|
|
The following list outlines field hierarchy, language-specific data types, requirements, and any applicable default values in service defaults configuration entries. Click on a property name to view additional details.
|
|
|
|
<Tabs>
|
|
<Tab heading="HCL and JSON" group="hcl">
|
|
|
|
- [`Kind`](#kind): string | required
|
|
- [`Name`](#name): string | required
|
|
- [`Namespace`](#namespace): string | `default` <EnterpriseAlert inline />
|
|
- [`Partition`](#partition): string | `default` <EnterpriseAlert inline />
|
|
- [`Meta`](#meta): map
|
|
- [`Protocol`](#protocol): string | `tcp`
|
|
- [`BalanceInboundConnections`](#balanceinboundconnections): string
|
|
- [`Mode`](#mode): string
|
|
- [`RateLimits`](#ratelimits): map | <EnterpriseAlert inline />
|
|
- [`InstanceLevel`](#ratelimits-instancelevel): map
|
|
- [`RequestsPerSecond`](#ratelimits-instancelevel-requestspersecond): number
|
|
- [`RequestsMaxBurst`](#ratelimits-instancelevel-requestsmaxburst): number
|
|
- [`Routes`](#ratelimits-instancelevel-routes): list maps
|
|
- [`PathExact`](#ratelimits-instancelevel-routes): string
|
|
- [`PathPrefix`](#ratelimits-instancelevel-routes): string
|
|
- [`PathRegex`](#ratelimits-instancelevel-routes): string
|
|
- [`RequestsPerSecond`](#ratelimits-instancelevel-routes): number | required
|
|
- [`RequestsMaxBurst`](#ratelimits-instancelevel-routes): number
|
|
- [`UpstreamConfig`](#upstreamconfig): map
|
|
- [`Overrides`](#upstreamconfig-overrides): map
|
|
- [`Name`](#upstreamconfig-overrides-name): string
|
|
- [`Namespace`](#upstreamconfig-overrides-namespace): string
|
|
- [`Peer`](#upstreamconfig-overrides-peer): string
|
|
- [`Protocol`](#upstreamconfig-overrides-protocol): string
|
|
- [`ConnectTimeoutMs`](#upstreamconfig-overrides-connecttimeoutms): int | `5000`
|
|
- [`MeshGateway`](#upstreamconfig-overrides-meshgateway): map
|
|
- [`mode`](#upstreamconfig-overrides-meshgateway): string
|
|
- [`BalanceOutboundConnections`](#upstreamconfig-overrides-balanceoutboundconnections): string
|
|
- [`Limits`](#upstreamconfig-overrides-limits): map
|
|
- [`MaxConnections`](#upstreamconfig-overrides-limits): number | `0`
|
|
- [`MaxPendingRequests`](#upstreamconfig-overrides-limits): number | `0`
|
|
- [`MaxConcurrentRequests`](#upstreamconfig-overrides-limits): number | `0`
|
|
- [`PassiveHealthCheck`](#upstreamconfig-overrides-passivehealthcheck): map
|
|
- [`Interval`](#upstreamconfig-overrides-passivehealthcheck): string | `0s`
|
|
- [`MaxFailures`](#upstreamconfig-overrides-passivehealthcheck): number | `0`
|
|
- [`EnforcingConsecutive5xx`](#upstreamconfig-overrides-passivehealthcheck): number | `0`
|
|
- [`MaxEjectionPercent`](#upstreamconfig-overrides-passivehealthcheck): number | `0`
|
|
- [`BaseEjectionTime`](#upstreamconfig-overrides-passivehealthcheck): string | `30s`
|
|
- [`Defaults`](#upstreamconfig-defaults): map
|
|
- [`Protocol`](#upstreamconfig-defaults-protocol): string
|
|
- [`ConnectTimeoutMs`](#upstreamconfig-defaults-connecttimeoutms): int | `5000`
|
|
- [`MeshGateway`](#upstreamconfig-defaults-meshgateway): map
|
|
- [`mode`](#upstreamconfig-defaults-meshgateway): string
|
|
- [`BalanceOutboundConnections`](#upstreamconfig-defaults-balanceoutboundconnections): string
|
|
- [`Limits`](#upstreamconfig-defaults-limits): map
|
|
- [`MaxConnections`](#upstreamconfig-defaults-limits): number | `0`
|
|
- [`MaxPendingRequests`](#upstreamconfig-defaults-limits): number | `0`
|
|
- [`MaxConcurrentRequests`](#upstreamconfig-defaults-limits): number | `0`
|
|
- [`PassiveHealthCheck`](#upstreamconfig-defaults-passivehealthcheck): map
|
|
- [`Interval`](#upstreamconfig-defaults-passivehealthcheck): string | `0s`
|
|
- [`MaxFailures`](#upstreamconfig-defaults-passivehealthcheck): number | `0`
|
|
- [`EnforcingConsecutive5xx`](#upstreamconfig-defaults-passivehealthcheck): number | `100`
|
|
- [`MaxEjectionPercent`](#upstreamconfig-defaults-passivehealthcheck): number | `0`
|
|
- [`BaseEjectionTime`](#upstreamconfig-defaults-passivehealthcheck): string | `30s`
|
|
- [`TransparentProxy`](#transparentproxy): map
|
|
- [`OutboundListenerPort`](#transparentproxy): number | `15001`
|
|
- [`DialedDirectly`](#transparentproxy ): boolean | `false`
|
|
- [`MutualTLSMode`](#mutualtlsmode): string
|
|
- [`EnvoyExtensions`](#envoyextensions): list
|
|
- [`Name`](#envoyextensions): string
|
|
- [`Required`](#envoyextensions): string
|
|
- [`Arguments`](#envoyextensions): map
|
|
- [`ConsulVersion`](#envoyextensions): string
|
|
- [`EnvoyVersion`](#envoyextensions): string
|
|
- [`Destination`](#destination): map
|
|
- [`Addresses`](#destination): list
|
|
- [`Port`](#destination): integer | `0`
|
|
- [`MaxInboundConnections`](#maxinboundconnections): number | `0`
|
|
- [`LocalConnectTimeoutMs`](#localconnecttimeoutms): number | `0`
|
|
- [`LocalRequestTimeoutMs`](#localrequesttimeoutms): number | `0`
|
|
- [`MeshGateway`](#meshgateway): map
|
|
- [`Mode`](#meshgateway): string
|
|
- [`ExternalSNI`](#externalsni): string
|
|
- [`Expose`](#expose): map
|
|
- [`Checks`](#expose-checks): boolean | `false`
|
|
- [`Paths`](#expose-paths): list
|
|
- [`Path`](#expose-paths): string
|
|
- [`LocalPathPort`](#expose-paths): integer | `0`
|
|
- [`ListenerPort`](#expose-paths): integer | `0`
|
|
- [`Protocol`](#expose-paths): string | `http`
|
|
|
|
</Tab>
|
|
<Tab heading="Kubernetes YAML" group="yaml">
|
|
|
|
- [`apiVersion`](#apiversion): string | required | must be set to `consul.hashicorp.com/v1alpha1`
|
|
- [`kind`](#kind): string
|
|
- [`metadata`](#metadata): map
|
|
- [`name`](#name): string
|
|
- [`namespace`](#namespace): string | `default`` | <EnterpriseAlert inline />
|
|
- [`spec`](#spec): map
|
|
- [`protocol`](#protocol): string | `tcp`
|
|
- [`balanceInboundConnections`](#balanceinboundconnections): string
|
|
- [`mode`](#mode): string
|
|
- [`rateLimits`](#spec-ratelimits): map | <EnterpriseAlert inline />
|
|
- [`instanceLevel`](#spec-ratelimits-instancelevel): map
|
|
- [`requestsPerSecond`](#spec-ratelimits-instancelevel-requestspersecond): number
|
|
- [`requestsMaxBurst`](#spec-ratelimits-instancelevel-requestsmaxburst): number
|
|
- [`routes`](#spec-ratelimits-instancelevel-routes): list maps
|
|
- [`pathExact`](#spec-ratelimits-instancelevel-routes): string
|
|
- [`pathPrefix`](#spec-ratelimits-instancelevel-routes): string
|
|
- [`pathRegex`](#spec-ratelimits-instancelevel-routes): string
|
|
- [`requestsPerSecond`](#spec-ratelimits-instancelevel-routes): number | required
|
|
- [`requestsMaxBurst`](#spec-ratelimits-instancelevel-routes): number
|
|
- [`upstreamConfig`](#upstreamconfig): map
|
|
- [`overrides`](#upstreamconfig-overrides): list
|
|
- [`name`](#upstreamconfig-overrides-name): string
|
|
- [`namespace`](#upstreamconfig-overrides-namespace): string
|
|
- [`peer`](#upstreamconfig-overrides-peer): string
|
|
- [`protocol`](#upstreamconfig-overrides-protocol): string
|
|
- [`connectTimeoutMs`](#upstreamconfig-overrides-connecttimeoutms): number | `5000`
|
|
- [`meshGateway`](#upstreamconfig-overrides-meshgateway): map
|
|
- [`mode`](#upstreamconfig-overrides-meshgateway): string
|
|
- [`balanceOutboundConnections`](#overrides-balanceoutboundconnections): string
|
|
- [`limits`](#upstreamconfig-overrides-limits): map
|
|
- [`maxConnections`](#upstreamconfig-overrides-limits): number | `0`
|
|
- [`maxPendingRequests`](#upstreamconfig-overrides-limits): number | `0`
|
|
- [`maxConcurrentRequests`](#upstreamconfig-overrides-limits): number | `0`
|
|
- [`passiveHealthCheck`](#upstreamconfig-overrides-passivehealthcheck): map
|
|
- [`interval`](#upstreamconfig-overrides-passivehealthcheck): string | `0s`
|
|
- [`maxFailures`](#upstreamconfig-overrides-passivehealthcheck): number | `0`
|
|
- [`enforcingConsecutive5xx`](#upstreamconfig-overrides-passivehealthcheck): number | `100`
|
|
- [`maxEjectionPercent`](#upstreamconfig-overrides-passivehealthcheck): number | `10`
|
|
- [`baseEjectionTime`](#upstreamconfig-overrides-passivehealthcheck): string | `30s`
|
|
- [`defaults`](#upstreamconfig-defaults): map
|
|
- [`protocol`](#upstreamconfig-defaults-protocol): string
|
|
- [`connectTimeoutMs`](#upstreamconfig-defaults-connecttimeoutms): number | `5000`
|
|
- [`meshGateway`](#upstreamconfig-defaults-meshgateway): map
|
|
- [`mode`](#upstreamconfig-defaults-meshgateway): string
|
|
- [`balanceOutboundConnections`](#upstreamconfig-defaults-balanceoutboundconnections): string
|
|
- [`limits`](#upstreamconfig-defaults-limits): map
|
|
- [`maxConnections`](#upstreamconfig-defaults-limits): number | `0`
|
|
- [`maxPendingRequests`](#upstreamconfig-defaults-limits): number | `0`
|
|
- [`maxConcurrentRequests`](#upstreamconfig-defaults-limits): number | `0`
|
|
- [`passiveHealthCheck`](#upstreamconfig-defaults-passivehealthcheck): map
|
|
- [`interval`](#upstreamconfig-defaults-passivehealthcheck): string | `0s`
|
|
- [`maxFailures`](#upstreamconfig-defaults-passivehealthcheck): number | `0`
|
|
- [`enforcingConsecutive5xx`](#upstreamconfig-defaults-passivehealthcheck): number | `100`
|
|
- [`maxEjectionPercent`](#upstreamconfig-defaults-passivehealthcheck): number | `10`
|
|
- [`baseEjectionTime`](#upstreamconfig-defaults-passivehealthcheck): string | `30s`
|
|
- [`transparentProxy`](#transparentproxy): map
|
|
- [`outboundListenerPort`](#transparentproxy): number | `15001`
|
|
- [`dialedDirectly`](#transparentproxy): boolean | `false`
|
|
- [`mutualTLSMode`](#mutualtlsmode): string
|
|
- [`envoyExtensions`](#envoyextensions): list
|
|
- [`name`](#envoyextensions): string
|
|
- [`required`](#envoyextensions): string
|
|
- [`arguments`](#envoyextensions): map
|
|
- [`consulVersion`](#envoyextensions): string
|
|
- [`envoyVersion`](#envoyextensions): string
|
|
- [`destination`](#destination): map
|
|
- [`addresses`](#destination): list
|
|
- [`port`](#destination): number | `0`
|
|
- [`maxInboundConnections`](#maxinboundconnections): number | `0`
|
|
- [`localConnectTimeoutMs`](#localconnecttimeoutms): number | `0`
|
|
- [`localRequestTimeoutMs`](#localrequesttimeoutms): number | `0`
|
|
- [`meshGateway`](#meshgateway): map
|
|
- [`mode`](#meshgateway): string
|
|
- [`externalSNI`](#externalsni): string
|
|
- [`expose`](#expose): map
|
|
- [`checks`](#expose-checks): boolean | `false`
|
|
- [`paths`](#expose-paths): list
|
|
- [`path`](#expose-paths): string
|
|
- [`localPathPort`](#expose-paths): number | `0`
|
|
- [`listenerPort`](#expose-paths): number | `0`
|
|
- [`protocol`](#expose-paths): string | `http`
|
|
|
|
</Tab>
|
|
</Tabs>
|
|
|
|
## Complete configuration
|
|
|
|
When every field is defined, a service-defaults configuration entry has the following form:
|
|
|
|
<Tabs>
|
|
<Tab heading="HCL" group="hcl">
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "<service_name>"
|
|
Namespace = "default"
|
|
Partition = "default"
|
|
Meta = {
|
|
Key = "value"
|
|
}
|
|
Protocol = "tcp"
|
|
BalanceInboundConnections = "exact_balance"
|
|
Mode = "<mode for directing traffic>"
|
|
RateLimits = {
|
|
InstanceLevel = {
|
|
RequestsPerSecond = <average number of requests allowed to the service per second>
|
|
RequestsMaxBurst = <max number of concurrent requests allowed to the service>
|
|
Routes = [ # Configure RequestsPerSecond and one route matching parameter for each member of the list
|
|
{
|
|
PathExact = "<limits apply paths that match this value>"
|
|
PathPrefix = "<limits apply to paths starting with this value>"
|
|
PathRegex = "<limits apply to paths that match this expression>"
|
|
RequestsPerSecond = <overrides top-level limit>
|
|
RequestMaxBurst = <overrides top-level limit>
|
|
}
|
|
]
|
|
}
|
|
UpstreamConfig = {
|
|
Overrides = {
|
|
Name = "<name of upstreams to override>"
|
|
Namespace = "<namespace containing upstreams to override>"
|
|
Peer = "<peer name of upstream service>"
|
|
Protocol = "<protocol for requests to upstream services>"
|
|
ConnectTimeoutMs = 5000
|
|
MeshGateway = {
|
|
mode = "<mode for mesh gateway services>"
|
|
}
|
|
BalanceOutboundConnections = "exact_balance"
|
|
Limits = {
|
|
MaxConnections = 0
|
|
MaxPendingRequests = 0
|
|
MaxConcurrentRequests = 0
|
|
}
|
|
PassiveHealthCheck = {
|
|
Interval = "0s"
|
|
MaxFailures = 0
|
|
EnforcingConsecutive5xx = 100
|
|
MaxEjectionPercent = 10
|
|
BaseEjectionTime = "30s"
|
|
}
|
|
}
|
|
Defaults = {
|
|
Protocol = "<default protocol for upstream traffic>"
|
|
ConnectTimeoutMs = 5000
|
|
MeshGateway = {
|
|
mode = "<default mode for mesh gateway services>"
|
|
}
|
|
BalanceOutboundConnections = "exact_balance"
|
|
Limits = {
|
|
MaxConnections = 0
|
|
MaxPendingRequests = 0
|
|
MaxConcurrentRequests = 0
|
|
}
|
|
PassiveHealthCheck = {
|
|
Interval = "0s"
|
|
MaxFailures = 0
|
|
EnforcingConsecutive5xx = 100
|
|
MaxEjectionPercent = 10
|
|
BaseEjectionTime = "30s"
|
|
}
|
|
}
|
|
}
|
|
TransparentProxy = {
|
|
OutboundListenerPort = 15001
|
|
DialedDirectly = false
|
|
}
|
|
MutualTLSMode = "strict" # only supported when services are in transparent proxy mode
|
|
EnvoyExtensions = [
|
|
{
|
|
Name = "<name of the extension>"
|
|
Required = `false`
|
|
Arguments = { <specific to each extension> }
|
|
ConsulVersion = "<Consul version constraint for applying the extension>"
|
|
EnvoyVersion = "<Envoy version constraint for applying the extension>"
|
|
}
|
|
]
|
|
Destination = {
|
|
Addresses = [
|
|
"<first IP address>",
|
|
"<second IP address>"
|
|
]
|
|
Port = 0
|
|
}
|
|
MaxInboundConnections = 0
|
|
LocalConnectTimeoutMs = 5000
|
|
LocalRequestTimeoutMs = "15s"
|
|
MeshGateway = {
|
|
Mode = "<default mode for mesh gateway services>"
|
|
}
|
|
ExternalSNI = "<SNI server host>"
|
|
Expose = {
|
|
Checks = false
|
|
Paths = [
|
|
{
|
|
Path = "</local/dir>"
|
|
LocalPathPort = 0
|
|
LocalListenerPort = 0
|
|
Protocol = "http"
|
|
}
|
|
]
|
|
}
|
|
```
|
|
</Tab>
|
|
|
|
<Tab heading="YAML">
|
|
|
|
```yaml
|
|
apiVersion: consul.hashicorp.com/v1alpha1
|
|
kind: ServiceDefaults
|
|
metadata:
|
|
name: <name of the service you are configuring>
|
|
namespace: default
|
|
spec:
|
|
protocol: tcp
|
|
balanceInboundConnections: exact_balance
|
|
mode: <mode for directing traffic>
|
|
rateLimits:
|
|
instanceLevel:
|
|
requestsPerSecond: <average number of requests allowed to the service per second>
|
|
requestsMaxBurst: <max number of concurrent requests allowed to the service>
|
|
routes: # Configure RequestsPerSecond and one route matching parameter for each member of the list
|
|
- pathExact: "<limits apply paths that match this value>"
|
|
pathPrefix: "<limits apply to paths starting with this value>"
|
|
pathRegex: "<limits apply to paths that match this expression>"
|
|
requestsPerSecond: <overrides top-level limit>
|
|
requestMaxBurst: <overrides top-level limit>
|
|
upstreamConfig:
|
|
overrides:
|
|
- name: <name of upstreams to override>
|
|
namespace: <namespace containing upstream>
|
|
peer: <peer name of upstream services>
|
|
protocol: <protocol for the upstream listener>
|
|
connectTimeoutMs: 5000
|
|
meshGateway:
|
|
mode: <mode for mesh gateway services>
|
|
balanceOutboundConnections: exact_balance
|
|
limits:
|
|
maxConnections: 0
|
|
maxPendingRequests: 0
|
|
maxConcurrentRequests: 0
|
|
passiveHealthCheck:
|
|
interval: "0s"
|
|
maxFailures: 0
|
|
enforcingConsecutive5xx: 100
|
|
maxEjectionPercent: 10
|
|
baseEjectionTime: "30s"
|
|
defaults:
|
|
protocol: <default protocol for the upstream listener>
|
|
connectTimeoutMs: 5000
|
|
meshGateway:
|
|
mode: <default mode for mesh gateway services>
|
|
balanceOutboundConnections: exact_balance
|
|
limits:
|
|
maxConnections: 0
|
|
maxPendingRequests: 0
|
|
maxConcurrentRequests: 0
|
|
passiveHealthCheck:
|
|
interval: "0s"
|
|
maxFailures: 0
|
|
enforcingConsecutive5xx: 100
|
|
maxEjectionPercent: 10
|
|
baseEjectionTime: "30s"
|
|
transparentProxy:
|
|
outboundListenerPort: 15001
|
|
dialedDirectly: false
|
|
mutualTLSMode: strict
|
|
envoyExtensions:
|
|
- name: <name of the extension>
|
|
required: false
|
|
arguments:
|
|
- <specific to each extension>
|
|
consulVersion: <Consul version constraint for applying the extension>
|
|
envoyVersion: <Envoy version constraint for applying the extension>
|
|
destination:
|
|
addresses:
|
|
- <first hostname or IP address>
|
|
<second hostname or IP address>
|
|
port: 0
|
|
maxInboundConnections: 0
|
|
meshGateway:
|
|
mode: <default mode for mesh gateway services>
|
|
externalSNI: <name of TLS SNI outside o f the mesh>
|
|
expose:
|
|
checks: false
|
|
paths:
|
|
- path: <HTTP path to expose through Envoy>
|
|
localPathPort: 0
|
|
listenerPort: 0
|
|
protocol: http
|
|
```
|
|
|
|
</Tab>
|
|
|
|
<Tab heading="JSON">
|
|
|
|
```json
|
|
{
|
|
"Kind": "ServiceDefaults",
|
|
"Name": "<name of the service you are configuring>",
|
|
"Namespace": "default",
|
|
"Partition": "default",
|
|
"Meta": {
|
|
"<Keys>": "<Values>"
|
|
},
|
|
"Protocol": "tcp",
|
|
"BalanceInboundConnections": "exact_balance",
|
|
"Mode": "<mode for directing traffic>",
|
|
"RateLimits": {
|
|
"InstanceLevel": {
|
|
"RequestsPerSecond": 100,
|
|
"RequestsMaxBurst": 200,
|
|
"Routes": [
|
|
{
|
|
"PathExact": "<limits apply paths that match this value>",
|
|
"PathPrefix": "<limits apply to paths starting with this value>",
|
|
"PathRegex": "<limits apply to paths that match this expression>",
|
|
"RequestsPerSecond": 50,
|
|
"RequestMaxBurst": 100
|
|
}
|
|
]
|
|
}
|
|
},
|
|
"UpstreamConfig": {
|
|
"Overrides": [{
|
|
"Name": "<name of upstream>",
|
|
"Namespace": "<namespace containing upstream - Consul Enterprise>",
|
|
"Peer": "<peer name of the upstream service>",
|
|
"Protocol": "<protocol for the upstream listener>",
|
|
"ConnectTimeoutMs": 5000,
|
|
"MeshGateway": {
|
|
"Mode": "<mode for mesh gateway services>"
|
|
},
|
|
"BalanceOutboundConnections": "exact_balance",
|
|
"Limits": {
|
|
"MaxConnections": 0,
|
|
"MaxPendingRequests": 0,
|
|
"MaxConcurrentRequests": 0
|
|
},
|
|
"PassiveHealthCheck": {
|
|
"Interval": "0s",
|
|
"MaxFailures": 0,
|
|
"EnforcingConsecutive5xx": 100,
|
|
"MaxEjectionPercent": 10,
|
|
"BaseEjectionTime": "30s"
|
|
}
|
|
}]
|
|
},
|
|
"Defaults": {
|
|
"Protocol": "<protocol for the upstream listener>",
|
|
"ConnectTimeoutMs": 5000,
|
|
"MeshGateway": {
|
|
"Mode": "<mode for mesh gateway services>"
|
|
},
|
|
"BalanceOutboundConnections": "exact_balance",
|
|
"Limits": {
|
|
"MaxConnections": 0,
|
|
"MaxPendingRequests": 0,
|
|
"MaxConcurrentRequests": 0
|
|
},
|
|
"PassiveHealthCheck": {
|
|
"Interval": "0s",
|
|
"MaxFailures": 0,
|
|
"EnforcingConsecutive5xx": 100,
|
|
"MaxEjectionPercent": 10,
|
|
"BaseEjectionTime": "30s"
|
|
}
|
|
},
|
|
"TransparentProxy": {
|
|
"OutboundListenerPort": 15001,
|
|
"DialedDirectly": false
|
|
},
|
|
"MutualTLSMode": "strict",
|
|
"EnvoyExtensions": [{
|
|
"Name": "<name of the extension>",
|
|
"Required": false,
|
|
"Arguments": {
|
|
"<specific to each extension>": "<specific to each extension>"
|
|
},
|
|
"ConsulVersion": "<Consul version constraint for applying the extension>",
|
|
"EnvoyVersion": "<Envoy version constraint for applying the extension>"
|
|
}],
|
|
"Destination": {
|
|
"Addresses": [
|
|
"<first hostname or IP address>",
|
|
"<second hostname or IP address>"
|
|
],
|
|
"Port": 0
|
|
},
|
|
"MaxInboundConnections": 0,
|
|
"MeshGateway": {
|
|
"Mode": "<mode for mesh gateway services>"
|
|
},
|
|
"ExternalSNI": "<name of TLS SNI outside o f the mesh>",
|
|
"Expose": {
|
|
"Checks": false,
|
|
"Paths": [{
|
|
"Path": "<HTTP path to expose through Envoy>",
|
|
"LocalPathPort": 0,
|
|
"ListenerPort": 0,
|
|
"Protocol": "http"
|
|
}]
|
|
}
|
|
}
|
|
```
|
|
|
|
</Tab>
|
|
|
|
</Tabs>
|
|
|
|
## Specification
|
|
|
|
This section provides details about the fields you can configure in the service defaults configuration entry.
|
|
|
|
<Tabs>
|
|
<Tab heading="HCL and JSON" group="hcl">
|
|
|
|
### `Kind`
|
|
|
|
Specifies the configuration entry type. The value must be set to `service-defaults`.
|
|
|
|
#### Values
|
|
|
|
- Default: none
|
|
- This field is required.
|
|
- Data type: String value that must be set to `service-defaults`.
|
|
|
|
### `Name`
|
|
|
|
Specifies the name of the service you are setting the defaults for.
|
|
|
|
#### Values
|
|
|
|
- Default: none
|
|
- This field is required.
|
|
- Data type: String
|
|
|
|
### `Namespace` <EnterpriseAlert inline/>
|
|
|
|
Specifies the Consul namespace that the configuration entry applies to.
|
|
|
|
#### Values
|
|
|
|
- Default: `default`
|
|
- Data type: String
|
|
|
|
### `Partition` <EnterpriseAlert inline/>
|
|
|
|
Specifies the name of the Consul admin partition that the configuration entry applies to. Refer to [Admin Partitions](/consul/docs/enterprise/admin-partitions) for additional information.
|
|
|
|
#### Values
|
|
|
|
- Default: `default`
|
|
- Data type: String
|
|
|
|
### `Meta`
|
|
|
|
Specifies a set of custom key-value pairs to add to the [Consul KV](/consul/docs/dynamic-app-config/kv) store.
|
|
|
|
#### Values
|
|
|
|
- Default: none
|
|
- Data type: Map of one or more key-value pairs.
|
|
- keys: String
|
|
- values: String, integer, or float
|
|
|
|
### `Protocol`
|
|
|
|
Specifies the default protocol for the service. In service mesh use cases, the `protocol` configuration is required to enable the following features and components:
|
|
|
|
- [observability](/consul/docs/connect/observability)
|
|
- [service splitter configuration entry](/consul/docs/connect/config-entries/service-splitter)
|
|
- [service router configuration entry](/consul/docs/connect/config-entries/service-router)
|
|
- [L7 intentions](/consul/docs/connect/intentions#l7-traffic-intentions)
|
|
|
|
You can set the global protocol for proxies in the [`proxy-defaults`](/consul/docs/connect/config-entries/proxy-defaults#default-protocol) configuration entry, but the protocol specified in the `service-defaults` configuration entry overrides the `proxy-defaults` configuration.
|
|
|
|
#### Values
|
|
|
|
- Default: `tcp`
|
|
- You can specify one of the following String values:
|
|
- `tcp` (default)
|
|
- `http`
|
|
- `http2`
|
|
- `grpc`
|
|
|
|
Refer to [Set the default protocol](#set-the-default-protocol) for an example configuration.
|
|
|
|
### `BalanceInboundConnections`
|
|
|
|
Specifies the strategy for allocating inbound connections to the service across Envoy proxy threads. The only supported value is `exact_balance`. By default, no connections are balanced. Refer to the [Envoy documentation](https://cloudnative.to/envoy/api-v3/config/listener/v3/listener.proto.html#config-listener-v3-listener-connectionbalanceconfig) for details.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `Mode`
|
|
|
|
Specifies a mode for how the service directs inbound and outbound traffic.
|
|
|
|
- Default: None
|
|
- You can specify the following string values:
|
|
- `direct`: The proxy's listeners must be dialed directly by the local application and other proxies.
|
|
- `transparent`: The service captures inbound and outbound traffic and redirects it through the proxy. The mode does not enable the traffic redirection. It instructs Consul to configure Envoy as if traffic is already being redirected.
|
|
|
|
### `RateLimits` <EnterpriseAlert inline />
|
|
|
|
Map containing an instance-level configuration for limiting the service's traffic rate.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `RateLimits{}.InstanceLevel`
|
|
|
|
Map containing a set of request rate limit configurations for instances of the service.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `RateLimits{}.InstanceLevel{}.RequestsPerSecond`
|
|
|
|
Specifies the average number of requests per second allowed to the service. If the [`RequestsMaxBurst`](#ratelimits-instancelevel-requestsmaxburst) parameter is set, the number of requests per second to the service can temporarily exceed the limit specified in the `RequestsPerSecond` up to the value specified in `RequestsMaxBurst`. Internally, this is the refill rate of the token bucket used for rate limiting.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Integer
|
|
|
|
### `RateLimits{}.InstanceLevel{}.RequestsMaxBurst`
|
|
|
|
Specifies the maximum number of concurrent requests momentarily allowed to the service. When the limit is reached, Consul blocks additional requests. You must specify a value equal to or greater than the [`RequestsPerSecond`](#ratelimits-instancelevel-requestspersecond) parameter. If unspecified, this parameter defaults to `RequestsPerSecond`. Internally, this is the maximum size of the token bucket used for rate limiting.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Integer
|
|
|
|
### `RateLimits{}.InstanceLevel{}.Routes`
|
|
|
|
Specifies a list of rate limiting configurations to apply to specific routes to the service. Each member of the `Routes` list must configure the `RequestsPerSecond` parameter and one of the following route-matching parameters:
|
|
|
|
- `PathExact`
|
|
- `PathPrefix`
|
|
- `PathRegex`
|
|
|
|
Consul applies the rate limit configuration to the first matching route for each request. Refer to [Examples](#enable-request-rate-limit-on-a-prefixed-path) for example configurations.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
The following table describes the parameters you can specify in the `Routes` map:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `PathExact` | Specifies the exact path to match on the request path. When using this field, do not configure `PathPrefix` or `PathRegex` in the same `Routes` map. | String | None |
|
|
| `PathPrefix` | Specifies the path prefix to match on the request path. When using this field, do not configure `PathExact` or `PathRegex` in the same `Routes` map. | String | None |
|
|
| `PathRegex` | Specifies a regular expression to match on the request path. When using this field, do not configure `PathExact` or `PathPrefix` in the same `Routes` map. The syntax is proxy-specific. When [using Envoy](/consul/docs/connect/proxies/envoy), refer to [the documentation for Envoy v1.11.2 or newer](https://github.com/google/re2/wiki/Syntax) or [the documentation for Envoy v1.11.1 or older](https://en.cppreference.com/w/cpp/regex/ecmascript), depending on the version of Envoy you use. | String | None |
|
|
| `RequestsPerSecond` | Specifies the average number of requests per second allowed to the service. Overrides the [`RequestsPerSecond`](#ratelimits-instancelevel-requestspersecond) parameter specified for the service. | Integer | None |
|
|
| `RequestsMaxBurst` | Specifies the maximum number of concurrent requests temporarily allowed to the service. When the limit is reached, Consul blocks additional requests. You must specify a value equal to or greater than the `Routes.RequestsPerSecond` parameter. Overrides the [`RequestsMaxBurst`](#ratelimits-instancelevel-requestsmaxburst) parameter specified for the service. | Integer | None |
|
|
|
|
### `UpstreamConfig`
|
|
|
|
Controls default upstream connection settings and custom overrides for individual upstream services. If your network contains federated datacenters, individual upstream configurations apply to all pairs of source and upstream destination services in the network. Refer to the following fields for details:
|
|
|
|
- [`UpstreamConfig.Overrides`](#upstreamconfig-overrides)
|
|
- [`UpstreamConfig.Defaults`](#upstreamconfig-defaults)
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `UpstreamConfig.Overrides[]`
|
|
|
|
Specifies options that override the [default upstream configurations](#upstreamconfig-defaults) for individual upstreams.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: List
|
|
|
|
### `UpstreamConfig.Overrides[].Name`
|
|
|
|
Specifies the name of the upstream service that the configuration applies to. We recommend that you do not use the `*` wildcard to avoid applying the configuration to unintended upstreams.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `UpstreamConfig.Overrides[].Namespace` <EnterpriseAlert inline/>
|
|
|
|
Specifies the namespace containing the upstream service that the configuration applies to. Do not use the `*` wildcard to prevent the configuration from applying to unintended upstreams.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `UpstreamConfig.Overrides[].Peer`
|
|
|
|
Specifies the peer name of the upstream service that the configuration applies to. The `*` wildcard is not supported.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `UpstreamConfig.Overrides[].Protocol`
|
|
Specifies the protocol to use for requests to the upstream listener.
|
|
|
|
We recommend configuring the protocol in the main [`Protocol`](#protocol) field of the configuration entry so that you can leverage [L7 features](/consul/docs/connect/manage-traffic). Setting the protocol in an upstream configuration limits L7 management functionality.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
|
|
### `UpstreamConfig.Overrides[].ConnectTimeoutMs`
|
|
|
|
Specifies how long in milliseconds that the service should attempt to establish an upstream connection before timing out.
|
|
|
|
We recommend configuring the upstream timeout in the [`connection_timeout`](/consul/docs/connect/config-entries/service-resolver#connecttimeout) field of the `service-resolver` configuration entry for the upstream destination service. Doing so enables you to leverage [L7 features](/consul/docs/connect/manage-traffic). Configuring the timeout in the `service-defaults` upstream configuration limits L7 management functionality.
|
|
|
|
#### Values
|
|
|
|
- Default: `5000`
|
|
- Data type: Integer
|
|
|
|
### `UpstreamConfig.Overrides[].MeshGateway`
|
|
|
|
Map that contains the default mesh gateway `mode` field for the upstream. Refer to [Service Mesh Proxy Configuration](/consul/docs/connect/gateways/mesh-gateway#service-mesh-proxy-configuration) in the mesh gateway documentation for additional information.
|
|
|
|
#### Values
|
|
|
|
- Default: `none`
|
|
- You can specify the following string values for the `mode` field:
|
|
- `none`: The service does not make outbound connections through a mesh gateway. Instead, the service makes outbound connections directly to the destination services.
|
|
- `local`: The service mesh proxy makes an outbound connection to a gateway running in the same datacenter.
|
|
- `remote`: The service mesh proxy makes an outbound connection to a gateway running in the destination datacenter.
|
|
|
|
|
|
### `UpstreamConfig.Overrides[].BalanceOutboundConnections`
|
|
|
|
Sets the strategy for allocating outbound connections from the upstream across Envoy proxy threads.
|
|
|
|
#### Values
|
|
|
|
The only supported value is `exact_balance`. By default, no connections are balanced. Refer to the [Envoy documentation](https://cloudnative.to/envoy/api-v3/config/listener/v3/listener.proto.html#config-listener-v3-listener-connectionbalanceconfig) for details.
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `UpstreamConfig.Overrides[].Limits`
|
|
|
|
Map that specifies a set of limits to apply to when connecting to individual upstream services.
|
|
|
|
#### Values
|
|
|
|
The following table describes limits you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `MaxConnections` | Specifies the maximum number of connections a service instance can establish against the upstream. Define this limit for HTTP/1.1 traffic. | Integer | `0` |
|
|
| `MaxPendingRequests` | Specifies the maximum number of requests that are queued while waiting for a connection to establish. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
| `MaxConcurrentRequests` | Specifies the maximum number of concurrent requests. Define this limit for HTTP/2 traffic. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
|
|
Refer to the [upstream configuration example](#upstream-configuration) for additional guidance.
|
|
|
|
### `UpstreamConfig.Overrides[].PassiveHealthCheck`
|
|
|
|
Map that specifies a set of rules that enable Consul to remove hosts from the upstream cluster that are unreachable or that return errors.
|
|
|
|
#### Values
|
|
|
|
The following table describes passive health check parameters you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `Interval` | Specifies the time between checks. | String | `0s` |
|
|
| `MaxFailures` | Specifies the number of consecutive failures allowed per check interval. If exceeded, Consul removes the host from the load balancer. | Integer | `0` |
|
|
| `EnforcingConsecutive5xx` | Specifies a percentage that indicates how many times out of 100 that Consul ejects the host when it detects an outlier status. The outlier status is determined by consecutive errors in the 500-599 response range. | Integer | `100` |
|
|
| `MaxEjectionPercent` | Specifies the maximum percentage of an upstream cluster that Consul ejects when the proxy reports an outlier. Consul ejects at least one host when an outlier is detected regardless of the value. | Integer | `10` |
|
|
| `BaseEjectionTime` | Specifies the minimum amount of time that an ejected host must remain outside the cluster before rejoining. The real time is equal to the value of the `BaseEjectionTime` multiplied by the number of times the host has been ejected. | String | `30s` |
|
|
|
|
### `UpstreamConfig.Defaults`
|
|
|
|
Specifies configurations that set default upstream settings. For information about overriding the default configurations for in for individual upstreams, refer to [`UpstreamConfig.Overrides`](#upstreamconfig-overrides).
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `UpstreamConfig.Defaults.Protocol`
|
|
|
|
Specifies default protocol for upstream listeners.
|
|
|
|
We recommend configuring the protocol in the main [`Protocol`](#protocol) field of the configuration entry so that you can leverage [L7 features](/consul/docs/connect/manage-traffic). Setting the protocol in an upstream configuration limits L7 management functionality.
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `UpstreamConfig.Defaults.ConnectTimeoutMs`
|
|
|
|
Specifies how long in milliseconds that all services should continue attempting to establish an upstream connection before timing out.
|
|
|
|
For non-Kubernetes environments, we recommend configuring the upstream timeout in the [`connection_timeout`](/consul/docs/connect/config-entries/service-resolver#connecttimeout) field of the `service-resolver` configuration entry for the upstream destination service. Doing so enables you to leverage [L7 features](/consul/docs/connect/manage-traffic). Configuring the timeout in the `service-defaults` upstream configuration limits L7 management functionality.
|
|
|
|
- Default: `5000`
|
|
- Data type: Integer
|
|
|
|
### `UpstreamConfig.Defaults.MeshGateway`
|
|
|
|
Specifies the default mesh gateway `mode` field for all upstreams. Refer to [Service Mesh Proxy Configuration](/consul/docs/connect/gateways/mesh-gateway#service-mesh-proxy-configuration) in the mesh gateway documentation for additional information.
|
|
|
|
You can specify the following string values for the `mode` field:
|
|
|
|
- `None`: The service does not make outbound connections through a mesh gateway. Instead, the service makes outbound connections directly to the destination services.
|
|
- `local`: The service mesh proxy makes an outbound connection to a gateway running in the same datacenter.
|
|
- `remote`: The service mesh proxy makes an outbound connection to a gateway running in the destination datacenter.
|
|
|
|
### `UpstreamConfig.Defaults.BalanceOutboundConnections`
|
|
|
|
Sets the strategy for allocating outbound connections from upstreams across Envoy proxy threads. The only supported value is `exact_balance`. By default, no connections are balanced. Refer to the [Envoy documentation](https://cloudnative.to/envoy/api-v3/config/listener/v3/listener.proto.html#config-listener-v3-listener-connectionbalanceconfig) for details.
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `UpstreamConfig.Defaults.Limits`
|
|
|
|
Map that specifies a set of limits to apply to when connecting upstream services. The following table describes limits you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `MaxConnections` | Specifies the maximum number of connections a service instance can establish against the upstream. Define this limit for HTTP/1.1 traffic. | Integer | `0` |
|
|
| `MaxPendingRequests` | Specifies the maximum number of requests that are queued while waiting for a connection to establish. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
| `MaxConcurrentRequests` | Specifies the maximum number of concurrent requests. Define this limit for HTTP/2 traffic. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
|
|
### `UpstreamConfig.Defaults.PassiveHealthCheck`
|
|
|
|
Map that specifies a set of rules that enable Consul to remove hosts from the upstream cluster that are unreachable or that return errors. The following table describes the health check parameters you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `Interval` | Specifies the time between checks. | String | `0s` |
|
|
| `MaxFailures` | Specifies the number of consecutive failures allowed per check interval. If exceeded, Consul removes the host from the load balancer. | Integer | `0` |
|
|
| `EnforcingConsecutive5xx ` | Specifies a percentage that indicates how many times out of 100 that Consul ejects the host when it detects an outlier status. The outlier status is determined by consecutive errors in the 500-599 response range. | Integer | `100` |
|
|
| `MaxEjectionPercent` | Specifies the maximum percentage of an upstream cluster that Consul ejects when the proxy reports an outlier. Consul ejects at least one host when an outlier is detected regardless of the value. | Integer | `10` |
|
|
| `BaseEjectionTime` | Specifies the minimum amount of time that an ejected host must remain outside the cluster before rejoining. The real time is equal to the value of the `BaseEjectionTime` multiplied by the number of times the host has been ejected. | String | `30s` |
|
|
|
|
### `TransparentProxy`
|
|
|
|
Controls configurations specific to proxies in transparent mode. Refer to [Transparent Proxy Mode](/consul/docs/k8s/connect/transparent-proxy) for additional information.
|
|
|
|
You can configure the following parameters in the `TransparentProxy` block:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `OutboundListenerPort` | Specifies the port that the proxy listens on for outbound traffic. This must be the same port number where outbound application traffic is redirected. | Integer | `15001` |
|
|
| `DialedDirectly` | Enables transparent proxies to dial the proxy instance's IP address directly when set to `true`. Transparent proxies commonly dial upstreams at the `"virtual"` tagged address, which load balances across instances. Dialing individual instances can be helpful for stateful services, such as a database cluster with a leader. | Boolean | `false` |
|
|
|
|
### `MutualTLSMode`
|
|
|
|
Controls whether mutual TLS is required for incoming connections to this service. This setting is
|
|
only supported for services with transparent proxy enabled. We recommend only using `permissive`
|
|
mode if necessary while onboarding services to the service mesh.
|
|
|
|
You can specify the following string values for the `MutualTLSMode` field:
|
|
|
|
- `""`: When this field is empty, the value is inherited from the `proxy-defaults` config entry.
|
|
- `strict`: The sidecar proxy requires mutual TLS for incoming traffic.
|
|
- `permissive`: The sidecar proxy accepts mutual TLS traffic on the sidecar proxy service port,
|
|
and accepts any traffic on the destination service's port.
|
|
|
|
### `EnvoyExtensions`
|
|
|
|
List of extensions to modify Envoy proxy configuration. Refer to [Envoy Extensions](/consul/docs/connect/proxies/envoy-extensions) for additional information.
|
|
|
|
The following table describes how to configure values in the `EnvoyExtensions` map:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `Name` | Specifies the name of the extension. | String | None |
|
|
| `Required` | Specify `true` to require the extension to apply successfully. <p>Use this parameter to ensure that extensions required for secure communication are not unintentionally bypassed.</p><p>When Envoy fails to apply a required extension, Consul logs an error and skips all extensions, leaving xDS resources unchanged. </p> | String | None |
|
|
| `Arguments` | Specifies the arguments to pass to the extension. Refer to the documentation for the extension you want to implement for additional information. | Map | None |
|
|
| `ConsulVersion` | Specifies the Consul [version constraint](https://github.com/hashicorp/go-version) for the extension. Consul validates the version constraint against the runtime version during xDS updates. If a non-matching version is in use, Consul logs and skips the extension. <p>Use this parameter to avoid upgrade issues when a configured extension is not compatible with a new version of Consul.</p> | String | None |
|
|
| `EnvoyVersion` | Specifies the Envoy [version constraint](https://github.com/hashicorp/go-version) for the extension. Consul validates the version constraint against the version of the running Envoy proxy during xDS updates. If a non-matching version is in use, Consul logs and skips the extension. <p>Use this parameter to avoid upgrade issues when a configured extension is not compatible with a new version of Envoy.</p> | String | None |
|
|
|
|
### `Destination{}`
|
|
|
|
Configures the destination for service traffic through terminating gateways. Refer to [Terminating Gateway](/consul/docs/connect/gateways/terminating-gateway) for additional information.
|
|
|
|
To use the `Destination` block, proxy services must be in transparent proxy mode. Refer to [Enable transparent proxy mode](/consul/docs/k8s/connect/transparent-proxy/enable-transparent-proxy) for additional information.
|
|
|
|
You can configure the following parameters in the `Destination` block:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `Addresses` | Specifies a list of addresses for the destination. You can configure a list of hostnames and IP addresses. Wildcards are not supported. | List | None |
|
|
| `Port` | Specifies the port number of the destination. | Integer | `0` |
|
|
|
|
### `MaxInboundConnections`
|
|
|
|
Specifies the maximum number of concurrent inbound connections to each service instance.
|
|
|
|
- Default: `0`
|
|
- Data type: Integer
|
|
|
|
### `LocalConnectTimeoutMs`
|
|
|
|
Specifies the number of milliseconds allowed for establishing connections to the local application instance before timing out.
|
|
|
|
- Default: `5000`
|
|
- Data type: Integer
|
|
|
|
### `LocalRequestTimeoutMs`
|
|
|
|
Specifies the timeout for HTTP requests to the local application instance. Applies to HTTP-based protocols only. If not specified, inherits the Envoy default for route timeouts.
|
|
|
|
- Default: Inherits `15s` from Envoy as the default
|
|
- Data type: String
|
|
|
|
### `MeshGateway`
|
|
|
|
Specifies the default mesh gateway `mode` field for the service. Refer to [Service Mesh Proxy Configuration](/consul/docs/connect/gateways/mesh-gateway#service-mesh-proxy-configuration) in the mesh gateway documentation for additional information.
|
|
|
|
You can specify the following string values for the `mode` field:
|
|
|
|
- `none`: The service does not make outbound connections through a mesh gateway. Instead, the service makes outbound connections directly to the destination services.
|
|
- `local`: The service mesh proxy makes an outbound connection to a gateway running in the same datacenter.
|
|
- `remote`: The service mesh proxy makes an outbound connection to a gateway running in the destination datacenter.
|
|
|
|
### `ExternalSNI`
|
|
|
|
Specifies the TLS server name indication (SNI) when federating with an external system.
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `Expose`
|
|
|
|
Specifies default configurations for exposing HTTP paths through Envoy. Exposing paths through Envoy enables services to listen on `localhost` only. Applications that are not Consul service mesh-enabled can still contact an HTTP endpoint. Refer to [Expose Paths Configuration Reference](/consul/docs/proxies/proxy-config-reference#expose-paths-configuration-reference) for additional information and example configurations.
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `Expose.Checks`
|
|
|
|
Exposes all HTTP and gRPC checks registered with the agent if set to `true`. Envoy exposes listeners for the checks and only accepts connections originating from localhost or Consul's [`advertise_addr`](/consul/docs/agent/config/config-files#advertise_addr). The ports for the listeners are dynamically allocated from the agent's [`expose_min_port`](/consul/docs/agent/config/config-files#expose_min_port) and [`expose_max_port`](/consul/docs/agent/config/config-files#expose_max_port) configurations.
|
|
|
|
We recommend enabling the `Checks` configuration when a Consul client cannot reach registered services over localhost, such as when Consul agents run in their own pods in Kubernetes.
|
|
|
|
- Default: `false`
|
|
- Data type: Boolean
|
|
|
|
### `Expose.Paths[]`
|
|
|
|
Specifies a list of configuration maps that define paths to expose through Envoy when `Expose.Checks` is set to `true`. You can configure the following parameters for each map in the list:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `Path` | Specifies the HTTP path to expose. You must prepend the path with a forward slash (`/`). | String | None |
|
|
| `LocalPathPort` | Specifies the port where the local service listens for connections to the path. | Integer | `0` |
|
|
| `ListenPort` | Specifies the port where the proxy listens for connections. The port must be available. If the port is unavailable, Envoy does not expose a listener for the path and the proxy registration still succeeds. | Integer | `0` |
|
|
| `Protocol` | Specifies the protocol of the listener. You can configure one of the following values: <li>`http`</li><li>`http2`: Use with gRPC traffic</li> | Integer | `http` |
|
|
|
|
</Tab>
|
|
|
|
<Tab heading="YAML" group="yaml">
|
|
|
|
### `apiVersion`
|
|
|
|
Specifies the version of the Consul API for integrating with Kubernetes. The value must be `consul.hashicorp.com/v1alpha1`. The `apiVersion` field is not supported for non-Kubernetes deployments.
|
|
|
|
- Default: None
|
|
- This field is required.
|
|
- String value that must be set to `consul.hashicorp.com/v1alpha1`.
|
|
|
|
### `kind`
|
|
|
|
Specifies the configuration entry type. Must be ` ServiceDefaults`.
|
|
|
|
- Required: required
|
|
- String value that must be set to `ServiceDefaults`.
|
|
|
|
### `metadata`
|
|
|
|
Map that contains the service name, namespace, and admin partition that the configuration entry applies to.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Map containing the following strings:
|
|
- [`name`](#name)
|
|
- [`namespace`](#namespace)
|
|
- [`partition`](#partition)
|
|
|
|
|
|
### `metadata.name`
|
|
|
|
Specifies the name of the service you are setting the defaults for.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- This field is required
|
|
- Data type: String
|
|
|
|
### `metadata.namespace` <EnterpriseAlert inline/>
|
|
|
|
Specifies the Consul namespace that the configuration entry applies to. Refer to [Consul Enterprise](/consul/docs/k8s/crds#consul-enterprise) for information about how Consul namespaces map to Kubernetes Namespaces. Consul Community Edition (Consul CE) ignores the `metadata.namespace` configuration.
|
|
|
|
- Default: `default`
|
|
- Data type: String
|
|
|
|
### `spec`
|
|
|
|
Map that contains the details about the `ServiceDefaults` configuration entry. The `apiVersion`, `kind`, and `metadata` fields are siblings of the `spec` field. All other configurations are children.
|
|
|
|
### `spec.protocol`
|
|
|
|
Specifies the default protocol for the service. In service service mesh use cases, the `protocol` configuration is required to enable the following features and components:
|
|
|
|
- [observability](/consul/docs/connect/observability)
|
|
- [`service-splitter` configuration entry](/consul/docs/connect/config-entries/service-splitter)
|
|
- [`service-router` configuration entry](/consul/docs/connect/config-entries/service-router)
|
|
- [L7 intentions](/consul/docs/connect/intentions#l7-traffic-intentions)
|
|
|
|
You can set the global protocol for proxies in the [`ProxyDefaults` configuration entry](/consul/docs/connect/config-entries/proxy-defaults#default-protocol), but the protocol specified in the `ServiceDefaults` configuration entry overrides the `ProxyDefaults` configuration.
|
|
|
|
#### Values
|
|
|
|
- Default: `tcp`
|
|
- You can specify one of the following string values:
|
|
- `tcp`
|
|
- `http`
|
|
- `http2`
|
|
- `grpc`
|
|
|
|
Refer to [Set the default protocol](#set-the-default-protocol) for an example configuration.
|
|
|
|
### `spec.balanceInboundConnections`
|
|
|
|
Specifies the strategy for allocating inbound connections to the service across Envoy proxy threads. The only supported value is `exact_balance`. By default, no connections are balanced. Refer to the [Envoy documentation](https://cloudnative.to/envoy/api-v3/config/listener/v3/listener.proto.html#config-listener-v3-listener-connectionbalanceconfig) for details.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.mode`
|
|
|
|
Specifies a mode for how the service directs inbound and outbound traffic.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Required: optional
|
|
- You can specified the following string values:
|
|
|
|
- `direct`: The proxy's listeners must be dialed directly by the local application and other proxies.
|
|
- `transparent`: The service captures inbound and outbound traffic and redirects it through the proxy. The mode does not enable the traffic redirection. It instructs Consul to configure Envoy as if traffic is already being redirected.
|
|
|
|
### `spec.rateLimits` <EnterpriseAlert inline />
|
|
|
|
Map containing an instance-level configuration for limiting the service's traffic rate.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `spec.rateLimits{}.instanceLevel`
|
|
|
|
Map containing a set of request rate limit configurations for instances of the service.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
### `spec.rateLimits{}.instanceLevel{}.requestsPerSecond`
|
|
|
|
Specifies the average number of requests per second allowed to the service. If the [`requestsMaxBurst`](#spec-ratelimits-instancelevel-requestsmaxburst) parameter is set, the number of requests per second to the service can temporarily exceed the limit specified in the `requestsPerSecond` up to the value specified in `requestsMaxBurst`. Internally, this is the refill rate of the token bucket used for rate limiting.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Integer
|
|
|
|
### `spec.rateLimits{}.instanceLevel{}.requestsMaxBurst`
|
|
|
|
Specifies the maximum number of concurrent requests momentarily allowed to the service. When the limit is reached, Consul blocks additional requests. You must specify a value equal to or greater than the [`requestsPerSecond`](#spec-ratelimits-instancelevel-requestspersecond) parameter. If unspecified, this parameter defaults to `requestsPerSecond`. Internally, this is the maximum size of the token bucket used for rate limiting.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Integer
|
|
|
|
### `spec.rateLimits{}.instanceLevel{}.routes`
|
|
|
|
Specifies a list of rate limiting configurations to apply to specific routes to the service. Each member of the `routes` list must configure the `requestsPerSecond` parameter and one of the following route-matching parameters:
|
|
|
|
- `pathExact`
|
|
- `pathPrefix`
|
|
- `pathRegex`
|
|
|
|
Refer to [Examples](#enable-request-rate-limit-on-a-prefixed-path) for example configurations.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: Map
|
|
|
|
The following table describes the parameters you can specify in the `routes` map:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `pathExact` | Specifies the exact path to match on the request path. When using this field, do not configure `pathPrefix` or `pathRegex` in the same `routes` map. | String | None |
|
|
| `pathPrefix` | Specifies the path prefix to match on the request path. When using this field, do not configure `pathExact` or `pathRegex` in the same `routes` map. | String | None |
|
|
| `pathRegex` | Specifies a regular expression to match on the request path. When using this field, do not configure `pathExact` or `pathPrefix` in the same `routes` map. The syntax is proxy-specific. When [using Envoy](/consul/docs/connect/proxies/envoy), refer to [the documentation for Envoy v1.11.2 or newer](https://github.com/google/re2/wiki/Syntax) or [the documentation for Envoy v1.11.1 or older](https://en.cppreference.com/w/cpp/regex/ecmascript), depending on the version of Envoy you use. | String | None |
|
|
| `requestsPerSecond` | Specifies the average number of requests per second allowed to the service. Overrides the [`requestsPerSecond`](#spec-ratelimits-instancelevel-requestspersecond) parameter specified for the service. | Integer | None |
|
|
| `requestsMaxBurst` | Specifies the maximum number of concurrent requests temporarily allowed to the service. When the limit is reached, Consul blocks additional requests. You must specify a value equal to or greater than the `routes.requestsPerSecond` parameter. Overrides the [`requestsMaxBurst`](#spec-ratelimits-instancelevel-requestsmaxburst) parameter specified for the service. | Integer | None |
|
|
|
|
### `spec.upstreamConfig`
|
|
|
|
Specifies a map that controls default upstream connection settings and custom overrides for individual upstream services. If your network contains federated datacenters, individual upstream configurations apply to all pairs of source and upstream destination services in the network.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Map that contains the following configurations:
|
|
- [`UpstreamConfig.Overrides`](#upstreamconfig-overrides)
|
|
- [`UpstreamConfig.Defaults`](#upstreamconfig-defaults)
|
|
|
|
### `spec.upstreamConfig.overrides[]`
|
|
|
|
Specifies options that override the [default upstream configurations](#spec-upstreamconfig-defaults) for individual upstreams.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: List
|
|
|
|
### `spec.upstreamConfig.overrides[].name`
|
|
|
|
Specifies the name of the upstream service that the configuration applies to. Do not use the `*` wildcard to prevent the configuration from applying to unintended upstreams.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.upstreamConfig.overrides[].namespace` <EnterpriseAlert inline/>
|
|
|
|
Specifies the namespace containing the upstream service that the configuration applies to. Do not use the `*` wildcard to prevent the configuration from applying to unintended upstreams.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.upstreamConfig.overrides[].peer`
|
|
|
|
Specifies the peer name of the upstream service that the configuration applies to. The `*` wildcard is not supported.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.upstreamConfig.overrides[].protocol`
|
|
|
|
Specifies the protocol to use for requests to the upstream listener. We recommend configuring the protocol in the main [`protocol`](#protocol) field of the configuration entry so that you can leverage [L7 features](/consul/docs/connect/manage-traffic). Setting the protocol in an upstream configuration limits L7 management functionality.
|
|
|
|
#### Values
|
|
|
|
- Default: inherits the main [`protocol`](#protocol) configuration
|
|
- Data type: String
|
|
|
|
|
|
### `spec.upstreamConfig.overrides[].connectTimeoutMs`
|
|
|
|
Specifies how long in milliseconds that the service should attempt to establish an upstream connection before timing out.
|
|
|
|
We recommend configuring the upstream timeout in the [`connectTimeout`](/consul/docs/connect/config-entries/service-resolver#connecttimeout) field of the `ServiceResolver` CRD for the upstream destination service. Doing so enables you to leverage [L7 features](/consul/docs/connect/manage-traffic). Configuring the timeout in the `ServiceDefaults` upstream configuration limits L7 management functionality.
|
|
|
|
#### Values
|
|
|
|
- Default: `5000`
|
|
- Data type: Integer
|
|
|
|
### `spec.upstreamConfig.overrides[].meshGateway.mode`
|
|
|
|
Map that contains the default mesh gateway `mode` field for the upstream. Refer to [Connect Proxy Configuration](/consul/docs/connect/gateways/mesh-gateway#connect-proxy-configuration) in the mesh gateway documentation for additional information.
|
|
|
|
#### Values
|
|
|
|
You can specify the following string values for the `mode` field:
|
|
|
|
- `none`: The service does not make outbound connections through a mesh gateway. Instead, the service makes outbound connections directly to the destination services.
|
|
- `local`: The service mesh proxy makes an outbound connection to a gateway running in the same datacenter.
|
|
- `remote`: The service mesh proxy makes an outbound connection to a gateway running in the destination datacenter.
|
|
|
|
### `spec.upstreamConfig.overrides[].balanceInboundConnections`
|
|
|
|
Sets the strategy for allocating outbound connections from the upstream across Envoy proxy threads. The only supported value is `exact_balance`. By default, no connections are balanced. Refer to the [Envoy documentation](https://cloudnative.to/envoy/api-v3/config/listener/v3/listener.proto.html#config-listener-v3-listener-connectionbalanceconfig) for details.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.upstreamConfig.overrides[].limits`
|
|
|
|
Map that specifies a set of limits to apply to when connecting to individual upstream services.
|
|
|
|
#### Values
|
|
|
|
The following table describes limits you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `maxConnections` | Specifies the maximum number of connections a service instance can establish against the upstream. Define this limit for HTTP/1.1 traffic. | Integer | `0` |
|
|
| `maxPendingRequests` | Specifies the maximum number of requests that are queued while waiting for a connection to establish. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
| `maxConcurrentRequests` | Specifies the maximum number of concurrent requests. Define this limit for HTTP/2 traffic. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
|
|
### `spec.upstreamConfig.overrides[].passiveHealthCheck`
|
|
|
|
Map that specifies a set of rules that enable Consul to remove hosts from the upstream cluster that are unreachable or that return errors.
|
|
|
|
#### Values
|
|
|
|
The following table describes passive health check parameters you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `interval` | Specifies the time between checks. | String | `0s` |
|
|
| `maxFailures` | Specifies the number of consecutive failures allowed per check interval. If exceeded, Consul removes the host from the load balancer. | Integer | `0` |
|
|
| `enforcingConsecutive5xx ` | Specifies a percentage that indicates how many times out of 100 that Consul ejects the host when it detects an outlier status. The outlier status is determined by consecutive errors in the 500-599 response range. | Integer | `100` |
|
|
| `maxEjectionPercent` | The maximum % of an upstream cluster that can be ejected due to outlier detection. Defaults to 10% but will eject at least one host regardless of the value. | Integer | `10` |
|
|
| `baseEjectionTime` | The base time that a host is ejected for. The real time is equal to the base time multiplied by the number of times the host has been ejected and is capped by max_ejection_time (Default 300s). Defaults to 30000ms or 30s. | String | `30s` |
|
|
|
|
### `spec.upstreamConfig.defaults`
|
|
|
|
Map of configurations that set default upstream configurations for the service. For information about overriding the default configurations for in for individual upstreams, refer to [`spec.upstreamConfig.overrides`](#spec-upstreamconfig-overrides).
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: List
|
|
|
|
### `spec.upstreamConfig.defaults.protocol`
|
|
|
|
Specifies default protocol for upstream listeners. We recommend configuring the protocol in the main [`Protocol`](#protocol) field of the configuration entry so that you can leverage [L7 features](/consul/docs/connect/manage-traffic). Setting the protocol in an upstream configuration limits L7 management functionality.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.upstreamConfig.default.connectTimeoutMs`
|
|
|
|
Specifies how long in milliseconds that all services should continue attempting to establish an upstream connection before timing out.
|
|
|
|
We recommend configuring the upstream timeout in the [`connectTimeout`](/consul/docs/connect/config-entries/service-resolver#connecttimeout) field of the `ServiceResolver` CRD for upstream destination services. Doing so enables you to leverage [L7 features](/consul/docs/connect/manage-traffic). Configuring the timeout in the `ServiceDefaults` upstream configuration limits L7 management functionality.
|
|
|
|
#### Values
|
|
|
|
- Default: `5000`
|
|
- Data type: Integer
|
|
|
|
### `spec.upstreamConfig.defaults.meshGateway.mode`
|
|
|
|
Specifies the default mesh gateway `mode` field for all upstreams. Refer to [Service Mesh Proxy Configuration](/consul/docs/connect/gateways/mesh-gateway#service-mesh-proxy-configuration) in the mesh gateway documentation for additional information.
|
|
|
|
#### Values
|
|
|
|
You can specify the following string values for the `mode` field:
|
|
|
|
- `none`: The service does not make outbound connections through a mesh gateway. Instead, the service makes outbound connections directly to the destination services.
|
|
- `local`: The service mesh proxy makes an outbound connection to a gateway running in the same datacenter.
|
|
- `remote`: The service mesh proxy makes an outbound connection to a gateway running in the destination datacenter.
|
|
|
|
### `spec.upstreamConfig.defaults.balanceInboundConnections`
|
|
|
|
Sets the strategy for allocating outbound connections from upstreams across Envoy proxy threads. The only supported value is `exact_balance`. By default, no connections are balanced. Refer to the [Envoy documentation](https://cloudnative.to/envoy/api-v3/config/listener/v3/listener.proto.html#config-listener-v3-listener-connectionbalanceconfig) for details.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.upstreamConfig.defaults.limits`
|
|
|
|
Map that specifies a set of limits to apply to when connecting upstream services.
|
|
|
|
#### Values
|
|
|
|
The following table describes limits you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `maxConnections` | Specifies the maximum number of connections a service instance can establish against the upstream. Define this limit for HTTP/1.1 traffic. | Integer | `0` |
|
|
| `maxPendingRequests` | Specifies the maximum number of requests that are queued while waiting for a connection to establish. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
| `maxConcurrentRequests` | Specifies the maximum number of concurrent requests. Define this limit for HTTP/2 traffic. An L7 protocol must be defined in the [`protocol`](#protocol) field for this limit to take effect. | Integer | `0` |
|
|
|
|
### `spec.upstreamConfig.defaults.passiveHealthCheck`
|
|
Map that specifies a set of rules that enable Consul to remove hosts from the upstream cluster that are unreachable or that return errors.
|
|
|
|
#### Values
|
|
|
|
The following table describes the health check parameters you can configure:
|
|
|
|
| Limit | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `interval` | Specifies the time between checks. | String | `0s` |
|
|
| `maxFailures` | Specifies the number of consecutive failures allowed per check interval. If exceeded, Consul removes the host from the load balancer. | Integer | `0` |
|
|
| `enforcingConsecutive5xx ` | Specifies a percentage that indicates how many times out of 100 that Consul ejects the host when it detects an outlier status. The outlier status is determined by consecutive errors in the 500-599 response range. | Integer | `100` |
|
|
| `MaxEjectionPercent` | Specifies the maximum percentage of an upstream cluster that Consul ejects when the proxy reports an outlier. Consul ejects at least one host when an outlier is detected regardless of the value. | Integer | `10` |
|
|
| `BaseEjectionTime` | Specifies the minimum amount of time that an ejected host must remain outside the cluster before rejoining. The real time is equal to the value of the `BaseEjectionTime` multiplied by the number of times the host has been ejected. | String | `30s` |
|
|
|
|
### `spec.transparentProxy`
|
|
|
|
Map of configurations specific to proxies in transparent mode. Refer to [Transparent Proxy](/consul/docs/connect/transparent-proxy) for additional information.
|
|
|
|
#### Values
|
|
|
|
You can configure the following parameters in the `TransparentProxy` block:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `outboundListenerPort` | Specifies the port that the proxy listens on for outbound traffic. This must be the same port number where outbound application traffic is redirected. | Integer | `15001` |
|
|
| `dialedDirectly` | Enables transparent proxies to dial the proxy instance's IP address directly when set to `true`. Transparent proxies commonly dial upstreams at the `"virtual"` tagged address, which load balances across instances. Dialing individual instances can be helpful for stateful services, such as a database cluster with a leader. | Boolean | `false` |
|
|
|
|
### `spec.mutualTLSMode`
|
|
|
|
Controls whether mutual TLS is required for incoming connections to this service. This setting is
|
|
only supported for services with transparent proxy enabled. We recommend only using `permissive`
|
|
mode if necessary while onboarding services to the service mesh.
|
|
|
|
#### Values
|
|
|
|
You can specify the following string values for the `MutualTLSMode` field:
|
|
|
|
- `""`: When this field is empty, the value is inherited from the `proxy-defaults` config entry.
|
|
- `strict`: The sidecar proxy requires mutual TLS for incoming traffic.
|
|
- `permissive`: The sidecar proxy accepts mutual TLS traffic on the sidecar proxy service port,
|
|
and accepts any traffic on the destination service's port.
|
|
|
|
### `spec.envoyExtensions`
|
|
|
|
List of extensions to modify Envoy proxy configuration. Refer to [Envoy Extensions](/consul/docs/connect/proxies/envoy-extensions) for additional information.
|
|
|
|
#### Values
|
|
|
|
The following table describes how to configure values in the `envoyExtensions` map:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `name` | Specifies the name of the extension. | String | None |
|
|
| `required` | Specify `true` to require the extension to apply successfully. <p>Use this parameter to ensure that extensions required for secure communication are not unintentionally bypassed.</p><p>When Envoy fails to apply a required extension, Consul logs an error and skips all extensions, leaving xDS resources unchanged. </p> | String | None |
|
|
| `arguments` | Specifies the arguments to pass to the extension. Refer to the documentation for the extension for additional information. | Map | None |
|
|
| `consulVersion` | Specifies the Consul [version constraint](https://github.com/hashicorp/go-version) for the extension. Consul validates the version constraint against the runtime version during xDS updates. If a non-matching version is in use, Consul logs and skips the extension. <p>Use this parameter to avoid upgrade issues when a configured extension is not compatible with a new version of Consul.</p> | String | None |
|
|
| `envoyVersion` | Specifies the Envoy [version constraint](https://github.com/hashicorp/go-version) for the extension. Consul validates the version constraint against the version of the running Envoy proxy during xDS updates. If a non-matching version is in use, Consul logs and skips the extension. <p>Use this parameter to avoid upgrade issues when a configured extension is not compatible with a new version of Envoy.</p> | String | None |
|
|
|
|
### `spec.destination`
|
|
|
|
Map of configurations that specify one or more destinations for service traffic routed through terminating gateways. Refer to [Terminating Gateway](/consul/docs/connect/gateways/terminating-gateway) for additional information.
|
|
|
|
To use the `destination` block, proxy services must be in transparent proxy mode. Refer to [Enable transparent proxy mode](/consul/docs/k8s/connect/transparent-proxy/enable-transparent-proxy) for additional information.
|
|
|
|
#### Values
|
|
|
|
You can configure the following parameters in the `destination` block:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `addresses` | Specifies a list of addresses for the destination. You can configure a list of hostnames and IP addresses. Wildcards are not supported. | List | None |
|
|
| `port` | Specifies the port number of the destination. | Integer | `0` |
|
|
|
|
### `spec.maxInboundConnections`
|
|
|
|
Specifies the maximum number of concurrent inbound connections to each service instance.
|
|
|
|
#### Values
|
|
|
|
- Default: `0`
|
|
- Data type: Integer
|
|
|
|
### `spec.localConnectTimeoutMs`
|
|
|
|
Specifies the number of milliseconds allowed for establishing connections to the local application instance before timing out.
|
|
|
|
#### Values
|
|
|
|
- Default: `5000`
|
|
- Data type: Integer
|
|
|
|
### `spec.localRequestTimeoutMs`
|
|
|
|
Specifies the timeout for HTTP requests to the local application instance. Applies to HTTP-based protocols only. If not specified, inherits the Envoy default for route timeouts.
|
|
|
|
#### Values
|
|
|
|
- Default of `15s` is inherited from Envoy
|
|
- Data type: String
|
|
|
|
### `spec.meshGateway.mode`
|
|
Specifies the default mesh gateway `mode` field for the service. Refer to [Service Mesh Proxy Configuration](/consul/docs/connect/gateways/mesh-gateway#service-mesh-proxy-configuration) in the mesh gateway documentation for additional information.
|
|
|
|
#### Values
|
|
|
|
You can specify the following string values for the `mode` field:
|
|
|
|
- `none`: The service does not make outbound connections through a mesh gateway. Instead, the service makes outbound connections directly to the destination services.
|
|
- `local`: The service mesh proxy makes an outbound connection to a gateway running in the same datacenter.
|
|
- `remote`: The service mesh proxy makes an outbound connection to a gateway running in the destination datacenter.
|
|
|
|
### `spec.externalSNI`
|
|
|
|
Specifies the TLS server name indication (SNI) when federating with an external system.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.expose`
|
|
|
|
Specifies default configurations for exposing HTTP paths through Envoy. Exposing paths through Envoy enables services to listen on `localhost` only. Applications that are not Consul service mesh-enabled can still contact an HTTP endpoint. Refer to [Expose Paths Configuration Reference](/consul/docs/connect/proxies/proxy-config-reference#expose-paths-configuration-reference) for additional information and example configurations.
|
|
|
|
#### Values
|
|
|
|
- Default: None
|
|
- Data type: String
|
|
|
|
### `spec.expose.checks`
|
|
|
|
Exposes all HTTP and gRPC checks registered with the agent if set to `true`. Envoy exposes listeners for the checks and only accepts connections originating from localhost or Consul's [`advertise_addr`](/consul/docs/agent/config/config-files#advertise_addr). The ports for the listeners are dynamically allocated from the agent's [`expose_min_port`](/consul/docs/agent/config/config-files#expose_min_port) and [`expose_max_port`](/consul/docs/agent/config/config-files#expose_max_port) configurations.
|
|
|
|
We recommend enabling the `Checks` configuration when a Consul client cannot reach registered services over localhost, such as when Consul agents run in their own pods in Kubernetes.
|
|
|
|
#### Values
|
|
|
|
- Default: `false`
|
|
- Data type: Boolean
|
|
|
|
### `spec.expose.paths[]`
|
|
|
|
Specifies an list of maps that define paths to expose through Envoy when `spec.expose.checks` is set to `true`.
|
|
|
|
#### Values
|
|
|
|
The following table describes the parameters for each map:
|
|
|
|
| Parameter | Description | Data type | Default |
|
|
| --- | --- | --- | --- |
|
|
| `path` | Specifies the HTTP path to expose. You must prepend the path with a forward slash (`/`). | String | None |
|
|
| `localPathPort` | Specifies the port where the local service listens for connections to the path. | Integer | `0` |
|
|
| `listenPort` | Specifies the port where the proxy listens for connections. The port must be available. If the port is unavailable, Envoy does not expose a listener for the path and the proxy registration still succeeds. | Integer | `0` |
|
|
| `protocol` | Specifies the protocol of the listener. You can configure one of the following values: <li>`http`</li><li>`http2`: Use with gRPC traffic</li> | Integer | `http` |
|
|
|
|
</Tab>
|
|
</Tabs>
|
|
|
|
## Example configurations
|
|
|
|
The following examples describe common `service-defaults` configurations.
|
|
|
|
### Set the default protocol
|
|
|
|
In the following example, protocol for the `web` service in the `default` namespace is set to `http`:
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "web"
|
|
Namespace = "default"
|
|
Protocol = "http"
|
|
```
|
|
|
|
```yaml
|
|
apiVersion: consul.hashicorp.com/v1alpha1
|
|
kind: ServiceDefaults
|
|
metadata:
|
|
name: web
|
|
spec:
|
|
protocol: http
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "web",
|
|
"Namespace": "default",
|
|
"Protocol": "http"
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|
|
|
|
You can also set the global default protocol for all proxies in the [`proxy-defaults` configuration entry](/consul/docs/connect/config-entries/proxy-defaults#default-protocol), but the protocol specified for individual service instances in the `service-defaults` configuration entry takes precedence over the globally-configured value set in the `proxy-defaults`.
|
|
|
|
### Upstream configuration
|
|
|
|
<Tabs>
|
|
<Tab heading="Consul CE">
|
|
|
|
The following example sets default connection limits and mesh gateway mode across all upstreams of the `dashboard` service.
|
|
It also overrides the mesh gateway mode used when dialing its `counting` upstream service.
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "dashboard"
|
|
|
|
UpstreamConfig = {
|
|
Defaults = {
|
|
MeshGateway = {
|
|
Mode = "local"
|
|
}
|
|
Limits = {
|
|
MaxConnections = 512
|
|
MaxPendingRequests = 512
|
|
MaxConcurrentRequests = 512
|
|
}
|
|
}
|
|
|
|
Overrides = [
|
|
{
|
|
Name = "counting"
|
|
MeshGateway = {
|
|
Mode = "remote"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
```
|
|
|
|
```yaml
|
|
apiVersion: consul.hashicorp.com/v1alpha1
|
|
kind: ServiceDefaults
|
|
metadata:
|
|
name: dashboard
|
|
spec:
|
|
upstreamConfig:
|
|
defaults:
|
|
meshGateway:
|
|
mode: local
|
|
limits:
|
|
maxConnections: 512
|
|
maxPendingRequests: 512
|
|
maxConcurrentRequests: 512
|
|
overrides:
|
|
- name: counting
|
|
meshGateway:
|
|
mode: remote
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "dashboard",
|
|
"UpstreamConfig": {
|
|
"Defaults": {
|
|
"MeshGateway": {
|
|
"Mode": "local"
|
|
},
|
|
"Limits": {
|
|
"MaxConnections": 512,
|
|
"MaxPendingRequests": 512,
|
|
"MaxConcurrentRequests": 512
|
|
}
|
|
},
|
|
"Overrides": [
|
|
{
|
|
"Name": "counting",
|
|
"MeshGateway": {
|
|
"Mode": "remote"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|
|
|
|
</Tab>
|
|
<Tab heading="Consul Enterprise">
|
|
|
|
The following example configures the default connection limits and mesh gateway mode for all of the `counting` service's upstreams. It also overrides the mesh gateway mode used when dialing the `dashboard` service in the `frontend` namespace.
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "dashboard"
|
|
Namespace = "product"
|
|
|
|
UpstreamConfig = {
|
|
Defaults = {
|
|
MeshGateway = {
|
|
Mode = "local"
|
|
}
|
|
Limits = {
|
|
MaxConnections = 512
|
|
MaxPendingRequests = 512
|
|
MaxConcurrentRequests = 512
|
|
}
|
|
}
|
|
|
|
Overrides = [
|
|
{
|
|
Name = "counting"
|
|
Namespace = "backend"
|
|
MeshGateway = {
|
|
Mode = "remote"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
```
|
|
|
|
```yaml
|
|
apiVersion: consul.hashicorp.com/v1alpha1
|
|
kind: ServiceDefaults
|
|
metadata:
|
|
name: dashboard
|
|
namespace: product
|
|
spec:
|
|
upstreamConfig:
|
|
defaults:
|
|
meshGateway:
|
|
mode: local
|
|
limits:
|
|
maxConnections: 512
|
|
maxPendingRequests: 512
|
|
maxConcurrentRequests: 512
|
|
overrides:
|
|
- name: counting
|
|
namespace: backend
|
|
meshGateway:
|
|
mode: remote
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "dashboard",
|
|
"Namespace": "product",
|
|
"UpstreamConfig": {
|
|
"Defaults": {
|
|
"MeshGateway": {
|
|
"Mode": "local"
|
|
},
|
|
"Limits": {
|
|
"MaxConnections": 512,
|
|
"MaxPendingRequests": 512,
|
|
"MaxConcurrentRequests": 512
|
|
}
|
|
},
|
|
"Overrides": [
|
|
{
|
|
"Name": "counting",
|
|
"Namespace": "backend",
|
|
"MeshGateway": {
|
|
"Mode": "remote"
|
|
}
|
|
}
|
|
]
|
|
}
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|
|
|
|
</Tab>
|
|
</Tabs>
|
|
|
|
### Terminating gateway destination
|
|
|
|
The following examples creates a default destination assigned to a terminating gateway. A destination
|
|
represents a location outside the Consul cluster. Services can dial destinations dialed directly when transparent proxy mode is enabled.
|
|
Proxy services must be in transparent proxy mode to configure destinations. Refer to [Enable transparent proxy mode](/consul/docs/k8s/connect/transparent-proxy/enable-transparent-proxy) for additional information.
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "test-destination"
|
|
Protocol = "tcp"
|
|
Destination {
|
|
Addresses = ["test.com","test.org"]
|
|
Port = 443
|
|
}
|
|
```
|
|
|
|
```yaml
|
|
apiVersion: consul.hashicorp.com/v1alpha1
|
|
kind: ServiceDefaults
|
|
metadata:
|
|
name: test-destination
|
|
spec:
|
|
protocol: tcp
|
|
destination:
|
|
addresses:
|
|
- "test.com"
|
|
- "test.org"
|
|
port: 443
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "test-destination",
|
|
"Protocol": "tcp",
|
|
"Destination": {
|
|
"Addresses": ["test.com","test.org"],
|
|
"Port": 443
|
|
}
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|
|
|
|
### Enable request rate limit for a service on all paths <EnterpriseAlert inline />
|
|
|
|
The following example configures the default behavior for a service named `billing`. This configuration limits each instance of the billing service to an average of 1000 requests per second, but allows the service to accept up to 1500 concurrent requests.
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "billing"
|
|
Protocol = "http"
|
|
|
|
RateLimit {
|
|
InstanceLevel {
|
|
RequestsPerSecond = 1000
|
|
RequestsMaxBurst = 1500
|
|
}
|
|
}
|
|
```
|
|
|
|
```yaml
|
|
kind: ServiceDefaults
|
|
name: billing
|
|
protocol: http
|
|
rateLimit:
|
|
instanceLevel:
|
|
requestsPerSecond: 1000
|
|
requestsMaxBurst: 1500
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "billing",
|
|
"Protocol": "http",
|
|
"RateLimit": {
|
|
"InstanceLevel": {
|
|
"RequestsPerSecond": 1000,
|
|
"RequestsMaxBurst": 1500
|
|
}
|
|
}
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|
|
|
|
### Enable request rate limit on a prefixed path <EnterpriseAlert inline />
|
|
|
|
The following example limits the request rate on all paths to `service-foo` to an average of `1000` requests per second but allows up to `1500` temporary concurrent requests. Request paths that begin with `/admin`, however, are limited to `20` requests per second.
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "service-foo"
|
|
Protocol = "http"
|
|
|
|
RateLimit {
|
|
InstanceLevel {
|
|
RequestsPerSecond = 1000
|
|
RequestsMaxBurst = 1500
|
|
|
|
Routes = [
|
|
{
|
|
PathPrefix = "/admin"
|
|
RequestsPerSecond = 20
|
|
}
|
|
]
|
|
}
|
|
}
|
|
```
|
|
|
|
```yaml
|
|
kind: service-defaults
|
|
name: service-foo
|
|
protocol: http
|
|
rateLimit:
|
|
instanceLevel:
|
|
requestsPerSecond: 1000
|
|
requestsMaxBurst: 1500
|
|
routes:
|
|
- pathPrefix: /admin
|
|
requestsPerSecond: 20
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "service-foo",
|
|
"Protocol": "http",
|
|
"RateLimit": {
|
|
"InstanceLevel": {
|
|
"RequestsPerSecond": 1000,
|
|
"RequestsMaxBurst": 1500,
|
|
"Routes": [
|
|
{
|
|
"PathPrefix": "/admin",
|
|
"RequestsPerSecond": 20
|
|
}
|
|
]
|
|
}
|
|
}
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|
|
|
|
### Enable request rate limits on multiple paths <EnterpriseAlert inline />
|
|
|
|
The following example configures the default behavior for a service named `billing`. This configuration limits each instance of the billing service depending on the path it received the request on. The service is limited to an average of 500 requests when the request is made on an HTTP path with the `/api` prefix. When an instance of the billing service receives a request from the `/login` path, it is limited to an average of 100 requests per second and 500 concurrent connections.
|
|
|
|
<CodeTabs tabs={[ "HCL", "Kubernetes YAML", "JSON" ]}>
|
|
|
|
```hcl
|
|
Kind = "service-defaults"
|
|
Name = "billing"
|
|
Protocol = "http"
|
|
|
|
RateLimit {
|
|
InstanceLevel {
|
|
Routes = [
|
|
{
|
|
PathPrefix = "/api"
|
|
RequestsPerSecond = 500
|
|
},
|
|
{
|
|
PathPrefix = "/login"
|
|
RequestsPerSecond = 100
|
|
RequestsMaxBurst = 500
|
|
}
|
|
]
|
|
}
|
|
}
|
|
```
|
|
|
|
```yaml
|
|
kind: service-defaults
|
|
name: billing
|
|
protocol: http
|
|
rateLimit:
|
|
instanceLevel:
|
|
routes:
|
|
- pathPrefix: /api
|
|
requestsPerSecond: 500
|
|
- pathPrefix: /login
|
|
requestsPerSecond: 100
|
|
requestsMaxBurst: 500
|
|
```
|
|
|
|
```json
|
|
{
|
|
"Kind": "service-defaults",
|
|
"Name": "billing",
|
|
"Protocol": "http",
|
|
"RateLimit": {
|
|
"InstanceLevel": {
|
|
"Routes": [
|
|
{
|
|
"PathPrefix": "/api",
|
|
"RequestsPerSecond": 500
|
|
},
|
|
{
|
|
"PathPrefix": "/login",
|
|
"RequestsPerSecond": 100,
|
|
"RequestsMaxBurst": 500
|
|
}
|
|
]
|
|
}
|
|
}
|
|
}
|
|
```
|
|
|
|
</CodeTabs>
|