Registered Service

Services in the registry represent logically a service that can be running in different compute platforms and different locations. The same service could be running on different Kubernetes clusters at the same time, on VMS, etc. A service in the registry represents an aggregated and logical view for all those individual services, and provides high-level features such as aggregated metrics.

Port

Port exposed by a service. Registration RPC will complete the instances field by assigning the physical services FQNs.

Field	Description	Validation Rule
number	uint32 REQUIRED A valid non-negative integer port number.	uint32 = {   lte: 65535   gte: 1 }
name	string Name assigned to the port.	–
instances	List of string OUTPUT_ONLY The list of FQNs of the instances that expose this port	–

RegisteredService

A service in the registry that represents an aggregated and logical view for all those individual services, and provides high-level features such as aggregated metrics.

Field	Description	Validation Rule
fqn	string OUTPUT_ONLY Fully-qualified name of the resource. This field is read-only.	–
displayName	string User friendly name for the resource.	–
etag	string The etag for the resource. This field is automatically computed and must be sent on every update to the resource to prevent concurrent modifications.	–
description	string A description of the resource.	–
shortName	string REQUIRED Short name for the service, used to uniquely identify it within the organization.	string = {   min_len: 1 }
hostname	string The hostname by which this service is accessed. Can correspond to the hostname of an internal service or that of a virtual host on a gateway.	–
ports	List of tetrateio.api.tsb.registry.v2.Port The set of ports on which this service is exposed.	–
subsets	List of string Subset denotes a specific version of a service. By default the 'version' label is used to designate subsets of a workload. Known subsets for the service.	–
serviceType	tetrateio.api.tsb.registry.v2.ServiceType REQUIRED Internal/external/load balancer service.	enum = {   defined_only: true }
instances	List of string OUTPUT_ONLY List of FQNs of the existing instances for this service. This is only available for internal and load balancer services and correspond to physical services in the onboarded clusters.	–
externalAddresses	List of string For kubernetes services of type load balancer, this field contains the list of lb hostnames or IPs assigned to the service.	–
state	tetrateio.api.tsb.registry.v2.State REQUIRED State of the service (registered/observed/controlled)	enum = {   defined_only: true }
source	string REQUIRED Source of the service: Kubernetes, Istio, Consul, etc.	string = {   min_len: 1 }
metrics	List of tetrateio.api.tsb.registry.v2.RegisteredService.MetricConfig OUTPUT_ONLY Services may expose different metrics. For example, a regular service may expose the usual red metrics for incoming requests. Services running in multiple clusters, may provide different aggregation levels, such as aggregation by cluster, by subset, etc. This list provides a complete list of all the aggregation keys that are available for this particular service. For example, a service that has instances in multiple clusters could provide the following metrics: - productpage-global: |productpage|bookinfo|| - productpage-cluster1: |productpage|bookinfo|cluster1| - productpage-cluster2: |productpage|bookinfo|cluster2| - productpage-v1: v1|productpage|bookinfo|| - productpage-v1-cluster1: v1|productpage|bookinfo|cluster1| This is only available for Observed and Controlled services.	–

MetricConfig

Configuration for metric aggregation

Field	Description	Validation Rule
name	string A user friendly name for this metric.	–
description	string A helpful description of what this metric represents.	–
aggregationKey	string An aggregation key that can be queried to get metrics for this service.	–
type	tetrateio.api.tsb.registry.v2.RegisteredService.MetricConfig.MetricType Type of the metric (single_instance/aggregated).	–
instance	string The FQN of the instance related with an instance metric. Will be empty for group metrics.	–

MetricType

MetricType denotes the relation of a metrics with a physical service instance.

Name	Number	Description
SINGLE_INSTANCE	0	A single instance metric config belongs to an specific physical service instance.
AGGREGATED	1	An aggregated metric config represents a set of physical services.

ServiceType

ServiceType denotes the exposition of a service in the mesh.

Name	Number	Description
INTERNAL	0	A regular service that is not directly exposed to the outside world.
LOADBALANCER	1	A load balancer service running only the proxy as the workload.
EXTERNAL	2	An external service (not in the mesh).

State

State denotes how deep is the knowledge of a service by the mesh. Meaning that if a service can be controlled, observed or none of these.

Name	Number	Description
REGISTERED	0	A registered service is a service that is known, but that cannot be observed (we can't get metrics for it) and cannot be controlled.
OBSERVED	1	An observed service is a known service that we can have metrics for. For example, a service running the Skywalking agents.
CONTROLLED	2	A controlled service is a service that is part of the mesh and has a proxy we can configure.