zipkin-api.yaml

#
# Copyright 2018-2019 The OpenZipkin Authors
#
# Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except
# in compliance with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software distributed under the License
# is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express
# or implied. See the License for the specific language governing permissions and limitations under
# the License.
#

swagger: "2.0"
info:
  version: "1.0.0"
  title: Zipkin API
  description: |
    Zipkin's Query api is rooted at `api/v1`, on a host that by default listens
    on port 9411. It primarily serves the zipkin-ui, although it includes a POST
    endpoint that can receive spans.
host: localhost:9411
basePath: /api/v1
schemes:
  - http
  - https
consumes:
  - application/json
produces:
  - application/json
paths:
  /services:
    get:
      description: |
        Returns a list of all service names associated with annotations.
      responses:
        '200':
          description: Succes
          schema:
            type: array
            items:
              type: string
        '400':
          description: Bad Request Error
  /spans:
    get:
      description: Get all the span names logged by a particular service
      parameters:
      - name: serviceName
        in: query
        required: true
        description: |
          Ex zipkin-server (required) - service that logged an annotation in a
          trace. The /services endpoint enumerates possible input values.
        type: string
      responses:
        '200':
          description: OK
          schema:
            type: array
            items:
              type: string
        '400':
          description: Bad Request Error
    post:
      description: |
          Uploads a list of spans encoded per content-type, for example json
          or thrift (TBinaryProtocol big-endian).
      consumes:
        - application/json
        - application/x-thrift
      produces: []
      parameters:
        - name: span
          in: body
          description: A list of spans that belong to any trace.
          required: true
          schema:
            $ref: "#/definitions/ListOfSpans"
      responses:
        '202':
          description: Accepted
  /traces:
    get:
      description: |
          Invoking this request retrieves traces matching the below filters.

          Results should be filtered against endTs, subject to limit and
          lookback. For example, if endTs is 10:20 today, limit is 10, and
          lookback is 7 days, traces returned should be those nearest to 10:20
          today, not 10:20 a week ago.

          Time units of endTs and lookback are milliseconds as opposed to
          microseconds, the grain of Span.timestamp. Milliseconds is a more
          familiar and supported granularity for query, index and windowing
          functions
      parameters:
        - name: serviceName
          in: query
          required: false
          description: |
            Ex zipkin-server - service that logged an annotation in a trace.
            Required when constraining on parameters except time and duration.
            The /services endpoint enumerates possible input values.
          type: string
        - name: spanName
          in: query
          required: false
          description: |
            Ex my_span_name - name of a span in a trace.
            Only return traces that contains spans with this name.
          type: string
        - name: annotationQuery
          in: query
          type: string
          required: false
          description: |
             Ex. `http.uri=/foo and retried` - If key/value (has an `=`),
             constrains against Span.binaryAnnotations of time string. If just
             a word, constrains against Span.annotations. Any values are AND
             against eachother. This means a span in the trace must match all of
             these.
        - name: minDuration
          in: query
          type: integer
          description: |
             Ex. 100000 (for 100ms). Only return traces whose `Span.duration` is
             greater than or equal to minDuration microseconds.
        - name: maxDuration
          in: query
          type: integer
          description: |
            Only return traces whose Span.duration is less than or equal to
            `maxDuration` microseconds. Only valid with minDuration.
        - name: endTs
          in: query
          type: integer
          format: int64
          description: |
            Only return traces where all Span.timestamp are at or before this
            time in epoch milliseconds. Defaults to current time.
        - name: lookback
          type: integer
          format: int64
          in: query
          description: |
            Only return traces where all Span.timestamp are at or after
            (`endTs - * lookback`) in milliseconds. Defaults to `endTs`, limited
            to a system parameter `QUERY_LOOKBACK`.
        - name: limit
          in: query
          default: 10
          type: integer
          description: |
            Maximum number of traces to return. Defaults to 10
      responses:
        '200':
          description: OK
          schema:
            $ref: "#/definitions/ListOfTraces"
  /trace/{traceId}:
    get:
      parameters:
        - name: traceId
          in: path
          required: true
          description: the 64 or 128-bit big endian, hex-encoded id of the trace as a path parameter.
          type: string
        - name: raw
          in: query
          required: false
          description: |
            Note this flag has no value. Ex. /trace/{traceId}?raw

            Normally, the trace endpoint cleans trace data. For example, it merges
            spans by id, adds missing timestamp or duration, corrects clock skew..

            Specifying this flag is a debug case, when you are debugging zipkin
            logic or zipkin instrumentation, and want to see the input to these
            adjusters. For example, this might explain or rule out clock skew.
          type: boolean
      responses:
        '200':
          description: OK
          schema:
            $ref: "#/definitions/Trace"
        '404':
          description: "`traceId` not found"
  /dependencies:
    get:
      description: |
        Returns dependency links derived from spans.

        Span names are in lowercase, rpc method for example. Conventionally,
        when the span name isn't known, name = "unknown".
      parameters:
        - name: endTs
          in: query
          description: |
            Only return links from spans where `Span.timestamp` are at or before
            this time in epoch milliseconds.
          required: true
          type: integer
          format: int64
        - name: lookback
          in: query
          description: |
            Only return links from spans where all Span.timestamp are at or after
            (`endTs - * lookback`) in milliseconds. Defaults to `endTs`, limited
            to a system parameter `QUERY_LOOKBACK`.
          type: integer
          format: int64
      responses:
        '200':
          description: OK
          schema:
            type: array
            title: ListOfDependencyLinks
            items:
              $ref: "#/definitions/DependencyLink"
definitions:
  Endpoint:
    type: object
    title: Endpoint
    description: The network context of a node in the service graph
    required:
      - serviceName
    properties:
      serviceName:
        type: string
        description: |
                    Lower-case label of this node in the service graph, such as "favstar". Set
                    to empty string if unknown.
                    
                    This is a primary label for trace lookup and aggregation, so it should be
                    intuitive and consistent. Many use a name from service discovery.
      ipv4:
        type: string
        format: ipv4
        description: |
                    The text representation of the primary IPv4 address associated with this
                    connection. Ex. 192.168.99.100 Absent if unknown.
      ipv6:
        type: string
        format: ipv6
        description: |
                    The text representation of the primary IPv6 address associated with a
                    connection. Ex. 2001:db8::c001 Absent if unknown.
                    
                    Prefer using the ipv4 field for mapped addresses.
      port:
        type: integer
        description: |
                    Depending on context, this could be a listen port or the client-side of a
                    socket. Absent if unknown. Please don't set to zero.
  Annotation:
    title: Annotation
    type: object
    required:
      - timestamp
      - value
    description: |
                Associates an event that explains latency with a timestamp.
                Unlike log statements, annotations are often codes. Ex. "sr" for ServerReceive
    properties:     
      timestamp:
        type: integer
        description: |
                    Epoch **microseconds** of this event.
                    
                    For example, 1502787600000000 corresponds to 2017-08-15 09:00 UTC
                    
                    This value should be set directly by instrumentation, using the most precise
                    value possible. For example, gettimeofday or multiplying epoch millis by 1000.
      value:
        type: string
        description: |
                  Usually a short tag indicating an event, like "sr"
                  
                  While possible to add larger data, such as garbage collection details, low
                  cardinality event names both keep the size of spans down and also are easy
                  to search against.
      endpoint:
        $ref: "#/definitions/Endpoint"
        description: |
                    The host that recorded this span, primarily for query by service name.
                    
                    Instrumentation should always record this. Usually, absent implies late
                    data. The IP address corresponding to this is usually the site local or
                    advertised service address. When present, the port indicates the listen
                    port.
  BinaryAnnotation:
    title: BinaryAnnotation
    type: object
    required:
      - key
      - value
    description: |
                Adds context to a span, for search, viewing and analysis.
                
                For example, a key "your_app.version" would let you lookup traces by version.
                A tag "sql.query" isn't searchable, but it can help in debugging when viewing
                a trace.
    properties:
      key:
        type: string
      value:
        type: string
      endpoint:
        $ref: "#/definitions/Endpoint"
        description: |
                    The host that recorded this span, primarily for query by service name.
                    
                    There is an exception, when the key is "sa", "ca" or "ma" this is an
                    address annotation. In such case, the endpoint is not what recorded the
                    span, rather the remote address. The value field is set to boolean true
                    in this case. This feature was refactored in v2 format as "remoteEndpoint"
  ListOfSpans:
    title: ListOfSpans
    description: 'A list of spans with possibly different trace ids, in no particular order'
    type: array
    items:
      $ref: "#/definitions/Span"
  Trace:
    title: Trace
    type: array
    description: 'List of spans who have the same trace id. Spans in the trace, and annotations in a span are sorted ascending by timestamp. ie first event should be first in the spans list.'
    items:
      $ref: "#/definitions/Span"
  ListOfTraces:
    title: ListOfTraces
    type: array
    items:
      $ref: "#/definitions/Trace"
  Span:
    title: Span
    type: object
    required:
      - traceId
      - id
      - name
    properties:
      traceId:
        type: string
        maxLength: 32
        minLength: 16
        pattern: "[a-z0-9]{16,32}"
        description: |
                    Randomly generated, unique identifier for a trace, set on all spans within it.
                    
                    Encoded as 16 or 32 lowercase hex characters in big endian byte order,
                    corresponding to 64 or 128 bits,
                    For example, a 128bit trace ID looks like 4e441824ec2b6a44ffdc9bb9a6453df3
      name:
        type: string
        description: |
                    The logical operation this span represents in lowercase (e.g. rpc method).
                    Set to empty string if unknown.
                    
                    As these are lookup labels, take care to ensure names are low cardinality.
                    For example, do not embed variables into the name.
      parentId:
        type: string
        pattern: "[a-z0-9]{16}"
        maxLength: 16
        minLength: 16
        description: 'The parent span ID or absent if this the root span in a trace.'
      id:
        type: string
        maxLength: 16
        minLength: 16
        pattern: "[a-z0-9]{16}"
        description: |
                    Unique 64bit identifier for this operation within the trace.
                    
                    Encoded as 16 lowercase hex characters. For example ffdc9bb9a6453df3
      timestamp:
        type: integer
        format: int64
        description: |
                    Epoch microseconds of the start of this span, possibly absent if
                    incomplete.
                    
                    For example, 1502787600000000 corresponds to 2017-08-15 09:00 UTC
                    
                    This value should be set directly by instrumentation, using the most
                    precise value possible. For example, gettimeofday or multiplying epoch
                    millis by 1000.
                    
                    There are three known edge-cases where this could be reported absent.
                     * A span was allocated but never started (ex not yet received a timestamp)
                     * The span's start event was lost
                     * Data about a completed span (ex tags) were sent after the fact
      duration:
        type: integer
        format: int64
        minimum: 1
        description: |
                    Duration in **microseconds** of the critical path, if known. Durations of less
                    than one are rounded up. Duration of children can be longer than their
                    parents due to asynchronous operations.
                    
                    For example 150 milliseconds is 150000 microseconds.
      debug:
        type: boolean
        description: |
                    True is a request to store this span even if it overrides sampling policy.
                    
                    This is true when the `X-B3-Flags` header has a value of 1.
      annotations:
        title: ListOfAnnotations
        type: array
        description: 'Associates events that explain latency with the time they happened.'
        items:
          $ref: '#/definitions/Annotation'
      binaryAnnotations:
        title: ListOfBinaryAnnotations
        type: array
        description: 'Binary Annotations are tags that give your span context for search, viewing and analysis.'
        items:
          $ref: '#/definitions/BinaryAnnotation'
  DependencyLink:
    title: DependencyLink
    type: object
    required:
      - parent
      - child
      - callCount
    properties:
      parent:
        type: string
      child:
        type: string
      callCount:
        type: integer
      errorCount:
        type: integer