# # Copyright 2018-2024 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: "2" title: Zipkin API description: | Zipkin's v2 API currently includes a POST endpoint that can receive spans. host: localhost:9411 basePath: /api/v2 schemes: - http - https consumes: - application/json paths: /services: get: description: | Returns a list of all service names associated with span endpoints. responses: '200': description: OK schema: type: array items: type: string '400': description: Bad Request Error /spans: get: description: Get all the span names recorded by a particular service parameters: - name: serviceName in: query required: true description: | Ex favstar (required) - Lower-case label of a node in the service graph. The /services endpoint enumerates possible input values. type: string responses: '200': description: OK schema: $ref: "#/definitions/ListOfSpans" '400': description: Bad Request Error post: summary: | Uploads a list of spans encoded per content-type, for example json. consumes: - application/json - application/x-protobuf parameters: - in: body name: spans 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 favstar (required) - Lower-case label of a node in the service graph. The /services endpoint enumerates possible input values. type: string - name: spanName in: query required: false description: | Ex get - 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.tags entres. If just a word, constrains against Span.annotations[].value or Span.tags[].key. Any values are AND against eachother. This means a span in the trace must match all of these. - name: minDuration in: query type: integer format: int64 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 format: int64 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 type: string maxLength: 32 minLength: 16 pattern: "[a-f0-9]{16,32}" description: | Trace identifier, set on all spans within it. Encoded as 16 or 32 lowercase hex characters corresponding to 64 or 128 bits. For example, a 128bit trace ID looks like 4e441824ec2b6a44ffdc9bb9a6453df3 responses: '200': description: OK schema: $ref: "#/definitions/Trace" '404': description: "`traceId` not found" /traceMany: get: description: | Invoking this request retrieves any traces with the specified IDs. Results return in any order, and can be empty. Use /trace/{traceId} to request a single trace ID: <2 trace IDs is a bad request. parameters: - name: traceIds in: query type: string minLength: 33 pattern: "([a-f0-9]{16,32},)+([a-f0-9]{16,32})" required: true description: | Comma delimited list of at least two unique trace identifiers. Encoded as 16 or 32 lowercase hex characters corresponding to 64 or 128 bits. For example, a 128bit trace ID looks like 4e441824ec2b6a44ffdc9bb9a6453df3 responses: '200': description: OK. List of traces that match the input traceIds, empty if none match. schema: $ref: "#/definitions/ListOfTraces" '400': description: Bad request. Less than two traceIds, or the list was redundant or malformed. /dependencies: get: description: | Returns service links derived from spans. 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 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" /autocompleteKeys: get: description: | Returns a subset of keys from Span.tags configured for value autocompletion. This helps sites populate common keys into the annotationQuery parameter of the /traces endpoint. For example, a UI can allow users to select site-specific keys from a drop-down as opposed to typing them in manually. This helps guide users towards the more correct keys and avoids typos or formatting problems. responses: '200': description: Success is a list of site-specific keys, such as environment. schema: type: array items: type: string '400': description: Bad Request Error /autocompleteValues: get: description: | Returns all known values of Span.tags for the given autocomplete key. Refer to the description of /autocompleteKeys for the use case. parameters: - name: key in: query required: true description: Name of the autocomplete key from the /autocompleteKeys endpoint. type: string responses: '200': description: | Success result is empty when there are no values or the key was not configured. schema: type: array items: type: string '400': description: Bad Request Error definitions: Endpoint: type: object title: Endpoint description: The network context of a node in the service graph properties: serviceName: type: string description: | Lower-case label of this node in the service graph, such as "favstar". Leave absent 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 description: | Associates an event that explains latency with a timestamp. Unlike log statements, annotations are often codes. Ex. "ws" for WireSend Zipkin v1 core annotations such as "cs" and "sr" have been replaced with Span.Kind, which interprets timestamp and duration. required: - timestamp - value properties: timestamp: type: integer format: int64 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 "error" 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. Tags: type: object title: Tags 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. additionalProperties: type: string 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.' items: $ref: "#/definitions/Span" ListOfTraces: title: ListOfTraces type: array items: $ref: "#/definitions/Trace" Span: title: Span description: | A span is a single-host view of an operation. A trace is a series of spans (often RPC calls) which nest to form a latency tree. Spans are in the same trace when they share the same trace ID. The parent_id field establishes the position of one span in the tree. The root span is where parent_id is Absent and usually has the longest duration in the trace. However, nested asynchronous work can materialize as child spans whose duration exceed the root span. Spans usually represent remote activity such as RPC calls, or messaging producers and consumers. However, they can also represent in-process activity in any position of the trace. For example, a root span could represent a server receiving an initial client request. A root span could also represent a scheduled job that has no remote context. type: object required: - traceId - id properties: traceId: type: string maxLength: 32 minLength: 16 pattern: "[a-f0-9]{16,32}" description: | Randomly generated, unique identifier for a trace, set on all spans within it. Encoded as 16 or 32 lowercase hex characters 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). Leave absent 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-f0-9]{16}" maxLength: 16 minLength: 16 description: 'The parent span ID or absent if this the root span in a trace.' id: type: string pattern: "[a-f0-9]{16}" maxLength: 16 minLength: 16 description: | Unique 64bit identifier for this operation within the trace. Encoded as 16 lowercase hex characters. For example ffdc9bb9a6453df3 kind: type: string enum: - CLIENT - SERVER - PRODUCER - CONSUMER description: | When present, kind clarifies timestamp, duration and remoteEndpoint. When absent, the span is local or incomplete. Unlike client and server, there is no direct critical path latency relationship between producer and consumer spans. * `CLIENT` * timestamp is the moment a request was sent to the server. (in v1 "cs") * duration is the delay until a response or an error was received. (in v1 "cr"-"cs") * remoteEndpoint is the server. (in v1 "sa") * `SERVER` * timestamp is the moment a client request was received. (in v1 "sr") * duration is the delay until a response was sent or an error. (in v1 "ss"-"sr") * remoteEndpoint is the client. (in v1 "ca") * `PRODUCER` * timestamp is the moment a message was sent to a destination. (in v1 "ms") * duration is the delay sending the message, such as batching. * remoteEndpoint is the broker. * `CONSUMER` * timestamp is the moment a message was received from an origin. (in v1 "mr") * duration is the delay consuming the message, such as from backlog. * remoteEndpoint - Represents the broker. Leave serviceName absent if unknown. 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. shared: type: boolean description: 'True if we are contributing to a span started by another tracer (ex on a different host).' localEndpoint: $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. remoteEndpoint: $ref: "#/definitions/Endpoint" description: | When an RPC (or messaging) span, indicates the other side of the connection. By recording the remote endpoint, your trace will contain network context even if the peer is not tracing. For example, you can record the IP from the `X-Forwarded-For` header or the service name and socket of a remote peer. annotations: type: array uniqueItems: true items: $ref: '#/definitions/Annotation' description: 'Associates events that explain latency with the time they happened.' tags: $ref: '#/definitions/Tags' description: 'Tags give your span context for search, viewing and analysis.' example: id: "352bff9a74ca9ad2" traceId: "5af7183fb1d4cf5f" parentId: "6b221d5bc9e6496c" name: "get /api" timestamp: 1556604172355737 duration: 1431 kind: "SERVER" localEndpoint: serviceName: "backend" ipv4: "192.168.99.1" port: 3306 remoteEndpoint: ipv4: "172.19.0.2" port: 58648 tags: http.method: "GET" http.path: "/api" DependencyLink: title: DependencyLink description: | The count of traced calls between services, or between a service and a broker. The direction of the link is parent to child, and can be one of: * client to server * producer to broker * broker to consumer Note: This is related to span ID count between a sender and receiver, but there is nuance that makes it more difficult than counting unique span IDs. Ex. the parent or child might be uninstrumented: detected via the remote endpoint. There can also be scenarios where both sides are instrumented. Please use existing tools such as zipkin-dependencies to derive links as they avoid under or over counting. type: object required: - parent - child - callCount properties: parent: type: string description: 'The service name of the caller: client or message producer or broker.' child: type: string description: 'The service name of the callee: server or message consumer or broker.' callCount: type: integer description: 'Total traced calls made from the parent to the child.' errorCount: type: integer description: 'Total traced calls made from the parent to the child known to be in error.'