g-greatdevaks / apigeecli

This is a tool to interact with Apigee APIs. The tool lets you manage (create, del, get, list) environments, proxies, etc.

About apigeecli

This is a tool to interact with Apigee APIs for Apigee hybrid and Apigee's managed offering. The tool lets you manage (Create,Get, List, Update, Delete, Export and Import) Apigee entities like proxies, products etc. The tools also helps you create Service Accounts in Google IAM to operate Apigee hybrid runtime.

Installation

apigeecli is a binary and you can download the appropriate one for your platform from here. Run this script to download & install the latest version (on Linux or Darwin)

curl -L https://raw.githubusercontent.com/apigee/apigeecli/main/downloadLatest.sh | sh -

NOTE: The signature is not verified and the original zip is not preserved.

To test the signature of the binary, import the gpg public key:

gpg --recv-keys --keyserver keyserver.ubuntu.com A714872F32F34390
gpg: key A714872F32F34390: public key "apigeecli (apigeecli) <[e-mail]>" imported
gpg: Total number processed: 1
gpg:               imported: 1

Use curl or wget to download the zip and sig files. Verify using:

gpg --verify apigeecli_<signature-file>.sig apigeecli_<original-file>.zip
gpg: Signature made Thu 05 May 2022 05:58:11 PM UTC
gpg:                using RSA key 72D11E3A3B1E9FE22110EC45A714872F32F34390
gpg:                issuer "[e-mail]"
gpg: Good signature from "apigeecli (apigeecli) <[e-mail]>" [unknown]
gpg: WARNING: This key is not certified with a trusted signature!
gpg:          There is no indication that the signature belongs to the owner.
Primary key fingerprint: 72D1 1E3A 3B1E 9FE2 2110  EC45 A714 872F 32F3 4390

Getting Started

User Tokens

The simplest way to get started with apigeecli is

token=$(gcloud auth print-access-token)

apigeecli orgs list -t $token

Set Preferences

If you are using the same GCP project for Apigee, then consider setting up preferences so they don't have to be included in every command. Preferences are written to the $HOME/.apigeecli folder

project=$(gcloud config get-value project | head -n 1)

apigeecli prefs set -o $project

Subsequent commands can be like this:

token=$(gcloud auth print-access-token)
apigeecli orgs get -t $token #fetches the org details of the org set in preferences

Access Token Generation

apigeecli can use the service account directly and obtain an access token.

apigeecli token gen -a serviceaccount.json

Parameters The following parameters are supported. See Common Reference for a list of additional parameters.

• --account -a (required) Service Account in json format

Use this access token for all subsequent calls (token expires in 1 hour)

Access Token Caching

apigeecli caches the OAuth Access token for subsequent calls (until the token expires). The access token is stored in $HOME/.apigeecli. This path must be readable/writeable by the apigeecli process.

apigeecli token cache -a serviceaccount.json

or

token=$(gcloud auth print-access-token)
apigeecli token cache -t $token

Container download

The lastest container version for apigeecli can be downloaded via

docker pull ghcr.io/apigee/apigeecli:latest

See this page for other versions.

What you need to know about apigeecli

You must have an account on Apigee to perform any apigeecli functions. These functions include: proxies, API Products, Environments, Org details etc.

You need to be familiar with basic concepts and features of Apigee such as API proxies, organizations, and environments.

For more information, refer to the Apigee API Reference.

Available Commands

Here is a list of available commands

Enviroment Variables

The following environment variables may be set to control the behavior of apigeecli. The default values are all false

• APIGEECLI_DEBUG=true enables debug log
• APIGEECLI_SKIPCACHE=true will not cache the access token on the disk
• APIGEECLI_ENABLE_RATELIMIT=true enables rate limiting when making Apigee APIs (at 1 API call every 100ms)
• APIGEECLI_NO_USAGE=true does not print usage when the command fails
• APIGEECLI_NO_ERRORS=true does not print error messages from the CLI (control plane error messages are displayed)
• APIGEECLI_DRYRUN=true does not execute Apigee control plane APIs

Generating API Proxies

apigeecli can generate API proxies from:

• OpenAPI 3.0 Specification
• GraphQL Schema
• A template/stub for Application Integration
• Cloud Endpoints/API Gateway OpenAPI 2.0 specification

Generating API Proxies from OpenAPI Specs

apigeecli allows the user to generate Apigee API Proxy bundles from an OpenAPI spec (only 3.0.x supported). The Apigee control plane does not support custom formats (ex: uuid). If you spec contains custom formats, consider the following flags

• --add-cors=true: Add a CORS policy
• --formatValidation=false: this disables validation for custom formats.
• --skip-policy=false: By default the OAS policy is added to the proxy (to validate API requests). By setting this to false, schema validation is not enabled and the control plane will not reject the bundle due to custom formats.

The following actions are automatically implemented when the API Proxy bundle is generated:

Security Policies

If the spec defines securitySchemes, for ex the following snippet:

components:
  securitySchemes:
    petstore_auth:
      type: oauth2
      flows:
        implicit:
          authorizationUrl: 'http://petstore.swagger.io/api/oauth/dialog'
          scopes:
            'write:pets': modify pets in your account
            'read:pets': read your pets
    api_key:
      type: apiKey
      name: api_key
      in: header

is interpreted as OAuth-v20 (verification only) policy and the VerifyAPIKey policy.

These security schemes can be added to the PreFlow by enabling the scheme globally

security:
  - api_key: []

Or within a Flow Condition like this

  '/pet/{petId}/uploadImage':
    post:
      ...
      security:
        - petstore_auth:
            - 'write:pets'
            - 'read:pets'

Dynamic target endpoints

apigeecli allows the user to dynamically set a target endpoint. These is especially useful when deploying target/backend applications to GCP's serverless platforms like Cloud Run, Cloud Functions etc. apigeecli also allows the user to enable Apigee'e Google authentication before connecting to the backend.

Set a dynamic target

apigeecli apis create -n petstore -f ./test/petstore.yaml --oas-target-url-ref=propertyset.petstore.url

This example dynamically sets the target.url message context variable. This variable is retrieved from a propertyset file. It is expected the user will separately upload an environment scoped propertyset file with this key.

Set a dynamic target for Cloud Run

apigeecli apis create -n petstore -f ./test/petstore.yaml --oas-google-idtoken-aud-ref=propertyset.petstore.aud --oas-target-url-ref=propertyset.petstore.url

This example dynamically sets the Google Auth audience and the target.url message context variable. These variables are retrieved from a propertyset file. It is expected the user will separately upload an environment scoped propertyset file with these keys. If you do not wish to user a property to set these values later, you can use --oas-google-idtoken-aud-literal to set the audience directly in the API Proxy.

While this example shows the use of Google IDToken, Google Access Token is also supported. To use Google Access Token, use the oas-google-accesstoken-scope-literal flag instead.

Traffic Management

apigeeli allow the user to add SpikeArrest or Quota policies. Since OpenAPI spec does not natively support the ability to specify such policies, a custom extension is used.

Quota custom extension

NOTE: This extension behaves differently when used with Swagger for Cloud Endpoints/API Gateway.

The following configuration allows the user to specify quota parameters in the API Proxy.

x-google-quota:
  - name: test1 # this is appended to the quota policy name, ex: Quota-test1
    interval-literal: 1 # specify the interval in the policy, use interval-ref to specify a variable
    timeunit-literal: minute # specify the timeUnit in the policy, use timeUnit-ref to specify a variable
    allow-literal: 1 # specify the allowed rate in the policy, use allow-ref to specify a variable
    identifier-literal: request.headers.api_key # optionally, set an identifier. If not set, the proxy will count every msg. Identifiers are variables in apigee

NOTE: literals cannot be combined with variables.

The following configuration allows the user to derive quota parameters from an API Product

x-google-quota:
  - name: test1 # this is appended to the quota policy name, ex: Quota-test1
    useQuotaConfigInAPIProduct: Verify-API-Key-api_key # specify the step name that contains the consumer identification. Must be OAuth or VerifyAPIKey step.

The above configurations are mutually exclusive.

SpikeArrest custom extension

The following configuration allows the user to specify Spike Arrest parameters in the API Proxy.

x-google-ratelimit:
  - name: test1 # this is appended to the quota policy name, ex: Spike-Arrest-test1
    rate-literal: 10ps # specify the allowed interval in the policy, use rate-ref to specify a variable
    identifier-ref: request.header.url #optional, specify msg ctx var for the identifier

Examples

See this OAS document for examples

Generating API Proxies from GraphQL Schemas

apigeecli allows the user to generate Apigee API Proxy bundles from a GraphQL schema. When generating a proxy, consider the following flags:

• --basepath: Specify a basePath for the GraphQL proxy
• --skip-policy=false: By default the GraphQL policy is added to the proxy (to validate API requests). By setting this to false, schema validation is not enabled.
• --target-url-ref: Specify a target endpoint location variable. For ex: --target-url-ref=propertyset.gql.url implies the GraphQL target location is available in an environment scoped property set called gql and the key is url

Generating an API Proxy template for Application Integration

apigeecli allows the user to generate an Apigee API Proxy bundle template for Application Integration. When generating the proxy, consider the following flags:

• --trigger: Specify the API trigger name of the Integration. This is also used as the basePath. Don't include api_trigger/
• --integration: Specify the Name of the Integration
• --name: Specify the Name of the API Proxy

Generating an API Proxy template from Cloud Endpoints/API Gateway artifacts

Cloud Endpoints and API Gateway use the extensions specified in an OpenAPI (FKA Swagger) document to apply policies. apigeecli can use those extensions to generate a similar Apigee API Proxy bundle. When generating the proxy, consider the following flags:

• --add-cors=true: Add a CORS policy

Limitations

• The protocol property in x-google-backend is ignored. All upstream/backend is treated as http 1.1
• The metrics property in x-google-management is not supported
• The quota unit is ignored in x-google-management is ignored. See below for quota behavior
• The extension x-google-endpoints is ignored. To add CORS, see above
• Mutiple security requirements: If more than one security policy, regardless of the security type, is set on a path, then the first one is enabled. In the following examples,

  /hello:
    get:
      operationId: hello
      security:
        - google_id_token: []
        - api_key: []

or in the following case,

  /hello:
    get:
      operationId: hello
      security:
        - google_id_token: []
          api_key: []

it cannot be determined which policy is applied. It is best to avoid Swagger documents with such configurations.

• If more than one x-google-jwt-locations are specified, then the first one is used. In the following example,

x-google-jwt-locations:
  # Expect header "Authorization": "MyBearerToken <TOKEN>"
  - header: "Authorization"
    value_prefix: "MyBearerToken "
  # expect query parameter "jwt_query_bar=<TOKEN>"
  - query: "jwt_query_bar"

query parameters are ignored. By default, if no location is specified, the JWT location is the Authorization header and value_prefix is Bearer <token>

---

Support

This is not an officially supported Google product