apidApigeeSync Mock Server

Overview

This Mock Server is used during unit tests of apidApigeeSync and has been designed to be run standalone for stand-alone development use as well as performance and load testing.

Build

From the apidApigeeSync base dir:

glide install

From apidApigeeSync/cmd/mockServer:

go build

You should now have an executable named “mockServer”.

Execute

Execute with the -h flag to see flags:

./mockServer -h
Usage of ./mockServer:
  -addDevEach duration
        add a developer each duration (default 0s)
  -bundleURI string
        a URI to a valid deployment bundle (default '')
  -numDeps int
        number of deployments in snapshot (default 2)
  -numDevs int
        number of developers in snapshot (default 2)
  -reliable
        if false, server will often send 500 errors (default true)
  -upDepEach duration
        update (replace) a deployment each duration (default 0s)
  -upDevEach duration
        update a developer each duration (default 0s)         

Note: Nothing is required.

The following are the values used by default by the Mock Server:

ReliableAPI: true
ClusterID: "cluster"
TokenKey: "key"
TokenSecret: "secret"
Scope: "scope"
Organization: "org"
Environment: "test"
NumDevelopers: 2
AddDeveloperEvery: 0
UpdateDeveloperEvery: 0
NumDeployments: 2
ReplaceDeploymentEvery: 0
Port: 9001

Put it to use

Set your apid configuration to point toward the Mock Server and have correct cluster, key, and secret values.

For example:

api_port: 9000
api_expvar_path: /expvar
events_buffer_size: 5
log_level: debug
apigeesync_proxy_server_base: http://localhost:9001
apigeesync_snapshot_server_base: http://localhost:9001
apigeesync_change_server_base: http://localhost:9001
apigeesync_consumer_key: key
apigeesync_consumer_secret: secret
apigeesync_cluster_id: cluster
#data_trace_log_level: debug
data_source: file:%s?_busy_timeout=20000

Now start apid. It should download the snapshot and changes as you configured for the Mock Server.

Try out a couple of APIs to verify:

curl -i -d "action=verify&key=1&uriPath=/&scopeuuid=scope" :9000/verifiers/apikey

curl -i :9000/deployments

Notes

Under high loads (eg. a large snapshot), apid may get timeout errors from sqlite. If you see this, you can work around it by increasing the _busy_timeout by adding a config item to your apid config:

data_source: file:%s?_busy_timeout=10000

The _busy_timeout value is in milliseconds, so the above value is 10s.