@fuzzy-ai/microservice
v1.3.0
Published
Simple microservice class used by Fuzzy.ai
Downloads
9
Readme
fuzzy.ai-microservice
This is the microservice class we use for Fuzzy.ai. The goal is to avoid re-writing a lot of boilerplate needed to set up an HTTP server and a database connection. It has a couple of nice characteristics that make this useful for us.
- It's configured using environment variables.
- It uses databank for data access.
- It uses express for the web interface.
- It uses Bunyan for logging.
We use Docker, so it dumps out its logs to stdout.
License
Copyright 2016 Fuzzy.ai
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.
Usage
You should be able to write pretty small microservice servers. Here's an example.
Microservice = require 'fuzzy.ai-microservice'
# Subclass Microservice
class BasicServer extends Microservice
# Override setupRoutes to add routes to your expressjs server
setupRoutes: (exp) ->
exp.get '/version', (req, res, next) ->
res.json {version: '0.1.0'}
# This one uses the built-in app authentication
exp.get '/who-am-i', @appAuthc, (req, res, next) ->
res.json {appName: req.appName}
server = new BasicServer()
server.start (err) ->
if err
console.error(err)
else
console.log("Server started.")
Calling a microservice
Note that you probably shouldn't invoke Microservice directly; you should use a sub-class. Here are the methods that you should use:
constructor(environment)
. Takes an environment as a parameter. If none is provided, uses process.env. The environment variables are changed into configuration options.start(callback)
. Start the microservice.callback
is called with either anerr
argument ornull
.stop(callback)
. Stop the microservice.callback
is called with either anerr
argument ornull
.
Methods to overload
These are methods that sub-classes of Microservice should overload.
getSchema()
. Return the Databank schema for the microservice. See https://github.com/e14n/databank#schemata for the format.setupMiddleware(exp)
. If you have any custom middleware to set up for the express serverexp
, do it here.setupParams(exp)
. Any custom params would go here. Good place forexp.param()
statements.setupRoutes(exp)
. All your routes should go here.startCustom(callback)
. If you need to have something happen after starting the server, do it here. This is a good time for ensuring databank items, for example.stopCustom(callback)
. If you need to do something before stopping (what?), do it here.environmentToConfig(env)
. Convert the environment to a config object.
These ones might be useful to overload if they're not working correctly.
getName()
. Return the name of the microservice. Used for error reporting and the like. Default implementation guesses from environment variables, so if you usenpm run start
for your microservice, you should be fine.getVersion()
. Ditto, but for the version.
Utility methods
These are some useful methods for microservice sub-classes to use.
envInt(env, key, def)
: Return the environment variable fromenv
atkey
as an integer, ordef
if the variable doesn't exist.envJSON(env, key, def)
: Return the environment variable fromenv
atkey
, parsed as JSON, ordef
if the variable doesn't exist.envBool(env, key, def)
: Return the environment variable fromenv
atkey
, interpreted as a boolean, ordef
if the variable doesn't exist. Case-insensitive variables that match "true", "yes", "on", or "1" are booleantrue
; ones that match "false", "no", "off", or "0" are booleanfalse
. Anything else gives an error.appAuthc(req, res, next)
: Middleware for checking the bearer token of a request against the configured app keys (see below). Will give the correct authorization error if none is allowed. Use this in your routes!slackMessage(type, message, icon, callback)
. Notification method for sending updates to Slack. Errors are sent to Slack by default, but you can send other notifications if you need to. You can send things to different hooks using the 'type' modifier. If there is no specific hook for that type (see SLACK_HOOK_SOMETHING below for how to do that), it will be sent via the default hook.dontLog(req, res, next)
. Middleware to use when you don't want to have a route logged. Useful for e.g. health-check URLs.
Environment variables
The system uses environment variables for configuration. This is great if you use Docker. We use Docker Compose, so that's even more great.
Here are the variables it uses by default.
PORT: The port to listen on. Defaults to 443 if
KEY
is set (see below), otherwise 80.ADDRESS: IP address to listen on. Defaults to '0.0.0.0', meaning all addresses.
HOSTNAME: hostname to use. Use address instead, usually.
KEY: SSL key to use. This is the full key, not the name of a file.
CERT: SSL cert to use. This is the full cert, not the name of a file.
LOG_LEVEL: Bunyan log level. Defaults to 'info'.
APP_KEY_SOMETHING: The app key that app 'something' will use to access this server. Supported by internal appAuthc.
MAX_UPLOAD_SIZE: Maximum size of an upload. Use a string with 'mb', 'gb' or 'kb' to define a size in bytes. Defaults to '50mb'.
SLACK_HOOK_SOMETHING: Hook for sending 'something' messages to Slack. This lets you specialise your Slack messages. By default, the error handler will use the 'error' hook, or it will fall back to the default. Note that hook 'SLACK_HOOK_SOMETHING' will get lowercased to 'something' when you need to send a slack message.
You can have a microservice grab more environment variables by overloading
environmentToConfig
.
environmentToConfig: (environment) ->
cfg = super environment
cfg.foo = environment.FOO
cfg