Connecting to a Redis instance from Cloud Functions

You can connect to a Redis instance from Cloud Functions by using Serverless VPC Access.

Setup

If you have already installed the Cloud SDK and have created a Redis instance, you can skip these steps.

1. Install the Cloud SDK and initialize:

   gcloud init
   

2. Follow the Quickstart Guide to create a Redis instance. Take note of the zone, IP address, and port of the Redis instance.

Configuring Serverless VPC Access

To connect from your Cloud Functions to your Redis instance's authorized VPC network, you must set up Serverless VPC Access.

1. Find your Redis instance's authorized network by running the command:

   gcloud beta redis instances describe [INSTANCE_ID] --region [REGION]
   

2. Follow the instructions at Creating a connector to create a Serverless VPC Access connector. Make sure you create the connector in the same region as your function, and make sure the connector is attached to the Redis instance's authorized VPC network. Remember the name of the connector.

Sample function

This sample function establishes a connection to a Redis instance from Cloud Functions.

Clone the repository for your desired programming language and navigate to the folder that contains the sample code:

Go

git clone https://github.com/GoogleCloudPlatform/golang-samples
cd golang-samples/functions/memorystore/redis

Node.js

git clone https://github.com/GoogleCloudPlatform/nodejs-docs-samples
cd nodejs-docs-samples/functions/memorystore/redis

Python

git clone https://github.com/GoogleCloudPlatform/python-docs-samples
cd python-docs-samples/functions/memorystore/redis

The sample code increments a Redis counter every time the function is triggered.

Go

This function uses the github.com/gomodule/redigo/redis client.

functions/memorystore/redis/visitcount.go

View on GitHub

// Package visitcount provides a Cloud Function that connects
// to a managed Redis instance.
package visitcount

import (
	"errors"
	"fmt"
	"log"
	"net/http"
	"os"

	"github.com/gomodule/redigo/redis"
)

var redisPool *redis.Pool

// initializeRedis initializes and returns a connection pool
func initializeRedis() (*redis.Pool, error) {
	redisHost := os.Getenv("REDISHOST")
	if redisHost == "" {
		return nil, errors.New("REDISHOST must be set")
	}
	redisPort := os.Getenv("REDISPORT")
	if redisPort == "" {
		return nil, errors.New("REDISPORT must be set")
	}
	redisAddr := fmt.Sprintf("%s:%s", redisHost, redisPort)

	const maxConnections = 10
	return &redis.Pool{
		MaxIdle: maxConnections,
		Dial: func() (redis.Conn, error) {
			c, err := redis.Dial("tcp", redisAddr)
			if err != nil {
				return nil, fmt.Errorf("redis.Dial: %v", err)
			}
			return c, err
		},
	}, nil
}

// VisitCount increments the visit count on the Redis instance
// and prints the current count in the HTTP response.
func VisitCount(w http.ResponseWriter, r *http.Request) {
	// Initialize connection pool on first invocation
	if redisPool == nil {
		// Pre-declare err to avoid shadowing redisPool
		var err error
		redisPool, err = initializeRedis()
		if err != nil {
			log.Printf("initializeRedis: %v", err)
			http.Error(w, "Error initializing connection pool", http.StatusInternalServerError)
			return
		}
	}

	conn := redisPool.Get()
	defer conn.Close()

	counter, err := redis.Int(conn.Do("INCR", "visits"))
	if err != nil {
		log.Printf("redis.Int: %v", err)
		http.Error(w, "Error incrementing visit count", http.StatusInternalServerError)
		return
	}
	fmt.Fprintf(w, "Visit count: %d", counter)
}

Node.js

This function uses the redis module.

functions/memorystore/redis/index.js

View on GitHub

const {promisify} = require('util');
const redis = require('redis');

const REDISHOST = process.env.REDISHOST || 'localhost';
const REDISPORT = process.env.REDISPORT || 6379;

const redisClient = redis.createClient(REDISPORT, REDISHOST);
redisClient.on('error', err => console.error('ERR:REDIS:', err));

const incrAsync = promisify(redisClient.incr).bind(redisClient);

exports.visitCount = async (req, res) => {
  try {
    const response = await incrAsync('visits');
    res.writeHead(200, {'Content-Type': 'text/plain'});
    res.end(`Visit count: ${response}`);
  } catch (err) {
    console.log(err);
    res.status(500).send(err.message);
  }
};

Python

This function uses the redis-py package.

functions/memorystore/redis/main.py

View on GitHub

import os

import redis

redis_host = os.environ.get('REDISHOST', 'localhost')
redis_port = int(os.environ.get('REDISPORT', 6379))
redis_client = redis.StrictRedis(host=redis_host, port=redis_port)


def visit_count(request):
    value = redis_client.incr('visits', 1)
    return 'Visit count: {}'.format(value)

Deploying the sample to Cloud Functions

Deploy the function using the gcloud command-line tool:

Go

gcloud functions deploy VisitCount \
--runtime go111 \
--trigger-http \
--region [REGION] \
--vpc-connector projects/[PROJECT_ID]/locations/[REGION]/connectors/[CONNECTOR_NAME] \
--set-env-vars REDISHOST=[REDIS_IP],REDISPORT=[REDIS_PORT]

Node.js

gcloud functions deploy visitCount \
--runtime nodejs10 \
--trigger-http \
--region [REGION] \
--vpc-connector projects/[PROJECT_ID]/locations/[REGION]/connectors/[CONNECTOR_NAME] \
--set-env-vars REDISHOST=[REDIS_IP],REDISPORT=[REDIS_PORT]

Python

gcloud functions deploy visit_count \
--runtime python37 \
--trigger-http \
--region [REGION] \
--vpc-connector projects/[PROJECT_ID]/locations/[REGION]/connectors/[CONNECTOR_NAME] \
--set-env-vars REDISHOST=[REDIS_IP],REDISPORT=[REDIS_PORT]

where:

• [PROJECT_ID] is your Google Cloud project's ID.
• [REGION] is the region where you want to deploy the application.
• [CONNECTOR_NAME] is the name of your connector.
• [REDIS_IP] and [REDIS_PORT] are the IP address and port number of your Redis instance.

After the function deployment finishes, see the visit count increase by sending a GET request to your function's URL:

Go

https://[REGION]-[PROJECT_ID].cloudfunctions.net/VisitCount

Node.js

https://[REGION]-[PROJECT_ID].cloudfunctions.net/visitCount

Python

https://[REGION]-[PROJECT_ID].cloudfunctions.net/visit_count