This tutorial is for using the Google Cloud platform with a Particle device, such as a Photon or Electron, to store data into a cloud database.
Updates to this document are here: https://github.com/rickkas7/google_cloud_tutorial
The beginning steps are nearly identical to the steps in the Particle Google Cloud Platform Tutorial, and you should also read that for additional tips.
This is also designed to be very similar to my Google Firebase Tutorial, except it uses Google Pub/Sub instead of Webhooks to access the database. Firebase is easier to get up and running quickly; this method is more powerful and scalable.
There is a lot of documentation for the Google Cloud, as there are many, many parts to it. I'll only be using Pub/Sub, Datastore and App Engine for this tutorial.
Sign up for the Google Cloud Platform Free Trial. When done, you should see your Google Cloud Platform dashboard:
To the right of Google Cloud Platform at the top of the window is a drop down menu. Select Create project.
You can see the test1 project from the Google Firebase tutorial, which didn't use the Google Cloud Platform, but clearly all of the pieces connect together somehow within Google.
This project is "test2". You'll be using the project ID frequently; it's the longer string with the unique number, "test2-146010" in this case.
The next step is to enable the Pub/Sub service. All of the services are in the menu at the top left corner of the window:
Select Pub/Sub (you'll probably have to scroll down) and enable the API:
And create a topic:
The first part of the topic name is fixed, but I added the test2 part at the end.
Note that your topic will be different, even if you're following this exactly, because the number at the beginning will be different. But it should follow the same pattern.
Now set the permissions for the new topic:
Select the checkbox to the left of the topic name (1), then click the Permissions button above it (2).
In the Add members box (1), enter:
particle-public@particle-public.iam.gserviceaccount.com
Then select Pub/Sub Publisher from the drop-down menu (2). You'll need to scroll down to find it.
And finally click Add.
You'll need log into to the Particle console.
Select the integrations icon on the left side (1), click the New Integration button (not pictured) and then Google cloud platform (3) when that appears.
There is a warning about steps you need to complete, but all of these should have been done above so you can confirm.
Then set the integration parameters:
Event Name is the Particle.publish name to match, and I used test2. Remember that event names are a prefix, so this integration will also be called for "test2data" and "test23" so beware!
Google Cloud Pub/Sub Topic is what I set above. Yours will have a different number in it.
projects/test2-146010/topics/test2
I set the the Device to Any to allow any of my devices to trigger the integration.
Then use the Test button to test it.
A green Success box should appear in the lower left corner of the window if the integration is working so far.
You can also test this using the Particle CLI.
particle publish test2 "testing" --private
Note that the Logs part of the Integration panel does not live update, so you'll need to exit and reopen the integration to see updates.
Here there are three log entries, two from my hitting the Test button and one from using the CLI. So far so good!
Back at the Google Cloud Platform Console, select the Activate Google Cloud Shell via the icon at the top of the screen.
Create a subscription:
gcloud beta pubsub subscriptions create test2 --topic test2
The topic "test2" at the end is the topic name, without the "projects/test2-146010/topics/" part, and must match what you set it to above.
Generate an event using the Particle CLI (not the Google console):
particle publish test2 "testing" --private
And then check to see if it arrived in the Google shell:
gcloud beta pubsub subscriptions pull test2 --auto-ack --limit 100
Hopefully, it did!
Later on, we'll add the Google Cloud SDK command lines that allow the same functions from your regular command line instead of a command line web browser window, but for now the web shell is pretty handy. Leave it open because we'll use it once more below.
If you've seen the Firebase example, this is the same as example 2, except for the event name. You'll want that to match what you used above for the Particle event name. I kept the Particle event name the same as the Google Pub/Sub name, but this isn't required.
This example just outputs some random data in JSON format for testing.
#include "Particle.h"
// Test Program #1 for Google Cloud
// Just generates some simple random data once a minute. Also sends up the device name.
#include <math.h> // This is just for cos and M_PI, used in generating random sample data
// Forward declarations
void publishData();
void deviceNameHandler(const char *topic, const char *data);
const unsigned long PUBLISH_PERIOD_MS = 60000;
const unsigned long FIRST_PUBLISH_MS = 10000;
const char *PUBLISH_EVENT_NAME = "test2data";
unsigned long lastPublish = FIRST_PUBLISH_MS - PUBLISH_PERIOD_MS;
int nextValue = 1;
String deviceName;
void setup() {
Serial.begin(9600);
Particle.subscribe("spark/", deviceNameHandler);
Particle.publish("spark/device/name");
}
void loop() {
if (millis() - lastPublish >= PUBLISH_PERIOD_MS) {
lastPublish = millis();
if (deviceName.length() > 0) {
publishData();
}
}
}
void publishData() {
// This just publishes some somewhat random data for testing
// a is a monotonically increasing integer
int a = nextValue++;
// double value b is a cosine, so the values will rise and fall nicely over 360 steps
double b = cos((double)(a % 360) * M_PI / 180.0);
// c is a random integer
int c = rand();
char buf[256];
snprintf(buf, sizeof(buf), "{\"a\":%d,\"b\":%.3f,\"c\":%d,\"n\":\"%s\"}", a, b, c, deviceName.c_str());
Serial.printlnf("publishing %s", buf);
Particle.publish(PUBLISH_EVENT_NAME, buf, PRIVATE);
}
void deviceNameHandler(const char *topic, const char *data) {
deviceName = data;
}
With that running, you can check the Google shell for events:
gcloud beta pubsub subscriptions pull test2 --auto-ack --max-messages 100
Looks good! Actually, that's a little hard to see, but the data contains the JSON data that is expected, like:
{"a":3,"b":0.999,"c":1270216262,"n":"test2"}
You can clean up the subscription that you created using this in the Google shell as you won't need this one anymore.
gcloud beta pubsub subscriptions delete test2
Also, you should see the events in the Particle console, in Logs tab. The events should look a lot like a regular webhook transaction:
Now comes the tricky part. The official Particle tutorial uses a node.js script that you run on your own computer, which is quick and easy.
It looks like Cloud Dataflow might be appropriate for processing large amounts of data. It's pretty complicated, so I think I'll wait on that for a future tutorial.
Cloud Functions are bits of code that get executed when something, like a pub/sub event, occur. These are the easiest way to get started and are the next example.
Another option is App Engine. At first this seemed like overkill, but it's actually not that difficult to use. And, since you get 28 instance hours of app engine per day in the free tier, if you don't use it too much it's still free. And, once you beyond simple samples, there is so much customization you can do, not only for storing the data, but also retrieving and displaying it. That seems ideal for a more complex example, say one that also had web pages for visualizing the data, etc.. A simple example follows after the cloud function example.
The Google Cloud SDK is required for both the cloud functions and app engines methods. This provides local command line tools on your computer so you can deploy the code. There are easy installers for Windows, Mac and Linux.
You'll probably want to install the beta components as well. These commands can also update an older install:
gcloud components update
gcloud components install beta
A good place to learn about cloud functions is the Google Cloud Functions Tutorial.
The app source is in the same place as this tutorial, in the 6-function directory.
https://github.com/rickkas7/google_cloud_tutorial
You will probably want to clone or download the repository to your computer and then use the 6-function directory for this tutorial.
cd 6-function
gcloud beta functions deploy datastoreTest --trigger-resource test2 --trigger-event google.pubsub.topic.publish
Replace the test2 in the --trigger-resource with your topic name, of course. If you have not already enabled cloud functions it will prompt you to enable them and retry.
Assuming you've set up the Particle console Google cloud integration, you can test it by publishing an event:
particle publish test2 '{"a":123}'
To test just the gcloud part:
gcloud beta pubsub topics publish test2 --message '{"a":123}'
To check the logs:
gcloud beta functions logs read --limit 50
And, if you go to the Google console Datastore page, you can see the log entries that were created:
And of course you can use the device firmware from the previous example to populate the Google Cloudstore data.
To delete this cloud function:
gcloud beta functions delete datastoreTest
Like the previous examples, the device publishes some random JSON data every minute, like this:
{"a":2,"b":0.999,"c":1085377743,"n":"test2"}
This is stored in the Google cloud data store, along with a timestamp.
The difference is that after storing the data in the cloud data store, the cloud function uses Particle.publish to publish the latest data to all devices in the account.
While this example just shows publish, the same technique could be used to read from the cloud data store and publish data. Or read from the cloud data store and call a function with data.
A good place to learn about cloud functions is the Google Cloud Functions Tutorial.
The app source is in the same place as this tutorial, in the 8-particle-api-function directory.
https://github.com/rickkas7/google_cloud_tutorial
You will probably want to clone or download the repository to your computer and then use the 8-particle-api-function directory for this tutorial.
You'll need to create a file config.json in the 8-particle-api-function directory containing a Particle access token for your account. You can find one in the settings panel at https://build.particle.io or you can generate one.
config.json:
{
"AUTH_TOKEN":"8c4943f741b8d5bc621b26c723d961b1ac81998d"
}
cd 8-particle-api-function
gcloud beta functions deploy datastoreApiTest --trigger-resource test2 --trigger-event google.pubsub.topic.publish
Replace the test2 in the --trigger-resource with your topic name, of course. If you have not already enabled cloud functions it will prompt you to enable them and retry.
This example requires special firmware to print the recent data from other devices to the serial debugging console. You can flash it like:
cd 8-particle-api-function
particle flash test4 8-photon.cpp
This flashes the test firmware to device "test4".
To check the logs:
gcloud beta functions logs read --limit 10
If you go to the Google console Datastore page, you can see the log entries that were created. In this example you can see that both test2 and test4 published data.
The serial debug logs (particle serial monitor) should look something like this:
0000001382 [hal.wlan] INFO: Bringing WiFi interface up with static IP
0000010000 [app] INFO: publishing {"a":1,"b":1.000,"c":1481765933,"n":"test2"}
0000013123 [app] INFO: recent: {"published_at":"2018-07-25T14:24:46.144Z","a":1,"b":1,"c":1481765933,"n":"test2"}
part0000070000 [app] INFO: publishing {"a":2,"b":0.999,"c":1085377743,"n":"test2"}
0000071381 [app] INFO: recent: {"published_at":"2018-07-25T14:25:46.144Z","a":2,"b":0.999,"c":1085377743,"n":"test2"}
0000087032 [app] INFO: recent: {"published_at":"2018-07-25T14:26:02.570Z","a":1,"b":1,"c":1481765933,"n":"test4"}
To delete this cloud function:
gcloud beta functions delete datastoreApiTest
A good place to learn about cloud functions is the Google Cloud Functions Tutorial.
The app source is in the same place as this tutorial, in the 7-sheets-function directory.
https://github.com/rickkas7/google_cloud_tutorial
You will probably want to clone or download the repository to your computer and then use the 7-sheets-function directory for this tutorial.
Log into the Google cloud console and select the menu (1), then APIs & Services, then Credentials (2), then click the Create Credentials button (3).
Then select Service account key.
Select New Service Account and enter a service account name. I entered spreadsheet-writer but you can select a different name.
Click Select a role then Project then Browser.
Download this key file in JSON and save it to credentials.json in the 7-sheets-function directory.
Also note the service account ID. In my case, it was:
spreadsheet-writer@test2-146010.iam.gserviceaccount.com
You'll need that later.
Copy and paste the API Key into a file config.json in the 7-sheets-function directory:
{
"key":"YOUR_API_KEY"
}
In APIs & Services click on Dashboard (1). Then Enable APIs and Services (2).
Search for and enable Google Sheets API.
Log into Google spreadsheets and create a new spreadsheet.
Click the blue Share button in the upper right. Enter the service account email you noted above. Mine was spreadsheet-writer@test2-146010.iam.gserviceaccount.com.
Also the URL. For example, mine was:
https://docs.google.com/spreadsheets/d/12jG-R4-OqefJfC4kxh4ZSIww8qK1rQkDQBGgzaz9zxY/edit#gid=0
The long part in the middle between the slashes is the spreadsheet ID. Mine was:
12jG-R4-OqefJfC4kxh4ZSIww8qK1rQkDQBGgzaz9zxY
Edit the config.json file you created above and add in the spreadsheetId along with the key:
{
"key":"AIzbSyC8Q3X4QazhjJHxy6ZdE1XrJIBUuFwup2o",
"spreadsheetId":"12jG-R4-OqefJfC4kxh4ZSIww8qK1rQkDQBGgzaz9zxY"
}
Now you have:
- Source code in 7-sheets-function
- config.json in that directory with your key and spreadsheetId
- credentials.json that you downloaded when you created your service key
Now deploy it:
cd 7-sheets-function
gcloud beta functions deploy sheetsTest --trigger-resource test2 --trigger-event google.pubsub.topic.publish
You can generate some test data from the Particle CLI or flash the test firmware to a device.
particle publish test2 '{"a":3,"b":0.999,"c":1270216262,"n":"test2"}'
Wait maybe 10 seconds and check the logs:
gcloud beta functions logs read --limit 10
And you can also check your spreadsheet. With any luck, you'll have a new row in it!
I was using the United States date format, but you can easily transform it by changing this line of code:
row.push(dateFormat(d, 'm/d/yyyy HH:MM:ss'));
If you are committing your code to source control (like github), make sure you don't commit the credentials.json or config.json files, as they have secrets you should not share.
There's a lot of boilerplate code in the index.js file in the source. But the important part is this:
// This generates the columns for the row to add to the spreadsheet
row.push(pubsubMessage.attributes.published_at);
const fields = ['a', 'b', 'c', 'n'];
fields.forEach(function(field) {
if (jsonData.hasOwnProperty(field)) {
row.push(jsonData[field]);
}
else {
row.push('');
}
});
What this does is form the spreadsheet row from the jsonData we got from the Photon.
The first column is the timestamp from the published_at field.
The next columns are the fields a, b, c, and n from the jsonData. You can customize this however you want, and even transform the data from the cloud function before saving it in the spreadsheet.
The Google Sheets with node.js documentation can help you do more advanced things with spreadsheet generation.
To delete this cloud function:
gcloud beta functions delete sheetsTest
In many cases, things will be much easier if you use Google Cloud Functions from the previous examples. There was a period of time, however, when Google Cloud Functions did not exist and the only way to do this was App Engine. The App Engine examples are still included here, because if you want to also provide a web interface, App Engine is still a great way to do both Express and Google Pub/Sub from a single container.
There are a bunch of options when using app engine, but since the official Particle example was already written in node.js and I'm familiar with that, I decided to use a flexible environment node.js configuration for my app engine.
Running through the quickstart will get the prerequisites installed and functioning. If you want to get a feel for how this works, go through the hello world, pub/sub and cloud datastore how-tos, which is how I figured out how to do this. Though note that the pub/sub app engine example uses push mode and I ended up using pull mode, so there are some differences there.
The app source is in the same place as this tutorial, in the 1-app directory.
https://github.com/rickkas7/google_cloud_tutorial
You will probably want to clone or download the repository to your computer and then use the 1-app directory for this tutorial.
We need a new pull subscription in order to implement the app engine versions of the code.
Select Pub/Sub from the main menu (the lines icon in the upper left corner of the window) in the Google Cloud console.
Then check the checkbox for your project and event (1) and hover to the right and select New subscription.
Then fill in the subscription name. The first part is fixed, I just added the test2db part. Select Pull as the delivery type.
The end goal is to run this code remotely on the Google cloud, but it takes a while to deploy a new VM running your code, so it's helpful and a huge time saver to be able to run it locally while you're developing your own code. Fortunately, the system is designed to make this (relatively) easy.
If you're only going to run the example in the cloud, you can skip to the next section, if you prefer.
Go to Google API Manager and select your project (test2, in my case).
Select Create credentials and then Service account key.
In the Credentials window, select App Engine default service account and a JSON key type and click Create. A new file will be downloaded to your computer.
My file was test2-4f42ca1a6f8b.json. Keep this file secure. Don't share it online or even check it into github. It's intended to be specific to your computer. I copied mine to a credentials directory that is at the same level as my app source directory. Doing it this way makes sure I won't accidentally commit the file to git.
If you want to test your code locally you must have Google Cloud SDK and node.js prerequisites installed as mentioned above.
cd 1-app
npm install
This will install all of the dependencies in the node_modules subdirectory.
The following commands will run the server locally:
export GOOGLE_APPLICATION_CREDENTIALS=../credentials/test2-4f42ca1a6f8b.json
export GCLOUD_PROJECT=test2-146010
export PUBSUB_VERIFICATION_TOKEN=PLIRKOQ413FAAS4H2MX5WQ6O
export PUBSUB_TOPIC=test2
export PUBSUB_SUBSCRIPTION_NAME=test2db
npm start
The GOOGLE_APPLICATION_CREDENTIALS line specifies the app credentials file I downloaded above. The name and path will vary depending on where you downloaded and moved it to and its name.
The rest of the parameters match the values in app.yaml, which you should also edit because that's what used when you deploy to the cloud. The fields are explained in greater detail in the next section if you're unsure of what to set them to.
When you run it, you should see something like:
$ npm start
> particle-datastore@0.0.1 start /Users/rickkas7/Documents/src/googlecloud/2simpleserver
> node app.js
App listening on port 8080
Press Ctrl+C to quit.
And, with any luck, if you turn on your photon with the code above, something like this:
event received { ackId: 'RUFeQBJMJxhESVMrQwsqWBFOBCEhPjA-RVNEUAYWLF1GSFE3GQhoUQ5PXiM_NSAoRREICBQFfH1xXVJ1WFUaB1ENGXJ8ZHI5UxEJC0NWKlVSEQ16bVxXOVcOGXp8aXVqUhUGBU1bVpLO7_FaZho9XxJLLD5-LC0',
id: '63081800867788',
data: { a: 24, b: 0.914, c: 1196032868, n: 'test4' },
attributes:
{ device_id: '1e0032123447343149111039',
event: 'test2data',
published_at: '2016-10-15T16:43:22.672Z' },
ack: [Function: bound ],
skip: [Function] }
stored in datastore { published_at: '2016-10-15T16:43:22.672Z',
a: 24,
b: 0.914,
c: 1196032868,
n: 'test4' }
And, if you go to the Google console Datastore page, you can see the log entries that were created:
The storeEvent function in app.js does the important work of getting things out of the pub/sub event and making it into the format you want to store in the database:
function storeEvent(message) {
var key = datastore.key('ParticleEvent');
// You can uncomment some of the other things if you want to store them in the database
var obj = {
// gc_pub_sub_id: message.id,
// device_id: message.attributes.device_id,
// event: message.attributes.event,
published_at: message.attributes.published_at
}
// Copy the data in message.data, the Particle event data, as top-level
// elements in obj. This breaks the data out into separate columns.
for (var prop in message.data) {
if (message.data.hasOwnProperty(prop)) {
obj[prop] = message.data[prop];
}
}
This adds the published_at, and then copies out the data from the Particle event sent by the Photon. That data originally looked like this:
{"a":29,"b":0.875,"c":1883488164,"n":"test4"}
If you add additional data to the event data JSON object from the Particle side, it automatically adds new columns in the database automatically. Neat!
That wasn't too bad, but the goal is to deploy to the cloud. Fortunately, this is almost easier than running locally.
Make sure you've updated app.yaml, in particular this section:
env_variables:
GCLOUD_PROJECT: test2-146010
PUBSUB_TOPIC: test2
PUBSUB_SUBSCRIPTION_NAME: test2db
# This token is used to verify that requests originate from your
# application. It can be any sufficiently random string.
PUBSUB_VERIFICATION_TOKEN: PLIRKOQ413FAAS4H2MX5WQ6O
GCLOUD_PROJECT is your project ID.
PUBSUB_TOPIC and PUBSUB_SUBSCRIPTION_NAME are in the Pub/Sub configuration.
The PUBSUB_VERIFICATION_TOKEN is just a random string. You can set it to any random string.
Once those are set, deploy it!
gcloud app deploy
This will take a while to complete, but when you're done, it should display this:
Updating service [default]...done.
Deployed service [default] to [https://test2-146010.appspot.com]
You can read logs from the command line by running:
$ gcloud app logs read
To view your application in the web browser run:
$ gcloud app browse
The browse function won't work for this app because I didn't implement any web browser pages. But the logs read command still works great:
gcloud app logs read
One thing to beware of: Deploying a new instance doesn't stop the old one first! Make sure you to clean up the old instances and versions from the console App Engine page.
This is an example of how to download your Google cloud datastore data.
The app source is in the same place as this tutorial, in the 4-download directory.
https://github.com/rickkas7/google_cloud_tutorial
You will probably want to clone or download the repository to your computer and then use the 4-download directory for this tutorial.
This example isn't really intended to run in the cloud; it's intended to run locally on your computer. However it uses the same Google authentication as you would for an app that can run either locally or in the cloud.
Also, this example assumes you've set up the Google Cloud Datastore example above and will reuse your service account key.
If you want to run this example you must have Google Cloud SDK and node.js prerequisites installed as mentioned above.
cd 4-download
npm install
This will install all of the dependencies in the node_modules subdirectory.
The following commands will run the server locally:
export GOOGLE_APPLICATION_CREDENTIALS=../credentials/test2-4f42ca1a6f8b.json
export GCLOUD_PROJECT=test2-146010
npm start
The GOOGLE_APPLICATION_CREDENTIALS line specifies the app credentials file I downloaded above. The name and path will vary depending on where you downloaded and moved it to and its name.
The GCLOUD_PROJECT is your project ID.
When you run it, you should see something like:
$ npm start
> particle-datastore@0.0.1 start /Users/rickkas7/Documents/src/googlecloud/4download
> node app.js
7,0.993,445349752,2017-02-10T17:47:19.269Z
6,0.995,553475508,2017-02-10T17:46:19.273Z
5,0.996,812669700,2017-02-10T17:45:19.273Z
4,0.998,1191391529,2017-02-10T17:44:19.272Z
3,0.999,1270216262,2017-02-10T17:43:19.276Z
2,0.999,1085377743,2017-02-10T17:42:19.273Z
1,1,1481765933,2017-02-10T17:41:19.276Z
The code is simple, and looks like this:
var Datastore = require('@google-cloud/datastore');
// Instantiate a datastore client
var datastore = Datastore();
const query = datastore.createQuery('ParticleEvent')
.limit(100)
.order('published_at', {
descending: true
});
datastore.runQuery(query, function(err, entities, info) {
if (err) {
console.log("database query error", err);
return;
}
for(var ii = 0; ii < entities.length; ii++) {
var data = entities[ii].data;
var csv = data.a + ',' + data.b + ',' + data.c + ',' + data.published_at;
console.log(csv);
}
});
It downloads the 100 most recent entries and prints them out as comma-separated values.
There's also an example in 5-excel that works like 4-download, except it saves the data in a file called out.xsls, and Excel-compatible spreadsheet file.
It uses xlsx to create the files.
The example above uses the cloud datastore. This example uses Firebase. It's just another database option from Google. I'm not certain there's a significant benefit to using it over Cloud Datastore, but here's how to use it if you'd like to.
If you've already done the cloud datastore example above can reuse the subscription if you've stopped that server. Otherwise, create a new subscription:
Select Pub/Sub from the main menu (the lines icon in the upper left corner of the window) in the Google Cloud console.
Then check the checkbox for your project and event (1) and hover to the right and select New subscription.
Then fill in the subscription name. The first part is fixed, I just added the test2db part. Select Pull as the delivery type.
Go to the Firebase console. Since we want to use our existing Google Cloud project test2, we'll use the Import Google Project option to add that.
The app source is in the same place as this tutorial, in the 3-firebase directory.
https://github.com/rickkas7/google_cloud_tutorial
You will probably want to clone or download the repository to your computer and then use the 3-firebase directory for this tutorial.
The end goal is to run this code remotely on the Google cloud, but it takes a while to deploy a new VM running your code, so it's helpful and a huge time saver to be able to run it locally while you're developing your own code. Fortunately, the system is designed to make this (relatively) easy.
If you're only going to run the example in the cloud, you can skip to the next section, if you prefer.
If you previously did the Google Cloud Datastore example above, you can reuse that service account key; you don't need to create a new one.
Go to Google API Manager and select your project (test2, in my case).
Select Create credentials and then Service account key.
In the Credentials window, select App Engine default service account and a JSON key type and click Create. A new file will be downloaded to your computer.
My file was test2-4f42ca1a6f8b.json. Keep this file secure. Don't share it online or even check it into github. It's intended to be specific to your computer. I copied mine to a credentials directory that is at the same level as my app source directory. Doing it this way makes sure I won't accidentally commit the file to git.
If you want to test your code locally you must have Google Cloud SDK and node.js prerequisites installed as mentioned above.
cd 3-firebase
npm install
This will install all of the dependencies in the node_modules subdirectory.
The following commands will run the server locally:
export GOOGLE_APPLICATION_CREDENTIALS=../credentials/test2-4f42ca1a6f8b.json
export GCLOUD_PROJECT=test2-146010
export PUBSUB_VERIFICATION_TOKEN=PLIRKOQ413FAAS4H2MX5WQ6O
export PUBSUB_TOPIC=test2
export PUBSUB_SUBSCRIPTION_NAME=test2db
npm start
The GOOGLE_APPLICATION_CREDENTIALS line specifies the app credentials file I downloaded above. The name and path will vary depending on where you downloaded and moved it to and its name.
The rest of the parameters match the values in app.yaml, which you should also edit because that's what used when you deploy to the cloud. The fields are explained in greater detail in the next section if you're unsure of what to set them to.
The PUBSUB_VERIFICATION_TOKEN is just a random string. You can set it to any random string.
When you run it, you should see something like:
$ npm start
> particle-firebase@0.0.1 start /Users/rickkas7/Documents/src/googlecloud/3firebase
> node app.js
App listening on port 8080
Press Ctrl+C to quit.
And, with any luck, if you turn on your photon with the code above, something like this:
event received { ackId: 'QV5AEkwkC0RJUytDCypYEU4EISE-MD5FU0RQBhYsXUZIUTcZCGhRDk9eIz81IChFEAUIFAV8fXFdUnVYVRoHUQ0ZcnxkJGhTE1cERQB7VVsRDXptXFc4UA0fe31ifWheGwMATVV7d9mKyYJvZho9XxJLLD5-Lz5F',
id: '106903824822865',
data: { a: 30, b: 0.866, c: 964776169, n: 'test4' },
attributes:
{ device_id: '1e0032123447343149111039',
event: 'test2data',
published_at: '2017-02-10T17:25:57.698Z' },
ack: [Function: bound ],
skip: [Function] }
And, if you go to the Google Firebase console you should be able to see the newly created data.
The storeEvent function in app.js does the important work of getting things out of the pub/sub event and making it into the format you want to store in the database:
function storeEvent(message) {
var particleEventRef = dbRef.child("ParticleEvent");
// You can uncomment some of the other things if you want to store them in the database
var obj = {
// gc_pub_sub_id: message.id,
// device_id: message.attributes.device_id,
// event: message.attributes.event,
published_at: message.attributes.published_at
}
// Copy the data in message.data, the Particle event data, as top-level
// elements in obj. This breaks the data out into separate columns.
for (var prop in message.data) {
if (message.data.hasOwnProperty(prop)) {
obj[prop] = message.data[prop];
}
}
particleEventRef.push().set(obj);
}
This adds the published_at, and then copies out the data from the Particle event sent by the Photon. That data originally looked like this:
{"a":29,"b":0.875,"c":1883488164,"n":"test4"}
If you add additional data to the event data JSON object from the Particle side, it automatically adds new columns in the database automatically. Neat!
The process is the same as for the Cloud Datastore example, so you can just follow the instructions there.
Here's a slightly modified version of the Photon firmware above that's better suited for the Electron. It saves the device name in retained memory so it doesn't need to be retrieved every time. It also goes to sleep between samplings to save battery.
#include "Particle.h"
// Test Program #4 for Firebase Integration
// Similar to #2, but optimized for the Electron. Goes to sleep between samples.
// Just generates some simple random data once a minute, stores in a per-device table
// using the device name as the key.
STARTUP(System.enableFeature(FEATURE_RETAINED_MEMORY));
#include <math.h> // This is just for cos and M_PI, used in generating random sample data
SYSTEM_THREAD(ENABLED);
// Forward declarations
void publishData();
void deviceNameHandler(const char *topic, const char *data);
const unsigned long POST_CONNECT_WAIT_MS = 8000;
const unsigned long NAME_WAIT_MS = 20000;
const unsigned long POST_PUBLISH_WAIT_MS = 2000;
const unsigned long MAX_CONNECT_TIME_MS = 120000;
const unsigned long SLEEP_TIME_SEC = 3600;
const char *PUBLISH_EVENT_NAME = "test2data";
const uint32_t RETAINED_DATA_MAGIC = 0xa2c7206a;
const size_t DEVICE_NAME_MAX_LEN = 31;
typedef struct {
uint32_t magic;
char deviceName[DEVICE_NAME_MAX_LEN + 1];
} RetainedData;
enum State {
START_STATE,
CONNECTED_WAIT_STATE,
POST_CONNECT_WAIT_STATE,
GET_NAME_STATE,
NAME_WAIT_STATE,
PUBLISH_STATE,
POST_PUBLISH_WAIT_STATE,
SLEEP_STATE
};
retained RetainedData retainedData = {0};
int nextValue = 1;
State state = START_STATE;
unsigned long stateTime = 0;
void setup() {
Serial.begin(9600);
}
void loop() {
switch(state) {
case START_STATE:
state = CONNECTED_WAIT_STATE;
stateTime = millis();
break;
case CONNECTED_WAIT_STATE:
if (Particle.connected()) {
Serial.println("connected to the cloud");
state = POST_CONNECT_WAIT_STATE;
}
else
if (millis() - stateTime >= MAX_CONNECT_TIME_MS) {
// Failed to connect to the cloud, go to sleep and try again later
Serial.println("failed to connect to cloud");
state = SLEEP_STATE;
}
break;
case POST_CONNECT_WAIT_STATE:
if (millis() - stateTime >= POST_CONNECT_WAIT_MS) {
state = GET_NAME_STATE;
}
break;
case GET_NAME_STATE:
if (retainedData.magic != RETAINED_DATA_MAGIC || retainedData.deviceName[0] == 0) {
memset(&retainedData, 0, sizeof(retainedData));
retainedData.magic = RETAINED_DATA_MAGIC;
Particle.subscribe("spark/", deviceNameHandler);
Particle.publish("spark/device/name");
state = NAME_WAIT_STATE;
stateTime = millis();
}
else {
Serial.printlnf("device name in retained memory %s", retainedData.deviceName);
state = PUBLISH_STATE;
stateTime = millis();
}
break;
case NAME_WAIT_STATE:
if (retainedData.deviceName[0] != 0) {
Serial.printlnf("device name from cloud %s", retainedData.deviceName);
state = PUBLISH_STATE;
stateTime = millis();
}
else
if (millis() - stateTime >= NAME_WAIT_MS) {
// Failed to connect to the cloud, go to sleep and try again later
Serial.println("failed to get device name");
state = SLEEP_STATE;
}
break;
case PUBLISH_STATE:
publishData();
state = POST_PUBLISH_WAIT_STATE;
stateTime = millis();
break;
case POST_PUBLISH_WAIT_STATE:
if (millis() - stateTime >= POST_PUBLISH_WAIT_MS) {
state = SLEEP_STATE;
}
break;
case SLEEP_STATE:
Serial.printlnf("going to sleep for %d seconds", SLEEP_TIME_SEC);
// This delay is just so you can see the serial print go out, feel free to remove if desired
delay(250);
System.sleep(SLEEP_MODE_DEEP, SLEEP_TIME_SEC);
// Not reached; when waking up from SLEEP_MODE_DEEP, the code starts again from setup()
break;
}
}
void publishData() {
// This just publishes some somewhat random data for testing
// a is a monotonically increasing integer
int a = nextValue++;
// double value b is a cosine, so the values will rise and fall nicely over 360 steps
double b = cos((double)(a % 360) * M_PI / 180.0);
// c is a random integer
int c = rand();
char buf[256];
snprintf(buf, sizeof(buf), "{\"a\":%d,\"b\":%.3f,\"c\":%d,\"n\":\"%s\"}", a, b, c, retainedData.deviceName);
Serial.printlnf("publishing %s", buf);
Particle.publish(PUBLISH_EVENT_NAME, buf, PRIVATE);
}
void deviceNameHandler(const char *topic, const char *data) {
if (strlen(data) <= DEVICE_NAME_MAX_LEN) {
strcpy(retainedData.deviceName, data);
}
else {
// Truncate name
strncpy(retainedData.deviceName, data, DEVICE_NAME_MAX_LEN);
retainedData.deviceName[DEVICE_NAME_MAX_LEN] = 0;
}
}