(Google) Chat me what you (tell Google Cloud) Build

(Google) Chat me what you (tell Google Cloud) Build

If your team is using Google Cloud Build for CI (Continuous Integration) process and is a client of Google Workspace (a.k.a G-Suite). We can setup a bot to send message when the build is finished.

This tutorial requires basic of Python.

Overview

1. Google Cloud Build

First of all is, we must have Google Cloud Build triggers (doc). If yes, go next.

2. Google Cloud Pub/Sub

There is a topic of Cloud Pub/Sub named “cloud-build” automatically when we have triggers. The message will be in this format.

{
  "name": "projects/123/locations/global/builds/ca523374-cf65-4acc-b13f-0c32b7c5f0b0",
  "id": "ca523374-cf65-4acc-b13f-0c32b7c5f0b0",
  "projectId": "project",
  "status": "SUCCESS",
  "source": {
    ...
  },
  "steps": [
    {
      ...
    }
  ],
  "results": {
      ...
  },
  "createTime": "2021-02-19T06:27:49.268719440Z",
  "startTime": "2021-02-19T06:27:50.900538766Z",
  "finishTime": "2021-02-19T06:29:32.266544Z",
  "timeout": "1200s",
  "queueTtl": "3600s",
  "logsBucket": "gs://log-bucket",
  "sourceProvenance": {
    ...
  },
  "buildTriggerId": "e6c46525-39be-4886-afe6-d98c18d6893e",
  "options": {
    ...
  },
  "logUrl": "https://console.cloud.google.com/cloud-build/builds/ca523374-cf65-4acc-b13f-0c32b7c5f0b0?project=123",
  "substitutions": {
    "BRANCH_NAME": "branch",
    "COMMIT_SHA": "b7738ee8b38e1a8551ec632e8e79733c9a546fa1",
    "REPO_NAME": "repo",
    "REVISION_ID": "b7738ee8b38e1a8551ec632e8e79733c9a546fa1",
    "SHORT_SHA": "b7738ee"
  },
  "tags": ["trigger-e6c46525-39be-4886-afe6-d98c18d6893e"],
  "timing": {
    "BUILD": {
      "startTime": "2021-02-19T06:28:03.203464932Z",
      "endTime": "2021-02-19T06:29:27.751418550Z"
    },
    "FETCHSOURCE": {
      "startTime": "2021-02-19T06:27:53.686032686Z",
      "endTime": "2021-02-19T06:28:03.203377780Z"
    }
  }
}
Sample payload from Google Cloud Build

3. Google Chat

Next is to create a room of Google Chat and an incoming webhook by this link.

4. Google Cloud Functions

After all, it’s our turn to create a Google Cloud Functions to connect to the Pub/Sub topic with a condition to send the pushed messages to the Google Chat room through the webhook.

Let’s start.

4.1 Write a function to handle messages from Pub/Sub

With these lines, the Google Cloud Functions will receive messages from a specific Pub/Sub topic.

import base64
import json

def cloudbuild_notifications(event, context):
    message = json.loads(base64.b64decode(event['data']).decode('utf-8'))
Extract message from event

A message comes into variable event. We extract the byte array in path “data” then decode and transform to json format. Now we have payload in variable message.

4.2 Create a function to send a Google Chat message

This doc explains us there are 2 types of Google Chat message that are texts and cards. Select cards to expose information in rich way.

from httplib2 import Http
from dateutil.parser import isoparse
from datetime import datetime, timezone
import pytz
    
def send_ggchat(payload):
    time_pattern = "%Y-%m-%d %H:%M:%S"
    time_timezone = pytz.timezone("Asia/Bangkok")
    time_start = isoparse(payload['timing']['FETCHSOURCE']['startTime']).astimezone(time_timezone)
    time_end = isoparse(payload['timing']['BUILD']['endTime']).astimezone(time_timezone)

    message =  """{{ "cards": [{{
        "header": {{
            "title": "Build Notification for Backend Core",
            "subtitle": "Build {build_id} is {status}"
        }},
        "sections": [{{
            "widgets": [
                {{"keyValue": {{"topLabel": "Repo", "content": "{repo}"}} }},
                {{"keyValue": {{"topLabel": "Branch", "content": "{branch}"}} }},
                {{"keyValue": {{"topLabel": "Commit", "content": "{commit}"}} }},
                {{"keyValue": {{"topLabel": "Created", "content": "{created_time}"}} }},
                {{"keyValue": {{"topLabel": "Status", "content": "{status}"}} }},
                {{"keyValue": {{"topLabel": "Duration (sec)", "content": "{duration}"}} }},
                {{"buttons": [{{
                    "textButton": {{
                        "text": "{build_id} log Link",
                        "onClick": {{"openLink": {{"url": "{log_url}"}} }}
                    }} 
                }} ] }}
            ]
        }} ]
    }} ] }} """.format(
                    build_id=payload['id'].split("-")[0],
                    status=payload['status'],
                    repo=payload['substitutions']['REPO_NAME'],
                    branch=payload['substitutions']['BRANCH_NAME'],
                    commit=payload['substitutions']['SHORT_SHA'],
                    created_time=time_start.strftime(time_pattern),
                    duration=(time_end - time_start).total_seconds(),
                    log_url=payload['logUrl'],
                    )

    message_headers = {'Content-Type': 'application/json; charset=UTF-8'}

    http_obj = Http()
    response = http_obj.request(
        uri="https://chat.googleapis.com/v1/spaces/abc/messages?key=key&token=token",
        method='POST',
        headers=message_headers,
        body=json.dumps(json.loads(message))
    )
Create and send message card to Google Chat

These are my design.

  1. Display id of the build and branch in short format.
  2. Date and time from Pub/Sub message is in nanosecond format such as "2021-02-19T06:28:04.329309285Z". We apply method .isoparse() of library dateutil to transform that string to datetime.
  3. Update timezone from UTC to “Asia/Bangkok” using the library pytz.
  4. Add links to the build log.
  5. Compute build duration since a start time of "FETCHSOURCE" to an end time of "BUILD".

A time to call webhook is to use method Http().request(). The body will be an properly encoded json string via json.dumps(json.loads()).

Here is an example card.

4.3 embed the URL to an environment variable

Hardcoding the URL is not a good idea. We choose to embed it as an environment variable called "_URL" then refer it in code with the variable "GGCHAT_URL".

import os

GGCHAT_URL = os.environ.get('_URL', 'Environment variable does not exist')

def send_ggchat(payload):
    # code ...
    response = http_obj.request(
        uri=GGCHAT_URL,
        method='POST',
        headers=message_headers,
        body=json.dumps(json.loads(message))
    )

4.4 assemble all into main.py

Calling the function send_ggchat() in function cloudbuild_notifications(). This is all of our function in the file “main.py”.

import base64
import json
from httplib2 import Http
from dateutil.parser import isoparse
from datetime import datetime, timezone
import pytz
import os

GGCHAT_URL = os.environ.get('_URL', 'Environment variable does not exist')
    
def send_ggchat(payload):
    time_pattern = "%Y-%m-%d %H:%M:%S"
    time_timezone = pytz.timezone("Asia/Bangkok")
    time_start = isoparse(payload['timing']['FETCHSOURCE']['startTime']).astimezone(time_timezone)
    time_end = isoparse(payload['timing']['BUILD']['endTime']).astimezone(time_timezone)

    message =  """{{ "cards": [{{
        "header": {{
            "title": "Build Notification for Backend Core",
            "subtitle": "Build {build_id} is {status}"
        }},
        "sections": [{{
            "widgets": [
                {{"keyValue": {{"topLabel": "Repo", "content": "{repo}"}} }},
                {{"keyValue": {{"topLabel": "Branch", "content": "{branch}"}} }},
                {{"keyValue": {{"topLabel": "Commit", "content": "{commit}"}} }},
                {{"keyValue": {{"topLabel": "Created", "content": "{created_time}"}} }},
                {{"keyValue": {{"topLabel": "Status", "content": "{status}"}} }},
                {{"keyValue": {{"topLabel": "Duration (sec)", "content": "{duration}"}} }},
                {{"buttons": [{{
                    "textButton": {{
                        "text": "{build_id} log Link",
                        "onClick": {{"openLink": {{"url": "{log_url}"}} }}
                    }} 
                }} ] }}
            ]
        }} ]
    }} ] }} """.format(
                    build_id=payload['id'].split("-")[0],
                    status=payload['status'],
                    repo=payload['substitutions']['REPO_NAME'],
                    branch=payload['substitutions']['BRANCH_NAME'],
                    commit=payload['substitutions']['SHORT_SHA'],
                    created_time=time_start.strftime(time_pattern),
                    duration=(time_end - time_start).total_seconds(),
                    log_url=payload['logUrl'],
                    )

    message_headers = {'Content-Type': 'application/json; charset=UTF-8'}

    http_obj = Http()
    response = http_obj.request(
        uri=GGCHAT_URL,
        method='POST',
        headers=message_headers,
        body=json.dumps(json.loads(message))
    )
    print(response)


def cloudbuild_notifications(event, context):
    message = json.loads(base64.b64decode(event['data']).decode('utf-8'))
    send_ggchat(message)
main.py

4.5 Create requirements.txt

This is because we have imported many external libraries. Need this to tell Google Cloud Functions to prepare the following library on deployment.

python-dateutil
httplib2
pytz
requirements.txt

4.6 Prepare an environment variable file

Because we have environments variables, we also write this file as a reference.

_URL: https://chat.googleapis.com/v1/spaces/abc/messages?key=key&token=token
env.yaml

Name this “env.yaml”.

4.7 Arrange files and folders

prepare files like this.

.
├── env.yaml
└── source
    ├── main.py
    └── requirements.txt

1 directory, 3 files

4.8 deploy function

To deploy via terminal/cmd, make sure Google Cloud SDK is already install in the machine (how to install). Go to the folder “source” and submit this command.

gcloud functions \
--project [project_id] deploy cloudbuild-notifications \
--entry-point cloudbuild_notifications \
--runtime python38 \
--retry \
--env-vars-file ../env.yaml \
--trigger-topic cloud-builds

The “project_id” must be matched with the one of Pub/Sub or the function cannot be triggered.

4.9 Review it

These message are displayed mean the function is deployed successfully.

Deploying function (may take a while - up to 2 minutes)...⠹                                                                                                
For Cloud Build Stackdriver Logs, visit: https://console.cloud.google.com/logs/viewer?project=project&advancedFilter=resource.type%3Dbuild%0Aresource.labels.build_id%3Dd13aeb6e-d08d-417e-9e2d-5bdfc55aa9e5%0AlogName%3Dprojects%2Fproject%2Flogs%2Fcloudbuild
Deploying function (may take a while - up to 2 minutes)...done.                                                                                            
availableMemoryMb: 256
buildId: d13aeb6e-d08d-417e-9e2d-5bdfc55aa9e5
entryPoint: cloudbuild_notifications
environmentVariables:
  _URL: https://chat.googleapis.com/v1/spaces/abc/messages?key=key&token=token
eventTrigger:
  eventType: google.pubsub.topic.publish
  failurePolicy:
    retry: {}
  resource: projects/project/topics/cloud-builds
  service: pubsub.googleapis.com
ingressSettings: ALLOW_ALL
labels:
  deployment-tool: cli-gcloud
name: projects/project/locations/us-central1/functions/cloudbuild-notifications
runtime: python38
serviceAccountEmail: [email protected]
sourceUploadUrl: https://storage.googleapis.com/xxx
status: ACTIVE
timeout: 60s
updateTime: '2021-02-19T10:56:09.074Z'
versionId: '1'
Sample output after we deployed it

Here is the repo of this tutorial.

GitHub - bluebirz/cloudbuild-to-googlechat
Contribute to bluebirz/cloudbuild-to-googlechat development by creating an account on GitHub.



Show Comments