Cloud Run: endpoint that runs a function as background job - python

I am trying to deploy a rest api in cloud run where one endpoint launches an async job. The job is defined inside a function in the code.
It seems one way to do it is to use Cloud Task, but this would mean to make a self-call to another endpoint of the deployed api. Specifically, to create an auxiliary endpoint that contains the job code (e.g. /run-my-function) and another one to set the queue to cloud task that launches the /run-my-function?
Is this the right way to do it or I have misunderstand something? In case it's the right way how to specify the url of the /run-my-function endpoint without explicitly hard-code the cloud run deployed uRL name?
The code for the endpoint that launches the endpoint with the run-my-function code would be:
from google.cloud import tasks_v2
client = tasks_v2.CloudTasksClient()
project = 'myproject'
queue = 'myqueue'
location = 'mylocation'
url = 'https://cloudrunservice-abcdefg-ca.b.run.app/run-my-function'
service_account_email = '12345#cloudbuild.gserviceaccount.com'
parent = client.queue_path(project, location, queue)
task = {
"http_request": {
"http_method": tasks_v2.HttpMethod.POST,
'url': url,
"oidc_token": {"service_account_email": service_account_email},
}
}
response = client.create_task(parent=parent, task=task)
However, this requires to hard-code the service name https://cloudrunservice-abcdefg-ca.b.run.app and to define an auxiliary endpoint /run-my-function that can be called via http

In your code you are able to get the Cloud Run URL without hardcoding it or setting it in an environment variable.
You can have a look to a previous article that I wrote, in the gracefull termison part. I provide a working code in Go, not so difficult to re-implement in Python.
Here the principle:
Get the Region and the project Number from the Metadata server. Keep in mind that Cloud Run has specific metadata like the region
Get the K_SERVICE env var (it's a standard Cloud Run env var)
Perform a call to the Cloud Run Rest API to get the service detail and customize the request with the data got previously
Extract the status.url JSON entry from the response.
Now you have it!
Let me know if you have difficulties to achieve that. I'm not good at Python, but I will be able to write that piece of code!

Related

Determine the project used by Vision Api

I'm working on a project on gcp using the OCR service from Google Vision.
My code is simple, I've declared a vision ImageAnnotatorClient like this:
client = vision.ImageAnnotatorClient()
and then call some batch annotation request like this:
response = client.batch_annotate_images(requests=[....])
There is a way to know which project will be used for the billing of the requests?
I've not explicitly set the credential and, in this case, the documentation (https://googleapis.dev/python/vision/latest/vision_v1/image_annotator.html) says that
if none are specified, the client will attempt to ascertain the credentials from the environment.
I need to know what "ascertain the credentials from the environment" means in this case, since I've alrteady done some requests and I wold like to know on which of the projects I'm working right now the requests will be charged.

How to share a global dict between Cloud Run instances in python?

I'm building a flask server in python with Cloud Run, for a chatbot to call.
Sometimes if user wants to do something with the chatbot, the bot need ask the user to login to a 3rd party server before doing the things.
I have two routes:
Route 1 is "/login", it returns a simple iframe which will open a login page in a 3rd party server, generate a "session_id", and save some info I already get to a global variable dict called "runtimes" with the "session_id" as key, so that I can use it later when visitor successfully logged in.
Route 2 is "/callback/<session_id>". After user successfully login to its account, the 3rd party server will call this route with a token in url parameters. Then I will use the "session_id" to read the saved info from "runtimes", and do later things.
It works well in my local machine. But in Google Cloud Run, because it support multiple instances, sometimes it will trigger a new instance when server calls "callback", so it cannot get the "runtime" because they are in different instances.
I know that I can save the runtimes dict to a database to solve this problem, but it looks too overkill...Just not seem right.
Is there any easy way that I can make the "runtimes" be shared between instances?
The solution here is to use a central point of storage: database, memorystore, firestore,... something out of Cloud Run itself.
You can also try the Cloud Run execution runtime v2 that allow you to mount a network disk, such as Cloud Storage or Filestore. You can imagine to store the session data in a file which has the name of the session ID.
Note: On Cloud Run side, something is cooking, but it's not 100% safe, it will be a best effort. A database backup will be required even with that new feature

calling a google cloud function from google app script

I have a Google Cloud Function that I would like to call from my Google App Script on a Google Form submission.
The process will be: 1)user submits google form, 2)there will be a trigger (onformsubmit) that will run the app script function 3) app script function will trigger cloud function.
So far:
The script trigger works, in the logs it's listening correctly.
The cloud function works, I tested it in the Cloud function testing interface and when I run it from there, it does what I need it to do which is to update a google sheet as well as upload data to BigQuery.
The problem comes from calling that function from App Script that I have associated with my google form submission trigger. There seems to be no communication there, as cloud function logs don't show anything happening at trigger submission.
This is my app script code:
function onSubmit() {
var url = "myurl"
const token = ScriptApp.getIdentityToken()
var options = {
'method' : 'get',
'headers': {"Authorization":"Bearer "+ token}
};
var data = UrlFetchApp.getRequest(url,options);
return data
}
And my Cloud function is a HTTP one in Python and starts with:
def numbers(request):
Some troubleshooting:
When I test it, the execution log shows no errors
If I try to change UrlFetchApp to .fetch or change getIdentityToken to
getOAuthToken I get a 401 error for both
I added the following to my oauthScopes:
"openid",
"https://www.googleapis.com/auth/cloud-platform",
"https://www.googleapis.com/auth/script.container.ui",
"https://www.googleapis.com/auth/script.external_request",
"https://www.googleapis.com/auth/documents"```
I'm running both from the same Google Cloud account
I added myself to permissions in Cloud Function settings too
Any ideas of why the two aren't communicating would be appreciated!
I was able to resolve this in case anyone has a similar issue. Since my email address was associated with an organizational account, my Apps Script and GCP didn't allow the correct permissions.
In the settings of Apps Script, I couldn't change the GCP account with that function because the GCP was outside of my organization. Once I set up the Cloud Function on my organizations GCP, I was able to change the account manually in the settings and my function worked properly on the trigger.

setting up a python http function in firestore

I have an app that is meant to integrate with third-party apps. These apps should be able to trigger a function when data changes.
The way I was envisioning this, I would use a node function to safely prepare data for the third parties, and get the url to call from the app's configuration on firestore. I would call that url from the node function, and wait for it to return, updating results as necessary (actually, triggering a push notification). -- these third-party functions would tend to be python functions, so my demo should be in python.
I have the initial node function and firestore setup so that I am currently triggering a ECONNREFUSED -- because I don't know how to set up the third-party function.
Let's say this is the function I need to trigger:
def hello_world(request):
request_json = request.get_json()
if request_json and 'name' in request_json:
name = request_json['name']
else:
name = 'World'
return 'Hello, {}!\n'.format(name)
Do I need to set up a separate gcloud account to host this function, or can I include it in my firestore functions? If so, how do I deploy this to firestore? Typically with my node functions, I am running firebase deploy and it automagically finds my functions from my index.js file.
If you're asking whether Cloud Functions that are triggered by Cloud Firestore can co-exist in a project with Cloud Functions that are triggered by HTTP(S) requests, then the answer is "yes they can". There is no need to set up a separate (Firebase or Cloud) project for each function type.
However: when you deploy your Cloud Functions through the Firebase CLI with firebase deploy, it will remove any functions that it finds in the project, that are not in the code. If you have functions both in Python and in Node.js, there is never a single codebase that contains both, so a blanket deploy would always delete some of your functions. So in that case, you should use the granular deploy option of the Firebase CLI.

Is there a better way to access my public api?

I am new to Flask.
I have a public api, call it api.example.com.
#app.route('/api')
def api():
name = request.args.get('name')
...
return jsonify({'address':'100 Main'})
I am building an app on top of my public api (call it www.coolapp.com), so in another app I have:
#app.route('/make_request')
def index():
params = {'name':'Fred'}
r = requests.get('http://api.example.com', params=params)
return render_template('really_cool.jinja2',address=r.text)
Both api.example.com and www.coolapp.com are hosted on the same server. It seems inefficient the way I have it (hitting the http server when I could access the api directly). Is there a more efficient way for coolapp to access the api and still be able to pass in the params that api needs?
Ultimately, with an API powered system, it's best to hit the API because:
It's user testing the API (even though you're the user, it's what others still access);
You can then scale easily - put a pool of API boxes behind a load balancer if you get big.
However, if you're developing on the same box you could make a virtual server that listens on localhost on a random port (1982) and then forwards all traffic to your api code.
To make this easier I'd abstract the API_URL into a setting in your settings.py (or whatever you are loading in to Flask) and use:
r = requests.get(app.config['API_URL'], params=params)
This will allow you to make a single change if you find using this localhost method isn't for you or you have to move off one box.
Edit
Looking at your comments you are hoping to hit the Python function directly. I don't recommend doing this (for the reasons above - using the API itself is better). I can also see an issue if you did want to do this.
First of all we have to make sure the api package is in your PYTHONPATH. Easy to do, especially if you're using virtualenvs.
We from api import views and replace our code to have r = views.api() so that it calls our api() function.
Our api() function will fail for a couple of reasons:
It uses the flask.request to extract the GET arg 'name'. Because we haven't made a request with the flask WSGI we will not have a request to use.
Even if we did manage to pass the request from the front end through to the API the second problem we have is using the jsonify({'address':'100 Main'}). This returns a Response object with an application type set for JSON (not just the JSON itself).
You would have to completely rewrite your function to take into account the Response object and handle it correctly. A real pain if you do decide to go back to an API system again...
Depending on how you structure your code, your database access, and your functions, you can simply turn the other app into package, import the relevant modules and call the functions directly.
You can find more information on modules and packages here.
Please note that, as Ewan mentioned, there's some advantages to using the API. I would advise you to use requests until you actually need faster requests (this is probably premature optimization).
Another idea that might be worth considering, depending on your particular code, is creating a library that is used by both applications.

Categories

Resources