Have questions about serverless architectures? Tune into our #AskGoogleCloud premiere on Friday, March 12 10AM PT for answers and a chance to chat live with Google Cloud’s serverless experts → goo.gle/3rFCJFl
Would love to see an API Gateway episode where you also cover exposing your API to 3-rd party developers and monitor their usage for the purpose of billing. Since AWS has a pretty strong offering on this front I'd like to see if Google can match it (with the Doc. generation/portal, SDK generation, etc.) Cheers :)
Excellent combination of overview and detail "how to" . How about another with multiple microservices behind a Google API Gateway with one or more deploy scripts
Where can we reach the selection grid mentioned at around 1:07 ? I generally miss such selection grids of GCP products in different depth, detail levels and targeted content consumer type ( business, dev, etc. ).
It would be nice if they show how the docker file looks for cloud run. Also a question: would cloud run a good option to host small web applications? (Like the REST example in the video but also have some http/React/Vue frontend things hosted as well via the container)
One more problem, would the "auto-sccale-to-0" be a cold start problem (high latency for the init call) if my site has very low traffic? Would it possible to config it the minimum container or pod running is one (or two if there is multiple version of code needs to run at the same time)?
@@FuyangLiu If you have very little traffic and are concerned that your users will experience a lot of cold starts, you should consider the new min-instances feature of Cloud Run: cloud.google.com/run/docs/configuring/min-instances. Hope this helps!
Agreed, App Engine is Serverless. It is also PaaS (Platform as a Service). In my book, all PaaS tools are also Serverless. But I am aware that some people draw a clear distinction between PaaS and Serverless. It's confusing.
Awesome. Great tutorial. Suggestion for future tutorial : An example tutorial for a full blown E2E "cloud native application deployment". I think I understand what the cloud functions, cloud app engine, and cloud run do. - But I want to see a real use case deployment where all these parts used and work together. - And to explain us some best practices when to use which in that example, on how to connect those pieces. - Not an in-depth tutorial for each parts, but rather an high level of how all those different pieces are working together. - Also, how the "Operation Suite" can be used in this entire workflow for monitoring/debugging. Thanks.
Also if you'd like a more real-world-like examples on Cloud Functions deployment management you might want to check out these boilerplates for http-triggered and background-triggered functions… github.com/buffolander/boilerplate-cloud-functions-http github.com/buffolander/boilerplate-cloud-functions-background
Cloud Functions gives a single managed HTTPS URL out of the box. How can i setup multiple endpoints using Cloud Functions for example a usecase to deploy REST APIs in Go ? Is it possible to achieve this without an API Gateway?
I think Cloud Run works better than Cloud Functions for REST APIs with multiple endpoints. With Cloud Run you still get an HTTPS URL out of the box, but it's easier to use common web frameworks like Gin for Go, Express for NodeJS, or Flask for Python. That makes it easier to provide the multiple endpoints that a REST API needs.
Hi Rahul, this is a great question and we actually answer it in our first episode of #AskGoogleCloud that’s premiering tomorrow March 12th at 10AM PT → goo.gle/3qFBa95 We’ll also have serverless experts who are going to be answering questions in real-time in the live chat. Drop by to ask your questions or say hello!
I've found Cloud Run or Cloud Functions are great for microservices. If you come from a Kubernetes background or want even more mangement of your services K8 works well too.
I build a similar POC for my beta project by using Cloud Function + PubSub + on-demand VM + Cloud object storage + Redis. I will use firebase in the future as it cheaper compares to Redis for Key-Value storage.
You can get the URL of a file in a Cloud Storage bucket in one of two ways: 1. Go to Cloud Storage and the right bucket in the Cloud Console. Click the "Copy URL" item next to the file you want. 2. Or compose the URL like this: BUCKET_NAME.storage.googleapis.com, where BUCKET_NAME is the name of the bucket you created.
Hi, great tutorial, I'm currently working on a similar project that requires cameras streaming images to the cloud, based on the concept you used is it possible to get info about the particular camera that uploaded the image
You can deploy source code without having to containerize it. Simply run "gcloud run deploy [SERVICE_NAME] --source ." in the same directory as your code. You won't need to create a Dockerfile or know anything about containers.
One quick difference: Cloud Run is vastly simpler. You essentially provide a container image and they run it. Also, you can have several cloud run services available in the same project. I noticed w/ App Engine (w/ my limited use) only allowed a single application per project, as it's much more complex. At least with App Engine, it supports WebSockets and minimum instances. Also: Cloud Run is *WAY* cheaper. It's also super easy to deploy to, IMHO.
Hi Senju, this is a great question and we actually answer it in our first episode of #AskGoogleCloud that’s premiering tomorrow March 12th at 10AM PT → goo.gle/3rFCJFl We’ll also have serverless experts who are going to be answering questions in real-time in the live chat. Drop by to ask your questions or say hello!
Have questions about serverless architectures?
Tune into our #AskGoogleCloud premiere on Friday, March 12 10AM PT for answers and a chance to chat live with Google Cloud’s serverless experts → goo.gle/3rFCJFl
Grant is arguably the most cheerful guy explaining tech topics on youtube.
@@AndreiGeorgescu-j9p although really simple explanations
Watching your videos gives me calmness.
Thanks to the Googlers that create these videos, they keep me engaged while having fun watching the content.
Hi Brian! Glad you like them :)
Thanks!! Just Started learning about Google Cloud and found this video useful.
I am getting more insights about the terminology that I am learning. This is very good.
Thanks a lot for sharing!! It was very useful. Greetings from Uruguay
Would love to see an API Gateway episode where you also cover exposing your API to 3-rd party developers and monitor their usage for the purpose of billing.
Since AWS has a pretty strong offering on this front I'd like to see if Google can match it (with the Doc. generation/portal, SDK generation, etc.)
Cheers :)
Excellent combination of overview and detail "how to" . How about another with multiple microservices behind a Google API Gateway with one or more deploy scripts
Excellent idea -- thank you, Glenda!
Such positive vibes and great information. What a cool end to the vid! Sense of humour too.
Manage traffic is a great feature. Especially with new app or new app feature deployments.
Where can we reach the selection grid mentioned at around 1:07 ? I generally miss such selection grids of GCP products in different depth, detail levels and targeted content consumer type ( business, dev, etc. ).
Great video, but note that that's a Green Anole, not a gecko :) They're very common here in South Louisiana. Cheers!
Thank you for the correction!
It would be nice if they show how the docker file looks for cloud run. Also a question: would cloud run a good option to host small web applications? (Like the REST example in the video but also have some http/React/Vue frontend things hosted as well via the container)
One more problem, would the "auto-sccale-to-0" be a cold start problem (high latency for the init call) if my site has very low traffic? Would it possible to config it the minimum container or pod running is one (or two if there is multiple version of code needs to run at the same time)?
Hi Fuyang! The Docker file is at github.com/momander/serverlesstoolbox/tree/master/critter-watch/critter-api
@@FuyangLiu If you have very little traffic and are concerned that your users will experience a lot of cold starts, you should consider the new min-instances feature of Cloud Run: cloud.google.com/run/docs/configuring/min-instances. Hope this helps!
@@TheMomander Thank you very much that sounds like some thing I want to try 👍👍👍
Nice and useful content, and great performance of the teachers. Thank you, Grant and Martin.
In 1:05, you showed App Engine as a Serverless Cloud, but it should be PaaS. Am I wrong?
Agreed, App Engine is Serverless. It is also PaaS (Platform as a Service). In my book, all PaaS tools are also Serverless. But I am aware that some people draw a clear distinction between PaaS and Serverless. It's confusing.
Awesome. Great tutorial.
Suggestion for future tutorial : An example tutorial for a full blown E2E "cloud native application deployment".
I think I understand what the cloud functions, cloud app engine, and cloud run do.
- But I want to see a real use case deployment where all these parts used and work together.
- And to explain us some best practices when to use which in that example, on how to connect those pieces.
- Not an in-depth tutorial for each parts, but rather an high level of how all those different pieces are working together.
- Also, how the "Operation Suite" can be used in this entire workflow for monitoring/debugging.
Thanks.
That's a great idea; thank you! I'm adding to our list of future episodes now.
Also if you'd like a more real-world-like examples on Cloud Functions deployment management you might want to check out these boilerplates for http-triggered and background-triggered functions…
github.com/buffolander/boilerplate-cloud-functions-http
github.com/buffolander/boilerplate-cloud-functions-background
@@buffolander9927 thanks, will look into it.
Great tutorial
Cloud Functions gives a single managed HTTPS URL out of the box. How can i setup multiple endpoints using Cloud Functions for example a usecase to deploy REST APIs in Go ? Is it possible to achieve this without an API Gateway?
I think Cloud Run works better than Cloud Functions for REST APIs with multiple endpoints. With Cloud Run you still get an HTTPS URL out of the box, but it's easier to use common web frameworks like Gin for Go, Express for NodeJS, or Flask for Python. That makes it easier to provide the multiple endpoints that a REST API needs.
Really helpful and thanks. By the way, the code samples .... are they on github.
Is Cloud Trace also available automatically for the Cloud Run Service?? For app-engine it is automatically available in the cloud trace dashboard.
Hi Rahul, this is a great question and we actually answer it in our first episode of #AskGoogleCloud that’s premiering tomorrow March 12th at 10AM PT → goo.gle/3qFBa95
We’ll also have serverless experts who are going to be answering questions in real-time in the live chat. Drop by to ask your questions or say hello!
This is really nice, thank you!
Thanks for this content! Do you think cloudrun is suitable for micro services? Or would a kubernetes cluster would be better for such a use case?
I've found Cloud Run or Cloud Functions are great for microservices. If you come from a Kubernetes background or want even more mangement of your services K8 works well too.
I build a similar POC for my beta project by using Cloud Function + PubSub + on-demand VM + Cloud object storage + Redis. I will use firebase in the future as it cheaper compares to Redis for Key-Value storage.
Nice! Firestore is a nice key-value storage that works well with Cloud Functions and a bunch of serverless products.
very nice tutorial! thanks guys
Anytime
Another awesome content.Omar ,can you let us know how you arrived at the photourl for the content in the s3 bucket?
You can get the URL of a file in a Cloud Storage bucket in one of two ways:
1. Go to Cloud Storage and the right bucket in the Cloud Console. Click the "Copy URL" item next to the file you want.
2. Or compose the URL like this: BUCKET_NAME.storage.googleapis.com, where BUCKET_NAME is the name of the bucket you created.
Hi, great tutorial, I'm currently working on a similar project that requires cameras streaming images to the cloud, based on the concept you used is it possible to get info about the particular camera that uploaded the image
Useful
Great video! Where can I find this handy selection grid?
Excellent tutorial!!
Thanks Alberto!
#AskGoogleCloud How do they compare in terms of local development, can I stop code in a debugger? Can I use the local emulator?
Does Cloud Run only serve containerized code ?
You can deploy source code without having to containerize it. Simply run "gcloud run deploy [SERVICE_NAME] --source ." in the same directory as your code. You won't need to create a Dockerfile or know anything about containers.
An amazing content. Thank you !
Thank you!
Great explanation , good job!!!
Thanks Vaibhav!
you could explain the difference between app engine flexible and cloud run.
One quick difference: Cloud Run is vastly simpler. You essentially provide a container image and they run it. Also, you can have several cloud run services available in the same project. I noticed w/ App Engine (w/ my limited use) only allowed a single application per project, as it's much more complex. At least with App Engine, it supports WebSockets and minimum instances. Also: Cloud Run is *WAY* cheaper. It's also super easy to deploy to, IMHO.
Please make Cloud Function available for Spark plan.
They are available in spark plan, but you are limited: no cron jobs, no egress, node 8 and usage limits are also smaller then in free blaze plan
Thanks, this was an engaging session!
Hi Vaibhav, we're so glad you found this video engaging :)
That was interesting
Can you please explain how to use cloud identity in GCP for user management?
Sounds like a great topic for another video. We're adding it to our list. Thanks you!
Thanks for for this video. can you share Cloudrun code to try on my end.
does it blend indeed
Greetings from Perú haha :)
Hey Iván! I loved Peru when I was there :) Maracuyá Sours were my favorite!
IMPRESSIVE 💖👍
Thanks!
Does cloud run support websocket?
Hi Senju, this is a great question and we actually answer it in our first episode of #AskGoogleCloud that’s premiering tomorrow March 12th at 10AM PT → goo.gle/3rFCJFl
We’ll also have serverless experts who are going to be answering questions in real-time in the live chat. Drop by to ask your questions or say hello!
Yes, it does. Do a UA-cam search for "Websockets with Cloud Run" and you will find a video where we explain how it works.
can you share the files?
soon we will have cc cameras detecting criminals around our houses collecting face recognition from criminals database.
We've only implemented animal tracking, but the code is open source :)
github.com/momander/serverlesstoolbox/tree/master/critter-watch
recomendo AWS