Steren Profile picture
@Google Cloud Run (https://t.co/hdiUrPJ9Gc) Product Manager

Apr 9, 2019, 67 tweets

🚀 Today at #GoogleNext19, we are launching Cloud Run.
Allowing you to run any stateless http container in a fully managed environment, paying only for the exact resources you use. 📦☁️
👉 cloud.google.com/run

😕 Developers love serverless, but feel limited by the available programming languages or the need to rewrite using functions.
👉 Cloud Run brings the properties of serverless to containers: no infra management, fast autoscaling and pay per use.

Sometimes, you want more control or have to run on top of Kubernetes, we are also introducing Cloud Run on GKE. Allowing you to run the same serverless workloads with the same developer experience, in your own GKE cluster. cloud.google.com/run/docs/quick…

🙏 to the thousands alpha and beta testers who have been using Cloud Run in the past months, providing feedback and reporting issues. You helped us shape Cloud Run!

Cloud Run and Cloud Run on GKE both implement the @KnativeProject API. An OSS project we introduced last year.
This allows you to deploy the same autoscaled services to a fully managed GCP environment, to your own GKE cluster, or anywhere #Kubernetes runs with #Knative.

😎I started a collection of awesome Cloud Run related things (articles, docs, containers, tools): github.com/steren/awesome…
Feel free to send suggestions via Pull Requests.

🎬 How to build with Cloud Build and deploy to Cloud Run by @BretMcG and @martinomander:

🌍 Cloud Run integrates with @Firebase Hosting.
This allows you to create webapps with dynamic endpoints using Cloud Run. All of this behind a global CDN out of the box: firebase.google.com/docs/hosting/c…

@Firebase Very excited to partner with @stackblitz for this amazing in browser developer experience:
1. Start by picking your stack
2. Get a GitHub repo and the prod app on Cloud Run
3. Iterate on the code using an in-browser VS Code-like editor

Read more at medium.com/stackblitz-blo…

@Firebase @stackblitz And we are very proud to count @datadog as a launch partner, allowing you to monitor your Cloud Run and Cloud Run on GKE services: datadoghq.com/blog/monitor-g…

@Firebase @stackblitz @datadog If you encounter issues or have questions about Cloud Run, please post on @StackOverflow with the tag `google-cloud-run`. Many googlers and myself are monitoring this tag and will be happy to help you

stackoverflow.com/questions/tagg…

@Firebase @stackblitz @datadog @StackOverflow As the name implies, Cloud Run pairs great with Cloud Build.

Did you know that you can set up Continuous Deployment of your git repository to Cloud Run using Cloud Build?
Docs at 👉 cloud.google.com/run/docs/conti…

@Firebase @stackblitz @datadog @StackOverflow Cloud Run now has direct support for Cloud SQL (in alpha). When enabled, we activate and configure the Cloud SQL proxy for you.

See temporary instructions here: stackoverflow.com/a/55892255/481…
We are working on publishing docs and moving it to Beta within 10 days. (cc @gabidavila)

@Firebase @stackblitz @datadog @StackOverflow @gabidavilacloud.run now redirects to the Cloud Run product page 😎

@Firebase @stackblitz @datadog @StackOverflow @gabidavila One month after the launch of Cloud Run, we are pleased to launch several new features to address your top requests:
🧮Cloud SQL support,
📈Metric at a glance,
🌍Europe and 🌏Asia regions opening soon.

Read more: cloud.google.com/blog/products/…

@Firebase @stackblitz @datadog @StackOverflow @gabidavila As promised, Cloud Run (cloud.run) is now available in 🌍Europe and 🌏Asia.
You can now picked from these 4 GCP regions:
- asia-northeast1 (Tokyo)
- europe-west1 (Belgium)
- us-central1 (Iowa)
- us-east1 (South Carolina)

@Firebase @stackblitz @datadog @StackOverflow @gabidavila Cloud Run will now only support HTTPS (because it's 2019)

@Firebase @stackblitz @datadog @StackOverflow @gabidavila 🏷️You can now set labels on your Cloud Run services.
cloud.google.com/run/docs/confi…

You can then filter billing, logs or metrics.
Use cases include cost center / team ownership, environments (env:prod)...

@Firebase @stackblitz @datadog @StackOverflow @gabidavila 🗄And today, we are announcing Cloud Run for Anthos, allowing you to get the same Cloud Run developer experience and scalability on-premises:
cloud.google.com/blog/topics/hy…

@Firebase @stackblitz @datadog @StackOverflow @gabidavila 👩‍💻The Cloud Run API is now "v1": cloud.google.com/run/docs/refer…

This new API is an implementation of the recently released Knative Serving v1 API.

"v1" means that you can expect stability and long term support. One step closer to general availability!

@Firebase @stackblitz @datadog @StackOverflow @gabidavila 📊 Many of you wanted to better understand how much Cloud Run scales and how many instances are scheduled. We added a new chart to the Metrics tab that should answer this question.

E.g. if ou see "2s/s", that means 2 instances were continuously serving traffic in this second.

@Firebase @stackblitz @datadog @StackOverflow @gabidavila 💥Since Beta, Cloud Run has integrated with Stackdriver Error Reporting out of the box (cloud.google.com/run/docs/error…)
We are now surfacing a summary of your top errors right in the Cloud Run user interface:

@Firebase @stackblitz @datadog @StackOverflow @gabidavila ☑️If you have 2 minutes, help us prioritize the next Cloud Run features by taking this short survey: forms.gle/irVTCQ5LTVLPXT…

🏥 Serverless containers for healthcare: Cloud Run is now covered by GCP HIPAA compliance cloud.google.com/security/compl…

🎬 I created a Youtube playlist of my recommended Cloud Run related videos:

🏅 Cloud Run is now Generally Available cloud.google.com/blog/products/…

🤝A broad ecosystem of partners are supporting Cloud Run: CI/CD, Observability and Security
cloud.google.com/blog/topics/pa…

🆕Along with General Availability, came many new cloud.run features.
@glaforge and I wrote about them in
cloud.google.com/blog/products/…

@glaforge 📜 The Service Level Agreement (SLA) for Cloud Run (fully managed) has been updated to 99.95% cloud.google.com/run/sla

@glaforge 📄The cloud.run product page has been re-designed and now includes a list of common use cases cloud.google.com/run/#use-cases

@glaforge If storing config as code or implementing GitOps is your thing, you will be glad to hear that you can now deploy to cloud.run from a local YAML file:

🌍🌎🌏 Cloud Run is now available in these 5 GCP regions: North Virginia, Oregon, Netherlands, Finland, and Taiwan

🔵🟢 Cloud Run now has support for rollbacks, gradual rollouts (blue/green deployments), and other traffic migration manipulations between Cloud Run Revisions.
Use the Cloud Console or `gcloud run services update-traffic`. More at cloud.google.com/run/docs/rollo…

🏗️The cloud.run user interface now displays link to the build information of the deployed container, as well as the repo and commit used to build it (if you are using a Cloud Build Trigger)

🖧 Your Cloud Run services can now connect to VPC networks, for example to access Memorystore Redis and Memcached.
Use it now:
1. Create a Serverless VPC Access connector:
cloud.google.com/vpc/docs/confi…
2. `$ gcloud alpha run services update SERVICE --vpc-connector CONNECTOR`

🏗️ @GCPcloud has released a set of Buildpacks to help you build containers in Go, Node.js, Python, Java and .Net without having to maintain any Dockerfile.
The resulting containers are optimized for production and for the Cloud Run environment.
github.com/GoogleCloudPla…

📝Cloud Run is now supported in Cloud Code for both @code and @intellijidea.
1 Bootstrap with samples
2 Explore your services
3 Deploy

- VSCode: marketplace.visualstudio.com/items?itemName…
- IntelliJ: plugins.jetbrains.com/plugin/8079-cl…

📋The Cloud Run user interface now allows you to copy a Cloud Run service.
Handy if you want to change your platform or move to a different region.

✏️ The user interface now also allows you to directly edit the service YAML
Handy if you prefer to edit raw config files instead of clicking the nice buttons we designed for you :)

🗺️Cloud Run is now available in Osaka, Sydney and Montréal!

You can now use Cloud CDN on top of Cloud Run.
By sending the right `Cache-Control` headers, you can drastically reduce the load on your Cloud Run service.

Get Cloud CDN via External HTTP(S) Load Balancing: cloud.google.com/load-balancing…

🌎🌍🌏Expose a global endpoint that routes requests to the closest Cloud Run service by using a HTTPS Global Load Balancer.
cloud.google.com/load-balancing…

📄🚀 Set up Continuous Deployment from git to Cloud Run in just a few click in the Cloud Run UI.
This flow is powered by a Cloud Build trigger that automatically builds and deploys when new changes are pushed to your repo. cloud.google.com/run/docs/conti…

Cloud Run can now connect to Shared VPCs.
Allowing you to connect to private IPs of resources from other GCP projects of the same organization.
Read more:
cloud.google.com/run/docs/confi…

🌏 Cloud Run is now available in Singapore (asia-southeast1)

🔵🟢 You can now reach a specific Cloud Run revision by "tagging" it, it is then available at a dedicate URL https://[tag]---[servie-url].a.run.app
This completes the gradual rollout story.
Read more on the @GCPcloud blog: cloud.google.com/blog/products/…

🔒Data Residency: We made it clearer that when you create a Cloud Run service in a region, your data and container stay in that region.
cloud.google.com/terms/data-res…

📄📦 When setting up Continuous Deployment from the Cloud Run user interface, you can now select a repository that contains Go, Node.js, Python Java or .NET Core code. It will be built using Google Cloud Buildpacks without needing a Dockerfile (github.com/GoogleCloudPla…)

📈 You can now pick up to 4GiB of RAM for your Cloud Run services.

⛩️ Introducing Google Cloud API Gateway. A fully managed gateway for your Cloud Run services.

See it in action at
Read more at cloud.google.com/api-gateway

We partnered with @pomerium_io, an open source identity-aware proxy, to enable you to build secure internal applications on Cloud Run.
cloud.google.com/community/tuto…

📝 Cloud Code for VSCode and IntelliJ now has support for running your Cloud Run locally, allowing you to run or debug in an emulated environment.

Blog post: cloud.google.com/blog/products/…

📺 Cloud Run now supports server-side HTTP and gRPC streaming.
This means HTTP responses are not limited to 32MB anymore if you use chunked transfer-encoding.

🕒 From 15min to 1h!
Your Cloud Run service can now take up to 3600 seconds to return a request.

💻💻💻💻 You can now allocate 4 vCPUs to your Cloud Run container instances.

🛑 Cloud Run now sends a SIGTERM signal before a container instance is shut down. If handled, CPU is allocated for 10s max

This allows for graceful instance termination, giving your code some time to close any database connection or send some last telemetry data.

🌐 Static outbound IP:
You can now route all outbound requests to your VPC network, allowing you to configure a static outbound IP address via Cloud NAT.
Check out @ahmetb's tutorial: cloud.google.com/run/docs/confi…

❄️🔥 Goodbye cold starts: You can now specify a minimum number of container instances to be kept warm and ready to serve requests.
When they are not serving requests, CPU is 10x cheaper than normal.
cloud.google.com/run/docs/confi…

🗺 Bringing Cloud Run to even more GCP regions: Hong Kong, Seoul, Jakarta, Mumbai, London, Frankfurt, Zurich, Sao Paulo. 21 regions and counting...

Did you know you can purchase domain names right from the GCP console with Cloud Domains?

The Cloud Run user interface now lets you to do that in context:

Cloud Run now integrates out of the box with Cloud Trace, capturing request latencies, allowing you to identify latency issue easily with its trace graph:

To help us shape the future of Cloud Run, participate in our research studies: g.co/userresearch/x…

⛗ Each Cloud Run container instance is now able to receive up to 250 concurrent requests (default is still 80).

🔎 We published a tutorial showing how to use Cloud Run revision URLs to implement previews of GitHub pull requests: cloud.google.com/run/docs/tutor…

📄 To simplify getting started with Cloud Run, we added a single command to build and deploy your source code:
$ gcloud beta run deploy --source .

- If a Dockerfile is present, it's used
- If not, language is automatically detected with GCP Buildpacks

cloud.google.com/blog/products/…

📈 You can now pick up to 8GiB of RAM for your Cloud Run services.

📶Cloud Run now allows to restrict ingress of your services to "internal traffic" and "internal traffic and traffic from Cloud Load Balancing".
Basically disabling the default endpoint, making the service "private" from a networking perspective.
cloud.google.com/run/docs/secur…

Share this Scrolly Tale with your friends.

A Scrolly Tale is a new way to read Twitter threads with a more visually immersive experience.
Discover more beautiful Scrolly Tales like this.

Keep scrolling