Restate

Restate

@restate.dev

Restate is the platform for building resilient applications that tolerate all infrastructure faults w/o the need for a PhD. https://restate.dev

235 followers5 following56 posts

Restate reposted

Restate v1.8.1 is out 🚀 🌊 Smooth hand-over when migrating Kafka topics to the new Kafka integration 🗑️ Remove several deployments at once ☸️ Kubernetes ConfigMap changes applied without restarts 🐛 Plus fixes, security updates and UI updates github.com/restatedev/r...

github.comRelease v1.8.1 · restatedev/restateRestate v1.8.1 v1.8.1 is a drop-in upgrade from v1.8.0, and you can roll back to v1.8.0 unless you enable signal recording. If you are still on v1.7.x, follow the v1.8.0 upgrade requirements. ✨ Hig...

Restate reposted

🚀 Restate 1.7.3 is out! Existing clusters can now migrate to VQueues and adopt flow control without reprovisioning. Also included: important memory-retention, reliability, metrics, and SQL fixes. Upgrade: github.com/restatedev/r...

github.comRelease v1.7.3 · restatedev/restateRestate 1.7.3 Restate 1.7.3 makes the flow-control capabilities introduced in v1.7 available to existing clusters. It also includes important fixes for memory retention, Prometheus compatibility, S...

Restate + Vercel AI SDK = enterprise-grade durability for your AI agents with just a few lines of code. Survive crashes, timeouts, and restarts - even for multi-turn conversations, human approval and long-running research agents. @vercel.com @sdk.vercel.ai

Restate reposted

AWS’s us-east-1 outage took down ChatGPT, Snapchat & Perplexity. The same happened in 2021: Netflix, Disney+, and Robinhood all went dark. The reason? Often a lack of true geo-replication. Here's how you can run apps with zero downtime during a region failure using Restate 🧵

View full thread

We ran a Restate app deployed in multiple regions, then killed an entire region. Result: ✅ Zero downtime ✅ Auto-failover in <60s ✅ No data loss ✅ All requests processed

AWS’s us-east-1 outage took down ChatGPT, Snapchat & Perplexity. The same happened in 2021: Netflix, Disney+, and Robinhood all went dark. The reason? Often a lack of true geo-replication. Here's how you can run apps with zero downtime during a region failure using Restate 🧵

Traditional geo-replication is brutally hard: - Complex distributed state management - Custom replication logic & conflict resolution - Leader election - Restructuring your entire application code It's an expert-level undertaking that costs companies months of engineering time.

With Restate, geo-replication is a deployment detail, not an application concern. Write normal business logic. Deploy across regions with one config line: 'default-replication = "{region: 2, node: 3}"'

​Last week, we announced that Restate Cloud is now publicly available. Join us on Wednesday for a hands-on walkthrough of Restate Cloud. Learn how to quickly set up your first project, and start building resilient applications, without managing any infra. luma.com/bdspjxof

🚀 Yesterday, we announced that 𝗥𝗲𝘀𝘁𝗮𝘁𝗲 𝗖𝗹𝗼𝘂𝗱 is now publicly available. Today, we’re announcing 𝗥𝗲𝘀𝘁𝗮𝘁𝗲 𝟭.𝟱! This release brings key features for building, controlling and observing your durable agents and workflows.

The Restate UI has improved so much over the last releases. The UI now shows a live timeline of execution steps, including retries, nested RPC calls, events, and even cancellation signals. And it's so beautiful!

🎉 Exciting news! 𝗥𝗲𝘀𝘁𝗮𝘁𝗲 𝗖𝗹𝗼𝘂𝗱 is now publicly available! Build workflows and AI agents with durable execution, observability, and production-grade availability without running any stateful infra yourself.

Restate reposted

🎥 Build Durable AI Agents with Restate + Vercel AI SDK Here are the highlights from our recent community meeting on building resilient AI agents using Restate with your favorite Agent SDK (like Vercel AI or OpenAI).

Join the community meeting next week to learn how combining Restate with your agent SDK (Vercel AI, OpenAI, ...) gives your agents a set of powerful features.

🚀 Restate 1.4 just dropped! 10x faster failure detection, automatic workload rebalancing, and 2.5x quicker startup. Your distributed applications just got more resilient with less downtime and better resource utilization. Build more, worry less. #Restate #Resilience restate.dev/blog/announc...

restate.devRestate 1.4: We've Got Your Resiliency CoveredRestate 1.4 improves cluster resiliency and workload balancing significantly. Experience less unavailability and achieve more with fewer resources.

Restate auto-retries temporary issues like network glitches. But for business logic failures ("Hotel fully booked"), retrying won't help. Instead, Restate helps you gracefully undo previous steps using sagas - each action has a matching "undo" step that runs when needed.

View full thread

When to use Sagas: - Business logic failures ("hotel full", "product not available") where retrying won't help - Cancelling ongoing requests, and letting them roll back Your system never ends up in a half-finished state!

Restate auto-retries temporary issues like network glitches. But for business logic failures ("Hotel fully booked"), retrying won't help. Instead, Restate helps you gracefully undo previous steps using sagas - each action has a matching "undo" step that runs when needed.

With Restate, implementing sagas is easy: ✅Record compensating (undo) actions in a list as your workflow progresses ✅ If a terminal error occurs, 𝗥𝗲𝘀𝘁𝗮𝘁𝗲 𝗴𝘂𝗮𝗿𝗮𝗻𝘁𝗲𝗲𝘀 𝗮𝗹𝗹 𝗰𝗼𝗺𝗽𝗲𝗻𝘀𝗮𝘁𝗶𝗼𝗻𝘀 𝗲𝘅𝗲𝗰𝘂𝘁𝗲

Serverless is great, until you need to manage state. This is where Restate comes in! Learn how to combine Knative with Restate to build serverless applications on Kubernetes that are not only scalable, but also stateful and resilient. 🧵

View full thread

Serverless is great, until you need to manage state. This is where Restate comes in! Learn how to combine Knative with Restate to build serverless applications on Kubernetes that are not only scalable, but also stateful and resilient. 🧵

View full thread

🔥Restate and Knative are a great match! - Knative gives you the scale-to-zero magic of serverless, on your own infra. - Restate lets you write state machines, orchestration logic, and sagas and run them resiliently on any serverless platform. No external orchestrators needed.

In the post, we show: ✅ How to run your Restate services on Knative ✅ How you can use Restate’s embedded K/V store ✅ An example of a signup flow in Go

Serverless is great, until you need to manage state. This is where Restate comes in! Learn how to combine Knative with Restate to build serverless applications on Kubernetes that are not only scalable, but also stateful and resilient. 🧵

Serverless platforms don’t work well for stateful orchestration logic. You are forced to push state out to a database or glue everything with queues and retries. That’s brittle, complex, and hard to debug.

We recently did a community meeting on Restate 1.3 and its new features. Have a look to learn more about how the Restate SDKs work, and how you can use them to implement resilient concurrent tasks. 👇 YouTube link below

The community meeting of Thursday May 8 will include demos of all the new features of Restate 1.3! Join to learn how the SDKs work under-the-hood and how to implement resilient concurrent applications with Restate.

There is a new Restate 1.3.1 release 🎉 It contains various fixes and quality of life improvements: * Improved Admin API OpenAPI spec * Improved file corruption protection * Protection against accidentally starting Restate from a wrong working dir Get it while it's 🔥 github.com/restatedev/r...

github.comRelease v1.3.1 · restatedev/restateRestate v1.3.1 Release highlights 🌟 Improved Admin API OpenAPI specification for better integrations with external tools. Use latest RocksDB table format which provides stricter checksum validatio...