March 2024 Summaries

Temporal's March newsletter highlights several updates and opportunities within their community, including the release of server version 1.23.0 with new features and depreciations, as well as the introduction of a Terraform provider and Workflow History Export in public preview. The newsletter details updates for PHP, Go, and Java SDKs, emphasizing the importance of reading release notes due to some breaking changes. It also showcases how companies are leveraging new automation features in Temporal Cloud for operational efficiency. Upcoming events include participation in QCon London, DevNexus, and the Great International Developer Summit, where attendees can meet the Temporal team. The newsletter promotes a series of on-demand webinars about building with Temporal Cloud, and introduces new training courses on workflow versioning across various SDKs, available for free. Additionally, the newsletter highlights community insights, including FireHydrant's testing of Temporal Signals and the advantages of Temporal Cloud in cloud disaster recovery and custom persistence layers.

Four new hands-on training courses have been released covering Versioning using Go, Java, Python, and TypeScript SDKs. These courses are designed for Temporal application developers to learn about the importance of versioning in application development and deployment. The courses cover when and why versioning is necessary, two techniques for versioning in Temporal, and a comparison of these techniques. Completion of one of these training courses will help developers understand how to apply versioning using their preferred SDK.

The text discusses the challenges of maintaining high availability when self-hosting Temporal at scale, and how Temporal Cloud provides high availability. Self-hosting Temporal requires expertise to maintain multiple independently scalable components such as a database, four independent services that make up the Temporal Server, and dealing with inevitable failures in distributed systems. While many developers successfully self-host Temporal, they may struggle to meet high availability SLAs and spend significant time and resources operating it. Temporal Cloud offers Temporal-as-a-service, properly tuning supporting database and services for the customer's load and ensuring their high availability. It provides fault tolerance by default, a 99.99% service level objective (SLO), and a 99.9% service level agreement (Contractual SLA). Disaster recovery features include zero RTO/RPO for availability zone failures and eight-hour maximum RTO/RPO for region failures. Temporal Cloud also offers Multi-Region Namespaces, which are currently in pre-release, providing failover capabilities to mitigate service outages due to regional failures and extending the contractual SLA to 99.99%.

Temporal can significantly benefit AI workflows due to its inherent capabilities around durability, scale, and failure handling. Its workflow orchestration and state management features are particularly useful for complex, long-running processes often found in AI applications. Key areas where Temporal can be helpful include workflow orchestration for AI pipelines, scalable and reliable machine learning model training, distributed data processing, continuous learning and model deployment, experimentation and versioning, efficient use of GPUs, scaling AI operations, event-driven and asynchronous execution, and observability and debugging. Getting started with Temporal involves diving into the getting started guide, experimenting with sample projects, and utilizing resources like documentation and community support.

Temporal Cloud Automation capabilities are revolutionizing how companies manage their cloud operations, enabling automation for tasks such as user management and namespace provisioning/deprovisioning. The newly added features include secure access via API Keys, command-line automation with tcld, and efficient infrastructure-as-code management through the Terraform provider. These tools not only enhance security but also provide operational efficiency gains. Use cases for Temporal Cloud Automation include centralized cloud operations, certificate rotation, user and access management, and team onboarding.

Temporal Cloud, a managed service for Temporal, offers better performance and lower latency compared to self-hosted clusters due to effective management and scaling by its team. The custom persistence layer of Temporal Cloud helps handle high throughput with more stable request latencies. This architecture is designed to support multiple tenants sharing the same compute and persistence layer, ensuring extra capacity for all customers during traffic spikes while addressing the noisy neighbor problem. Key features of this custom persistence solution include better sharding, a write-ahead log, and tiered storage of Workflow Event History. These enhancements result in higher scalability and lower latency for Temporal Cloud users.

The Temporal Cloud for Startups program aims to assist early-stage companies in achieving market success more quickly without depleting their resources. In collaboration with AWS's startup program, an exclusive offer of $3,000 USD in free Temporal Cloud credits is now available for AWS Activate customers. This partnership seeks to expand the reach of the Temporal Cloud program and support startups in growing their businesses using AWS services.

In a single Workflow Definition, it is possible to include Activities written in different languages. Each Activity requires a separate Worker and distinct task queue names for the Activities in each language. The Temporal SDK automatically converts language types for interoperability with other SDKs. To combine actions written in different languages, use distinct Activity Types that will be called during Workflow Execution by the language-specific Worker Entity. Examples of using multiple languages with Temporal include the Polyglot and Pendulum applications.

The Replay Conference is returning to Seattle for its third year, and Super Early Bird tickets are now available at a discounted price of $99 until March 6th. The conference will take place from September 18-20 at the Hyatt Regency Bellevue on Seattle's Eastside. Temporal Cloud's default Actions Per Second (APS) limit for all Namespaces has been raised to 400, and a new feature called Temporal Schedules is now available as a replacement for traditional Cron jobs for task scheduling. The .NET SDK is also being introduced with high-performance capabilities. Upcoming events include DevWorld in Amsterdam on February 29-March 1, VS Live! in Las Vegas from March 3-8, and Explore DDD Conference in Denver from March 12-15.

Temporal Cloud's Workflow History Export allows users to export closed workflow histories on a per-namespace basis to an AWS S3 cloud storage bucket. This feature is useful for long-term data storage, compliance, auditing, and further analysis. Currently in Public Preview, the cost associated with this feature involves each workflow exported accruing a single action, which will be shown as a separate line item per namespace on the invoice. To set up workflow export, users must create an AWS S3 bucket in the same region as the namespace and grant Temporal permission to write data via AWS CloudFormation. Once everything is set up correctly, Temporal will export history data from your namespace to S3 on an hourly basis.