architecture & infrastructure

10 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

Temporal Cloud does not manage applications or workers

Temporal Cloud provides the capabilities of Self-Hosted Temporal as a managed service but does not manage your applications or workers. Applications and services written using Temporal SDKs still run in your compute environment, and you have full control over how you secure your applications and services.

AWS PrivateLink for Temporal Cloud

Temporal Cloud supports AWS PrivateLink as a private connectivity option. If your infrastructure is in AWS, configure a PrivateLink endpoint for Temporal Cloud to allow workers and applications to reach Temporal Cloud over a private network path, avoiding traversal of the public internet.

Google Cloud Private Service Connect for Temporal Cloud

Temporal Cloud supports Google Cloud Private Service Connect (PSC) as a private connectivity option. If your infrastructure is in GCP, configure a PSC endpoint for Temporal Cloud to allow workers and applications to reach Temporal Cloud over a private network path, avoiding traversal of the public internet.

Temporal Cloud default SLA by Namespace scope

Single-Region Namespaces provide 99.9% uptime SLA with ≤ 8 hours RTO and ≤ 8 hours RPO. Same-Region Replication, Multi-Region Replication, and Multi-Cloud Replication all provide 99.99% uptime SLA with ≤ 20 minutes RTO and near-zero (≈ seconds) RPO.

High Availability features cost impact

Enabling High Availability (HA) features for namespaces doubles the consumption cost and provides a 99.99% contractual SLA guarantee against service errors.

Context propagation purpose and mechanism

Context propagation lets you pass custom key-value data from a Client to Workflows, and from Workflows to Activities and Child Workflows, without threading values through every function signature. Each SDK provides a context propagator interface you implement to control which values are injected and extracted. You register propagators on the Client, and the SDK calls them automatically at every boundary.

Common context propagation use cases

Common use cases for context propagation include propagating distributed tracing IDs such as OpenTelemetry trace context, passing tenant IDs for multi-tenant applications, and forwarding auth tokens or request-scoped metadata.

Default payload size limit

The Temporal Service enforces a maximum per-payload size. The default and recommended limit is 2 MB. This limit is fixed at 2 MB on Temporal Cloud. Self-hosted users can configure this limit, but it cannot exceed 2 MB on Cloud.

Multi-Cluster Replication: purpose and basics

Multi-Cluster Replication is a feature that asynchronously replicates Workflow Executions from active Clusters to other passive Clusters for backup and state reconstruction. When necessary for higher availability, Cluster operators can failover to any of the backup Clusters. The feature is considered experimental and not subject to normal versioning and support policy.

dcRedirectionPolicy enables request forwarding to active cluster

When a client connects to a standby Cluster, its requests can be forwarded to the active Cluster. This capability is off by default and is enabled by setting the dcRedirectionPolicy configuration parameter.

Give your agent this brain