Skip to main content

Security model - Temporal Cloud

What kind of security does Temporal Cloud provide?

The security model of Temporal Cloud encompasses applications, data, and the Temporal Cloud platform itself.

General platform security

For information about the general security features of Temporal, see our Platform security page.

Application and data​

What is the security model for applications and data in Temporal Cloud?

Code execution boundaries​

Temporal Cloud provides the capabilities of Temporal Server as a managed service; it does not manage your applications or Workers. Applications and services written using Temporal SDKs run in your computing environment, such as containers (Docker, Kubernetes) or virtual machines (in any hosting environment). You have full control over how you secure your applications and services.

Data Converter: Client-side encryption​

The optional Data Conversion capability of the Temporal Platform lets you transparently encrypt data before it's sent to Temporal Cloud and decrypt it when it comes out.

Data Conversion runs on your Temporal Workers and Clients; Temporal Cloud cannot see or decrypt your data. If you use this feature, data stored in Temporal Cloud remains encrypted even if the service itself is compromised.

By deploying a Codec Server you can securely decrypt data in the Temporal Web UI without sharing encryption keys with Temporal.

The platform​

What is the security model for the Temporal Cloud platform?

Namespace isolation​

The base unit of isolation in a Temporal environment is a Namespace. Each Temporal Cloud account can have multiple Namespaces. A Namespace (regardless of account) cannot interact with other Namespaces. Each Namespace is available through a secure gRPC (mTLS) endpoint and an HTTPS (TLS) endpoint. You can make these endpoints more secure by routing all communication through AWS PrivateLink.

Temporal Cloud is a multi-tenant service. Namespaces in the same environment are logically segregated. Namespaces do not share data processing or data storage across regional boundaries.

AWS PrivateLink allows you to establish a private connection between your Amazon Virtual Private Cloud (VPC) and Temporal Cloud, without routing traffic over the internet. This ensures that your communication with Temporal Cloud remains within the AWS network, providing an extra layer of security and privacy. It's important to note that PrivateLink enables a one-way connection from your VPC to Temporal Cloud. Temporal Cloud does not gain peering access to your VPC or any resources within it.

note

If you are interested in leveraging AWS PrivateLink in your Namespaces, create a support ticket that includes the following information:

  • AWS Region: The Region in which your connection will go through.
  • AWS Account Id: The account which contains the permissions to enable AWS PrivateLink.
  • Temporal Cloud Namespace names: The name of the Namespaces you want to enable AWS PrivateLink with.

To set up PrivateLink connectivity with Temporal Cloud, you'll need to perform the following high-level steps:

  1. Add a VPC Endpoint Interface in your AWS VPC.
  2. Configure security groups to allow TCP egress traffic to port 7233 (for gRPC communication).

Once set up, you can test your PrivateLink connectivity using:

  1. Non-Temporal tools like cURL (useful for testing from environments that restrict tool usage).
  2. Temporal CLI: by setting the --tls-server-name parameter to the PrivateLink endpoint.
  3. Temporal SDKs: by setting the server name argument to the PrivateLink endpoint.

Encryption​

Communication into and out of Namespaces is over TLS. All communication within our production environments is over TLS 1.3. Data is stored in two separate locations: an Elasticsearch instance (used when filtering Workflows in SDK clients, the CLI, or the Web UI) and the core Temporal Cloud persistence layer. Both are encrypted at rest with AES-256-GCM.

For more information, see Requirements for CA certificates in Temporal Cloud.

Identity​

Authentication to gRPC endpoints is provided by mTLS per Namespace.

For more information, see How to manage SAML authentication with Temporal Cloud.

Access​

Authorization is managed at the account and Namespace level. Users and systems are assigned one or more preconfigured roles. Users hold account-level Roles of administrators, developers, and read-only users. Systems and applications processes hold their own distinct roles.

Monitoring​

In addition to extensive system monitoring for operational and availability requirements, we collect and monitor audit logs from the AWS environment and all calls to the gRPC API (which is used by the SDKs, CLI, and Web UI). These audit logs can be made available for ingestion into your security monitoring system.

Testing​

We contract with a third party to perform a full-scope pentest (with the exception of social engineering) annually. Additionally, we perform targeted third-party and internal testing on an as-needed basis, such as when a significant feature is being released.

Internal Temporal access​

We restrict access to production systems to the small team of employees who maintain our production infrastructure. We log all access to production systems; shared accounts are not allowed. Access to all production systems is through SSO, with MFA enabled.

Access to AWS is granted only for limited periods of time, with a maximum of 8 hours. (For more information, see the blog post Rolling out access hours at Temporal.)

All Temporal engineering systems are secured by GitHub credentials, which require both membership in the Temporal GitHub organization and MFA. Access grants are reviewed quarterly.

Compliance​

Temporal Technologies is SOC 2 Type 2 certified and compliant with GDPR and HIPAA regulations. Compliance audits are available by request through our Contact page.