Yandex Cloud
Search
Contact UsGet started
  • Blog
  • Pricing
  • Documentation
  • All Services
  • System Status
    • Featured
    • Infrastructure & Network
    • Data Platform
    • Containers
    • Developer tools
    • Serverless
    • Security
    • Monitoring & Resources
    • ML & AI
    • Business tools
  • All Solutions
    • By industry
    • By use case
    • Economics and Pricing
    • Security
    • Technical Support
    • Customer Stories
    • Cloud credits to scale your IT product
    • Gateway to Russia
    • Cloud for Startups
    • Education and Science
    • Yandex Cloud Partner program
  • Blog
  • Pricing
  • Documentation
© 2025 Direct Cursus Technology L.L.C.
Yandex Serverless Containers
  • Comparison with other Yandex Cloud services
    • Overview
    • Container
    • Invoking a container
    • Runtime environment
    • Networking
    • Mounting external resources to a container file system
    • Long-lived containers
    • Container termination notifications
    • Dead Letter Queue
    • Container logs
    • Backups
    • Quotas and limits
  • Access management
  • Pricing policy
  • Terraform reference
  • Audit Trails events
  • Release notes
  • FAQ

In this article:

  • Supported languages and images
  • Container response
  • Environment variables
  • File system
  • Container instance resources
  • CPU
  • RAM
  • Ratio between RAM and vCPU
  • Metadata service
  1. Concepts
  2. Runtime environment

Runtime environment

Written by
Yandex Cloud
Updated at January 22, 2025
  • Supported languages and images
  • Container response
  • Environment variables
  • File system
  • Container instance resources
    • CPU
    • RAM
    • Ratio between RAM and vCPU
  • Metadata service

The application is run in a virtual environment.

Runtime environment characteristics:

  • Operating system: Ubuntu 22.04 LTS
  • Linux kernel: 5.15
  • Architecture: AMD64

Supported languages and imagesSupported languages and images

You can write your application in any programming language and have it use any image that meets the following requirements:

  • The executables are compiled under Linux x86_64.
  • Available image formats: Docker Image Manifest V 2, Schema 2 and OCI Image Format.

Container responseContainer response

The process which starts when invoking a container depends on the container operation mode.

Container invocation ends in one of the following cases:

  • Serverless Containers processed an HTTP request within the specified timeout and returned an HTTP response consistent with the container operation mode. The timeout includes the container instance launch time.
  • Serverless Containers failed to process an HTTP request within the specified timeout and returned a 504 error.
  • There is an infrastructure or initialization error with details provided in the response body.

Environment variablesEnvironment variables

The table provides a list of environment variables defined in the Serverless Containers runtime and available to an application. You cannot override these.

Key Value
PORT Network port the runtime uses to send HTTP requests to an application.
REQUEST_PATH Data on an HTTP request to a container if the container operation mode is Commands.

You can add other environment variables when you create a container revision or build a Docker image. The limit for maximum environment variable size, including variable names, is 4 KB.

You cannot calculate environment variables. Environment variable values are string constants. You can only calculate these within application code.

You can retrieve environment variables using standard programming language tools.

File systemFile system

The container image files and directories are readable on the virtual environment file system. The file system is writable. The space available for writing is limited. For more detail, please see Quotas and limits.

The file system is stored in RAM. Data written to the file system is not saved when a container instance is stopped.

Container instance resourcesContainer instance resources

CPUCPU

A CPU resource is allocated whenever a container instance is processing at least one request. Currently, no more than 4 vCPUs can be allocated.

RAMRAM

Each container instance is allocated RAM. The default value is 128 MB.

For example, RAM is consumed for:

  • File system writes
  • Container process launches
  • Request processing

Ratio between RAM and vCPURatio between RAM and vCPU

The number of available cores depends on the specified RAM size.

RAM CPU
128 MB to less than 4 GB 1 vCPU with a guaranteed vCPU share up to 100%
4 GB to less than 6 GB 2 vCPU with a guaranteed vCPU share up to 100%
6 GB to less than 8 GB 3 vCPU with a guaranteed vCPU share up to 100%
8 GB 4 vCPU with a guaranteed vCPU share up to 100%

Metadata serviceMetadata service

The metadata service is available inside the container's virtual runtime environment. Using it, you can get an IAM token for a service account if you specified the service account when creating the container revision.

Was the article helpful?

Previous
Invoking a container
Next
Networking
© 2025 Direct Cursus Technology L.L.C.