AWS S3

Deep dive into Parseable and AWS S3 setup

Prerequisites

  • New Parseable cluster requires a fully empty S3 bucket.

  • For optimum performance, please ensure the S3 bucket is in the same region as the compute (EC2, EKS, ECS, Lightsail or others) instances.

Authentication

Parseable supports authentication to S3 via

  • Access Key and Secret Key: The AWS access key and the secret key can be used to authenticate to AWS S3 bucket. To use this method, set the environment variables P_S3_ACCESS_KEY and P_S3_SECRET_KEY before starting the Parseable server(s).

  • Instance Metadata Service (IMDS): For Parseable instances running on EC2, AWS credentials can be sourced from the Instance Metadata Service (IMDS), avoiding the need for explicit P_S3_ACCESS_KEY and P_S3_SECRET_KEY. To use this method,

    • Ensure that Instance Metadata Service (IMDS) is enabled when creating the EC2 instance (under Advanced details section). Select the Metadata version to V1 and V2 (token optional). Please refer to the metadata service docs for more. You set P_AWS_IMDSV1_FALLBACK to true if you want to use the V1 method.

    • By default, Parseable uses the standard AWS metadata endpoint. If you’re using a custom metadata endpoint, add the optional environment variable P_AWS_METADATA_ENDPOINT to specify a custom endpoint URL for retrieving instance metadata.

  • IAM Roles for Service Accounts (IRSA): This is useful for EKS deployments. With IRSA you can associate an IAM role with a Kubernetes service account and configure Parseable to use the service account. To use this method, refer to Parseable EKS documentation.

Updated on