v3.0.0

What's New

BYOC Index Support

You can now manage Bring Your Own Cloud (BYOC) indexes using the pinecone_index resource. BYOC indexes are deployed into your own cloud environment via a Pinecone-managed environment identifier.

resource "pinecone_index" "byoc" {
  name      = "my-byoc-index"
  dimension = 1536
  spec = {
    byoc = {
      environment = "my-byoc-env-id"
    }
  }
}

Read Capacity Configuration

Serverless and BYOC indexes now support configurable read capacity via a read_capacity block inside the spec.serverless or spec.byoc block. Choose between on_demand (the default) or dedicated mode with fixed compute provisioning.

spec = {
  serverless = {
    cloud  = "aws"
    region = "us-east-1"
    read_capacity = {
      dedicated = {
        node_type = "b1"
        replicas  = 1
        shards    = 1
      }
    }
  }
}

Note: To switch from dedicated back to on_demand, explicitly set on_demand = {}. Removing the read_capacity block entirely will not change the mode already recorded in state.

Metadata Schema

Serverless, BYOC, and integrated indexes now support a schema block for controlling which metadata fields are indexed for filtering. By default all metadata is indexed; when schema is present, only fields listed with filterable = true are indexed.

Note: schema can only be configured at index creation time — changing it requires replacing the index.

spec = {
  serverless = {
    cloud  = "aws"
    region = "us-east-1"
    schema = {
      fields = {
        "category" = { filterable = true }
        "language"  = { filterable = true }
      }
    }
  }
}

Other Changes

Dependency updates (#71)

Changelog

• 9858479 Update README and supporting documentation for version bump (#72)
• e878eb5 Bump dependencies per Dependabot PRs #48, #54, #55, #61, #63, #64, #65, #66 (#71)
• 3421473 Implement MetadataSchema support for serverless, integrated, and BYOC indexes (#70)
• 3487fc7 Implement BYOC indexes and read_capacity support (#69)