v0.56.0

parruda/swarmv0.56.0Jun 3, 2026by github-actions[bot]

AI Summary

Adds a new Databricks sink, delay transform, and fixes a significant CPU regression introduced in version 0.50.0.

Key Highlights

  • New `databricks_zerobus` sink supporting OAuth 2.0 and Unity Catalog.
  • New `delay` transform for event delay based on fixed duration or condition.
  • HTTP sinks now support `retry_strategy` configuration.
  • Fixed CPU regression affecting metric normalization sinks.

Breaking Changes

  • The `greptimedb_metrics` and `greptimedb_logs` sinks now require GreptimeDB v1.x.

New Features

  • Databricks ZeroBus sink
  • Delay transform
  • HTTP retry_strategy configuration
  • zstd compression support
  • Parquet batch encoding in AWS S3
  • CPU regression fix

Full Release Notes

The [COSE team](https://opensource.datadoghq.com/about/#the-community-open-source-engineering-team) is excited to announce version `0.56.0`! šŸš€ 

## Release highlights

- Added a new `databricks_zerobus` sink that streams log data to Databricks Unity Catalog tables through the Zerobus ingestion service. The sink supports OAuth 2.0 authentication, automatic schema fetching from Unity Catalog, and protobuf batch encoding.
- Added a new `delay` transform that delays each event by a fixed duration. Events can also
  be delayed based on a condition which includes VRL transforms.
- HTTP-based sinks that use the shared retry helpers now support a `retry_strategy` configuration
option to control which HTTP response codes are retried. The `http` sink also includes a new
example showing how to retry only specific transient status codes.
- The `vector` sink now supports `zstd` compression in addition to `gzip`. This provides better
 compression ratios and performance for Vector-to-Vector communication.
- The `tag_cardinality_limit` transform received major enhancements: per-tag cardinality
  overrides (`per_tag_limits`), per-metric tracking isolation (`tracking_scope: per_metric`),
  a global key cap (`max_tracked_keys`), and the ability to opt entire metrics out of
  cardinality tracking.
- Parquet batch encoding in the `aws_s3` sink is now available out of the box in official release
  binaries for all users.
- Fixed a CPU regression introduced in 0.50.0 affecting all sinks that use metric normalization such as `prometheus_remote_write`, `aws_cloudwatch_metrics`, `statsd`, and others.
- Restored support for installing Vector on RHEL 8, Rocky Linux 8, AlmaLinux 8, and CentOS
  Stream 8, which had been broken since 0.55.0 due to an inadvertent glibc requirement bump.
- Unit tests now support an optional `expected_event_count` field on test outputs, allowing assertions on the number of events emitted by a transform.

## Breaking Changes

- The `greptimedb_metrics` and `greptimedb_logs` sinks now require GreptimeDB v1.x. Users running GreptimeDB v0.x must upgrade their GreptimeDB instance before upgrading Vector.

šŸ‘‰ View all new features and fixes in the release notes [here](https://vector.dev/releases/0.56.0).