Changelog

September 18, 2026Reliability
Improved data-type reliability
Fixes preserve floating-point precision, decode Kafka Connect decimals, improve MongoDB Extended JSON handling, and correct Snowpipe Streaming endpoints.
Read more
September 2, 2026Reliability
PostgreSQL 17 failover replication slots
New CDC pipelines can use PostgreSQL 17 failover replication slots, configurable in the Dashboard or Terraform.
Read more
July 26, 2026Reliability
Automatic schema reconciliation after redeploy
Pipelines with automatic schema changes enabled reconcile destination schemas after a non-first redeploy, with a two-hour per-pipeline cooldown.
Read more
July 19, 2026Reliability
MongoDB updates with large fields
Redundant updateDescription metadata is removed from MongoDB change-stream events, lowering the risk that large updates exceed BSON size limits.
Read more
April 2, 2026Reliability
MongoDB snapshot retries on network errors
MongoDB snapshot retries cover network-level failures and connection pool cleared errors, in addition to server-side cursor errors.
Read more
April 2, 2026Reliability
Resumable DynamoDB snapshots
DynamoDB snapshot config accepts a startAfterKey, an S3 key to resume from, so interrupted snapshots continue without reprocessing completed files.
Read more
April 2, 2026Reliability
Webhook failure alerts and dashboard visibility
The dashboard shows failing webhooks and how long they have been failing, and Artie emails alert recipients on failure and on recovery.
Read more
April 1, 2026Reliability
Auto-skip primary key constraints when using primaryKeysOverride
Tables with primaryKeysOverride configured automatically omit the PRIMARY KEY constraint from destination DDL, and the override now reaches Oracle pipelines.
Read more
January 12, 2026Reliability
Table-Level Monitoring
Monitors can be configured per table within a pipeline, with anomaly detection and threshold alerts on row count and data volume.
Read more
January 7, 2026Reliability
Improved Postgres Replication Slot Monitoring
The validation job monitors every replication slot on a Postgres instance and alerts when a third-party slot is inactive, lost, or retaining too much WAL.
Read more