Product Updates
Back to All
Added

Snowflake & Databricks — export Lytics platform metrics type: added

6 days ago

The Snowflake and Databricks integrations now include an Export Metrics workflow that lands raw time-series metric data — segment sizes, stream volumes, job activity, and the rest of the /v2/metric catalog — directly in your warehouse for analysis and reporting. Same controls as the existing BigQuery metric export, now available against two more destinations.

How it works

  • Pick a Metric Dimension (e.g. segment, stream, works), optionally narrow to a single Dimension ID, and choose a Metric Type scoped to that dimension. The full dimension/type matrix is shared across all three warehouse exports.
  • Choose a Lookback Range for the initial backfill — 0d, 7d, 14d, 30d (default), 60d, or 90d.
  • Toggle Keep Updated to run the export daily. Each run is append-only: only new rows since the last successful export are loaded, and existing rows are never modified or dropped.
  • Snowflake: reuses the Snowflake Direct Authorization for Bulk Export (with GCS storage integration). Stages gzipped NDJSON to GCS, then COPY INTO your table with VARCHAR(16777216) / TIMESTAMP_TZ(9) / FLOAT columns. Default table LYTICS_{AID}_METRICS.
  • Databricks: uses the standard Databricks Database Authorization. Stages gzipped CSV to a Lytics-managed S3 bucket, then COPY INTO a Delta table with STRING / TIMESTAMP / DOUBLE columns. Default table lytics_{AID}_metrics.

See Snowflake → Export Metrics and Databricks → Export Metrics for the full configuration walkthroughs.