Skip to content

[Filters] Add filters to DAG View – Runs tab #53046

@cmarteepants

Description

@cmarteepants

View

DAG View – Runs tab

Image

Filters to Add

  • Filter by Run After Range
    ↪ Helps find runs that were delayed or executed later than expected

  • Filter by Start Date
    ↪ Helps find runs that started during a specific window of time (e.g. following a system update)

  • Filter by End Date
    ↪ Useful to identify runs that ended recently or ended during a problematic period

  • Filter by Duration Range
    ↪ Identify long-running or stuck DAGs

  • Filter by DAG Version and/or Bundle Version
    ↪ Correlate failures or anomalies with recent code or deployment changes

  • Filter by Conf Key/Value
    ↪ Enables users to track down runs with specific parameters (e.g., a run triggered with a certain partner_id)

  • Filter by Logical Date Range
    ↪ Useful for identifying runs within a specific scheduled interval (e.g., a regression starting on June 1)

  • Filter by Triggered By (e.g. manual, CLI, sensor, asset, backfill)
    ↪ Quickly isolate runs from specific sources or trigger types

  • Filter by Consuming Asset
    ↪ Identify DAG runs triggered by a specific upstream asset event

  • Filter by Producing Asset
    ↪ Identify DAG runs producing a specific downstream asset event

API Support

⚠️ Partial

  • /dagRuns supports filtering by:

    • dag_id
    • state
    • run_type
    • execution_date (logical date)
  • Missing filter support for:

    • Conf content (e.g., key/value search)
    • Run-after range
    • Logical date range
    • DAG version or bundle version
    • Trigger source (manual, CLI, etc.)
    • Producing asset / Consuming asset

UI Impact

  • Filters could be placed above the DAG Run table (as dropdowns, date pickers, or input fields)
  • Filtering improves usability in high-throughput environments and large DAGs
  • Enables faster root cause analysis and run correlation

Notes

  • Conf filtering and trigger source filtering are especially useful for asset-aware and parameterized DAGs
  • Logical date and duration range are essential for tracking SLAs and performance regressions

Metadata

Metadata

Labels

area:UIRelated to UI/UX. For Frontend Developers.good first issuekind:featureFeature Requestskind:metaHigh-level information important to the community

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions