All pull requests

Pull requests

• [Core] Fix EvictLineage evicting all lineage due to negative byte accounting

  #67271·dragongu opened Oct 22, 2026Contributor·3 tasks done·

  ·
  3

• [Data] Distribute Parquet and ORC full schema planning

  #67269·jiangxt2 opened Oct 22, 2026Contributor·

  ·
  1

• [Data] Add Hive table projection, filter and limit pushdown

  #67268·jiangxt2 opened Oct 22, 2026Contributor·

  ·
  1

• [Data] Proactively scale down actor pools for backpressured operators to reduce resource waste

  community-contributiondata
  #67267·dragongu opened Oct 22, 2026Contributor·

  ·
  2

• [Data] Fix unifying a fixed-shape tensor array with a lower-ndim variable-shaped one

  community-contributiondata
  #67266·rayo5o opened Oct 22, 2026·

  ·
  1

  3

• [Core/Auth] Add RAY_AUTH_SKIP_METRICS to serve Ray's metrics endpoints without the auth token

  corego
  #67264·MortalHappiness opened Oct 22, 2026Member·

  ·
  1

  

• [Data] Fix write_lance compatibility with PyLance 6+ (#65129)

  community-contributiondata
  #67263·Pushkar-Shahi opened Oct 22, 2026·

  ·
  1

  4

• [release] Move data release tests off the ray-ml image to ray-torch (2/N)

  #67262·elliot-barn opened Oct 22, 2026Collaborator·

  ·

• [Train] Replace fan_out_to_workers with collect_results

  #67259·liulehui opened Oct 21, 2026Contributor·

  ·

• [Core/Auth] Create the auth token for ray.init() and Cluster() heads, and require one for ray start --head

  corego
  #67258·MortalHappiness opened Oct 21, 2026Member·

  ·
  1

  13

  

• [doc] Batch of small docs fixes

  coredocsgo
  #67257·dstrodtman opened Oct 21, 2026Contributor·

  ·
  12

  3

• [dashboard] Report GPU SM occupancy, tensor and DRAM activity through NVML GPM

  core
  #67253·prabhakar-dessai opened Oct 21, 2026·

  ·
  3

• [serve] Deflake test_ray_client on Windows

  goserve
  #67249·johntaylor-cell opened Oct 21, 2026Contributor·

  ·
  5

  

• [dashboard] Report GPU SM activity through NVML GPM

  core
  #67248·prabhakar-dessai opened Oct 21, 2026·

  ·
  7

• [dashboard] Add GPU busy metrics averaged over the whole poll window

  coredashboard
  #67247·prabhakar-dessai opened Oct 21, 2026·

  ·
  6

• [core][observability] Let the reporter agent record counters

  core
  #67246·prabhakar-dessai opened Oct 21, 2026·

  ·
  8

• [dashboard] Fix GPU utilization reporting in the dashboard agent

  coredocs
  #67245·prabhakar-dessai opened Oct 21, 2026·

  ·
  8

• [Data] Replace O(N) list.pop(0) with deque.popleft() in PushBasedShuffleTaskScheduler

  community-contributiondata
  #67243·deep1blue opened Oct 21, 2026·2 tasks done·

  ·
  4

• [Core] Fix issue #64692 by honoring client_processed_up_to also in the unordered actor task queue.

  community-contributioncore
  #67242·Pengkun-ZHU opened Oct 21, 2026·

  ·
  1

  4

• [serve][llm] Add opt-in GenAI workflow tracing

  #67241·aureleoules opened Oct 21, 2026·

  ·

• [data] Convert the write_sql and write_lance docstring examples to stock doctests

  datadocstring-onlygo
  #67238·dstrodtman opened Oct 21, 2026Contributor·

  ·
  5

• [ci] Skip polars docstring examples by name, apply Data doctest settings in the examples runner, and gate ml: data integration tests

  datadevproddocsgo
  #67237·dstrodtman opened Oct 21, 2026Contributor·

  ·

• fix:null rows in tensor columns

  community-contributiondata
  #67236·moiz-naeem opened Oct 21, 2026Contributor·

  ·
  1

  7

• [Core][Client] Add run_function_as_driver Ray Client prototype

  #67235·ryanaoleary opened Oct 21, 2026Contributor·6 of 8 tasks·

  ·
  1

• [serve] Skip the GCS availability-zone lookup for handles under Ray C…

  community-contributionserve
  #67233·da-daken opened Oct 21, 2026·

  ·