All pull requests

Pull requests

• fix(spark): render decimal, date and timestamp values in CDC images

  #20305·codope opened Oct 13, 2026Member·3 tasks done·

  ·
  1

  3

• fix(spark): honor 'earliest' begin instant in DataFrame CDC reads

  #20304·codope opened Oct 13, 2026Member·3 tasks done·

  ·
  2

  2

• fix(kafka-connect): close AdminClient after fetching topic partitions

  #20303·fhan688 opened Oct 13, 2026Contributor·3 tasks done·

  ·
  1

• feat(agent): Add first executable mutable Flink SQL MOR sink path to Hudi Architect

  #20302·fhan688 opened Oct 13, 2026Contributor·2 of 3 tasks·

  ·

• fix(flink): treat missing column stats as unknown after index expansion

  size:S
  #20298·csarushan1729 opened Oct 13, 2026·3 tasks done·

  ·
  1

  4

• fix(flink): retain files with missing column statistics

  size:M
  #20295·cshuo opened Oct 13, 2026Collaborator·2 of 3 tasks·

  ·
  1

  10

• fix(flink): preserve timestamp precision in predicates and indexes

  size:L
  #20294·cshuo opened Oct 13, 2026Collaborator·2 of 3 tasks·

  ·
  1

  5

• fix(flink): align bucket pruning with compound key placeholders

  size:M
  #20292·cshuo opened Oct 13, 2026Collaborator·2 of 3 tasks·

  ·
  1

  3

• [UBER] Add async pre-write rollback support

  size:M
  #20290·vamshipasunuru1 opened Oct 12, 2026Contributor·

  ·
  11

• fix(flink): preserve event-time delete ordering during precombine

  size:M
  #20288·cshuo opened Oct 12, 2026Collaborator·3 tasks done·

  ·
  1

  8

• Add Hadoop configuration directory hadoop_conf.dir support for Flink cross-cluster Hudi read/write

  size:L
  #20285·cbg-wx opened Oct 11, 2026·3 tasks·

  ·
  6

• fix(common): map every meta-fields mode but NONE to the legacy populated boolean

  size:S
  #20282·vamshipasunuru1 opened Oct 10, 2026Contributor·3 tasks done·

  ·
  8

• fix(core): write the key generator class for tables below version 8

  size:L
  #20276·yihua opened Oct 9, 2026Contributor·3 tasks done·

  ·
  1

  3

• fix(core): stop persisting hoodie.database.name in the table config

  size:L
  #20274·yihua opened Oct 9, 2026Contributor·3 tasks done·

  ·
  1

• fix(common): correct incremental fallback reads when a stale non-write instant precedes the first active commit

  size:L
  #20273·ad1happy2go opened Oct 9, 2026Collaborator·3 tasks done·

  ·
  1

  5

• fix(spark): list catalog partitions from the resolved catalog entry and fall back when the catalog is stale

  size:L
  #20272·yihua opened Oct 9, 2026Contributor·3 tasks done·

  ·
  1

• chore(deps): bump org.apache.flink:flink-table-planner_2.12 from 1.20.1 to 1.20.4 in /hudi-agent-gateway/fixtures/hudi-architect-flink-1.20

  dependenciesjavasize:XS
  #20270·dependabot[bot] opened Oct 9, 2026Bot·

  ·
  1

• perf(spark): stop deep schema comparisons per record in partial-update merging

  size:M
  #20269·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• fix(spark): keep bootstrap skeleton files and partitions as data skipping candidates

  size:S
  #20267·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• perf(timeline-service): check view freshness without the view lock

  size:L
  #20265·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• perf(index): reuse driver-resolved base files in the bloom key check

  size:L
  #20263·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• fix(writer): do not drop MOR records when the operation field is enabled

  size:M
  #20261·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• perf(metadata): read each metadata table HFile once per reader

  size:L
  #20259·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• fix(spark): use an expression index only for filters on its own function

  size:M
  #20256·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1

• perf(spark): build the prepped-write schema once per partition instead of per record

  size:S
  #20254·yihua opened Oct 9, 2026Contributor·3 tasks·

  ·
  1

  1