Skip to content

Update changelog for the v26.06.1 release [skip ci] #15300

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
NvTimLiu merged 1 commit into from
Jul 17, 2026
Merged

Update changelog for the v26.06.1 release [skip ci] #15300

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 14 additions & 1 deletion CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
# Change log
Generated on 2026-06-14
Generated on 2026-07-17

## Release 26.06

Expand All @@ -21,6 +21,7 @@ Generated on 2026-06-14
### Performance
|||
|:---|:---|
|[#15163](https://github.com/NVIDIA/spark-rapids/issues/15163)|Parquet Hadoop fallback readVectored uses 128 KiB copies and can regress remote scans|
|[#14283](https://github.com/NVIDIA/spark-rapids/issues/14283)|[FEA] Support join condition which has "cast to bigint"|
|[#14068](https://github.com/NVIDIA/spark-rapids/issues/14068)|[FEA] Iceberg planning overhead is larger than parquet planning.|
|[#14591](https://github.com/NVIDIA/spark-rapids/issues/14591)|[FEA] Support perf io in iceberg.|
Expand All @@ -30,6 +31,8 @@ Generated on 2026-06-14
### Bugs Fixed
|||
|:---|:---|
|[#15235](https://github.com/NVIDIA/spark-rapids/issues/15235)|[BUG] cudaErrorIllegalAddress while writing into the Delta table|
|[#15183](https://github.com/NVIDIA/spark-rapids/issues/15183)|[BUG] DBR 17.3 build fails after SessionCatalog listPartitions APIs added resolvedCatalogTable|
|[#14864](https://github.com/NVIDIA/spark-rapids/issues/14864)|[BUG] test_parquet_interleaved_file_splits_partition_value_alignment fails on Dataproc Serverless: os.walk cannot see GCS-backed spark_tmp_path|
|[#14981](https://github.com/NVIDIA/spark-rapids/issues/14981)|[BUG] DBR 14.3 CPU fallback MERGE with NOT MATCHED BY SOURCE can fail with GpuUnionExec|
|[#14986](https://github.com/NVIDIA/spark-rapids/issues/14986)|[BUG] Delta DELETE on Databricks 14.3 returns num_affected_rows = -1 for metadata-only (partition/whole-table) deletes|
Expand Down Expand Up @@ -75,6 +78,12 @@ Generated on 2026-06-14
### PRs
|||
|:---|:---|
|[#15299](https://github.com/NVIDIA/spark-rapids/pull/15299)|Update dependency version JNI to 26.06.1|
|[#15298](https://github.com/NVIDIA/spark-rapids/pull/15298)|[DOC] update download page for 26.06.1 release [skip ci]|
|[#15164](https://github.com/NVIDIA/spark-rapids/pull/15164)|Use configured copy buffer for Hadoop vectored reads|
|[#15228](https://github.com/NVIDIA/spark-rapids/pull/15228)|Update dependency version JNI to 26.06.1-SNAPSHOT|
|[#15211](https://github.com/NVIDIA/spark-rapids/pull/15211)|Fix DBR 17.3 build after SessionCatalog partition API change|
|[#15084](https://github.com/NVIDIA/spark-rapids/pull/15084)|Update changelog for the v26.06.0 release [skip ci]|
|[#15079](https://github.com/NVIDIA/spark-rapids/pull/15079)|Update dependency version private to 26.06.1|
|[#15024](https://github.com/NVIDIA/spark-rapids/pull/15024)|Fall back when Iceberg S3 PerfIO is unsupported|
|[#15066](https://github.com/NVIDIA/spark-rapids/pull/15066)|Use current repository name in GitHub workflows [skip ci]|
Expand Down Expand Up @@ -359,15 +368,19 @@ Generated on 2026-06-14
|[#14492](https://github.com/NVIDIA/spark-rapids/pull/14492)|Databricks run-tests: rsync test reports from scala2.13 for Spark 4.x|
|[#14464](https://github.com/NVIDIA/spark-rapids/pull/14464)|Fix oom of iceberg ci|
|[#14475](https://github.com/NVIDIA/spark-rapids/pull/14475)|Fix databricks build break caused by `NoIcebergProviderImpl`|
|[#14445](https://github.com/NVIDIA/spark-rapids/pull/14445)|Add runtime check for supported icberg version|
Comment thread
NvTimLiu marked this conversation as resolved.
Comment thread
NvTimLiu marked this conversation as resolved.
|[#14466](https://github.com/NVIDIA/spark-rapids/pull/14466)|Use long instead of int to track kudo buffer sizes and offsets in GpuShuffleCoalesce|
|[#14467](https://github.com/NVIDIA/spark-rapids/pull/14467)|Update doc for deletionVectors.predicatePushdown.enabled to note useMetadataRowIndex dependency [skip ci]|
|[#14468](https://github.com/NVIDIA/spark-rapids/pull/14468)|Fix compile break from BloomFilter.create deprecation|
|[#14460](https://github.com/NVIDIA/spark-rapids/pull/14460)|Enable the new DV-aware Delta reader by default|
|[#14398](https://github.com/NVIDIA/spark-rapids/pull/14398)|[AutoSparkUT] Fix CollectLimitExec GPU replacement for Command results (SPARK-19650)|
|[#14456](https://github.com/NVIDIA/spark-rapids/pull/14456)|Fix DB-17.3 integration tests in join_test.py|
|[#14455](https://github.com/NVIDIA/spark-rapids/pull/14455)|Fix DB-17.3 integration tests for JSON, Parquet, and ORC write|
|[#14435](https://github.com/NVIDIA/spark-rapids/pull/14435)|Add support for Coalesce reader for Delta Lake using cuDF deletion vector APIs|
|[#14270](https://github.com/NVIDIA/spark-rapids/pull/14270)|Add read support for nested data types in iceberg.|
|[#14446](https://github.com/NVIDIA/spark-rapids/pull/14446)|[AutoSparkUT] Fix cached table zero-column scan crash (issue #14098)|
|[#14400](https://github.com/NVIDIA/spark-rapids/pull/14400)|[AutoSparkUT] Re-enable 'normalize special floating numbers in subquery' test (issue #14116)|
|[#14293](https://github.com/NVIDIA/spark-rapids/pull/14293)|Close stream time gap in RowToColumnarIterator hasNext|
|[#14450](https://github.com/NVIDIA/spark-rapids/pull/14450)|Fix `_metadata` column incorrectly pruned when nested fields are referenced|
|[#14448](https://github.com/NVIDIA/spark-rapids/pull/14448)|check rapids shuffle configured instead of available before heartbeat init|
|[#14436](https://github.com/NVIDIA/spark-rapids/pull/14436)|[AutoSparkUT] Fix trunc/date_trunc to return null for invalid format strings|
Expand Down
Loading