| #8441 |
Memoizing DataGens in integration tests |
| #8516 |
Avoid calling Table.merge with BinaryType columns |
| #8515 |
Fix warning about deprecated parquet config |
| #8427 |
[Doc] address Spark RAPIDS NVAIE VDR issues [skip ci] |
| #8486 |
Move task completion listener registration to after variables are initialized |
| #8481 |
Removed spark.rapids.sql.castDecimalToString.enabled and enabled GPU decimal to string by default |
| #8485 |
Disable test_read_compressed_hive_text on CDH. |
| #8488 |
Adds note on multi-threaded shuffle targetting <= 200 partitions and on TCP keep-alive for UCX [skip ci] |
| #8414 |
Add support for computing remainder with Decimal128 operands with more precision on Spark 3.4 |
| #8433 |
Add regression test for regexp_replace hanging with some inputs |
| #8477 |
Fix input binding of grouping expressions for complete aggregations |
| #8464 |
Remove NOP Maven javadoc plugin definition |
| #8402 |
Bring back UCX 1.14 |
| #8470 |
Ensure the MT shuffle reader enables/disables with spark.rapids.shuff… |
| #8462 |
Fix compressed Hive text read on |
| #8458 |
Add check for negative id when creating new MR job id |
| #8437 |
Implement the bug fix for SPARK-41448 and shim it for Spark 3.2.4 and Spark 3.3.{2,3} |
| #8420 |
Fix reads for GZIP compressed Hive Text. |
| #8445 |
Document errors/warns in the logs during catalog shutdown [skip ci] |
| #8438 |
Revert "skip test_array_repeat_with_count_scalar for now (#8424)" |
| #8385 |
Reduce memory usage in GpuFileFormatDataWriter and GpuDynamicPartitionDataConcurrentWriter |
| #8304 |
Support combining small files for multi-threaded ORC reads |
| #8413 |
Stop double closing in json scan + skip test |
| #8430 |
Update docs for spark.rapids.filecache.checkStale default change [skip ci] |
| #8424 |
skip test_array_repeat_with_count_scalar to wait for fix #8409 |
| #8405 |
Change TimeAdd/Sub subquery tests to use min/max |
| #8408 |
Document conventional dist jar layout for single-shim deployments [skip ci] |
| #8394 |
Removed "peak device memory" metric |
| #8378 |
Use spillable batch with retry in GpuCachedDoublePassWindowIterator |
| #8392 |
Update IDEA dev instructions [skip ci] |
| #8387 |
Rename inconsinstent profiles in api_validation |
| #8374 |
Avoid processing empty batch in ParquetCachedBatchSerializer |
| #8386 |
Fix check to do positional indexing in ORC |
| #8360 |
use matrix to combine multiple jdk* jobs in maven-verify CI [skip ci] |
| #8371 |
Fix V1 column name match is case-sensitive when dropping partition by columns |
| #8368 |
Doc Update: Clarify both line anchors ^ and $ for regular expression compatibility [skip ci] |
| #8377 |
Avoid a possible race in test_empty_filter |
| #8354 |
[DOCS] Updating tools docs in spark-rapids [skip ci] |
| #8341 |
Enable CachedBatchWriterSuite.testCompressColBatch |
| #8264 |
Make tables spillable by default |
| #8364 |
Fix NullPointerException in ORC multithreaded reader where we access context that could be null |
| #8322 |
Avoid out of bounds on GpuInMemoryTableScan when reading no columns |
| #8342 |
Elimnate javac warnings |
| #8334 |
Add in support for filter on empty batch |
| #8355 |
Speed up github verify checks [skip ci] |
| #8356 |
Enable auto-merge from branch-23.06 to branch-23.08 [skip ci] |
| #8339 |
Fix withResource order in GpuGenerateExec |
| #8340 |
Stop calling contiguousSplit without splits from GpuSortExec |
| #8333 |
Fix GpuTimeAdd handling both input expressions being GpuScalar |
| #8302 |
Add support for DecimalType in Remainder for Spark 3.4 and DB 11.3 |
| #8325 |
Disable test_read_hive_fixed_length_char on Spark 3.4+. |
| #8327 |
Enable spark.sql.legacy.parquet.nanosAsLong for Spark 3.2.4 |
| #8328 |
Fix Hive text file write to deal with CUDF changes |
| #8309 |
Fix GpuTopN with offset for multiple batches |
| #8306 |
Update code to deal with new retry semantics |
| #8307 |
Full ordinal support in GetArrayItem |
| #8243 |
Enable retry for Parquet writes |
| #8295 |
Fix ORC reader for CHAR(N) columns written from Hive |
| #8298 |
Append new authorized user to blossom-ci whitelist [skip ci] |
| #8276 |
Fallback to CPU for enableDateTimeParsingFallback configuration |
| #8296 |
Fix Multithreaded Readers working with Unity Catalog on Databricks |
| #8273 |
Add support for escaped dot in character class in regexp parser |
| #8266 |
Add test to confirm correct behavior for decimal average in Spark 3.4 |
| #8291 |
Fix delta stats tracker conf |
| #8287 |
Fix Delta write stats if data schema is missing columns relative to table schema |
| #8286 |
Add Tencent cosn:// to default cloud schemes |
| #8283 |
Add split and retry support for filter |
| #8290 |
Pre-merge docker build stage to support containerd runtime [skip ci] |
| #8257 |
Support cuda12 jar's release [skip CI] |
| #8274 |
Add a unit test for reordered canonicalized expressions in BinaryComparison |
| #8265 |
Small code cleanup for pattern matching on Decimal type |
| #8255 |
Enable locals,patvars,privates unused Scalac checks |
| #8234 |
JDK17 build support in CI |
| #8256 |
Use env var with version files as fallback for IT DBR version |
| #8239 |
Add Spark 3.2.4 shim |
| #8221 |
[Doc] update getting started guide based on latest databricks env [skip ci] |
| #8224 |
Fix misinterpretation of Parquet's legacy ARRAY schemas. |
| #8241 |
Update to filecache API changes |
| #8244 |
Remove semicolon at the end of the package statement in Scala files |
| #8245 |
Remove redundant open of ORC file |
| #8252 |
Fix auto merge conflict 8250 [skip ci] |
| #8170 |
Update GpuRunningWindowExec to use OOM retry framework |
| #8218 |
Update to add 340 build and unit test in premerge and in JDK 11 build |
| #8232 |
Add integration tests for inferred schema |
| #8223 |
Use SupportsRuntimeV2Filtering in Spark 3.4.0 |
| #8233 |
cudf-udf integration test against python3.9 [skip ci] |
| #8226 |
Offset support for TakeOrderedAndProject |
| #8237 |
Use weak keys in executor broadcast plan cache |
| #8229 |
Upgrade to jacoco 0.8.8 for JDK 17 support |
| #8216 |
Add oom retry handling for GpuGenerate.fixedLenLazyArrayGenerate |
| #8191 |
Add in retry-work to GPU OutOfCore Sort |
| #8228 |
Partial JDK 17 support |
| #8227 |
Adjust defaults for better performance out of the box |
| #8212 |
Add file caching |
| #8179 |
Fall back to CPU for try_cast in Spark 3.4.0 |
| #8220 |
Batch install-file executions in a single JVM |
| #8215 |
Fix count from ORC files with no column names |
| #8192 |
Handle PySparkException in case of literal expressions |
| #8190 |
Fix element_at_index_zero integration test by using newer error message from Spark 3.4.0 |
| #8203 |
Clean up queued batches on task failures in RapidsShuffleThreadedBlockIterator |
| #8207 |
Support std aggregation in reduction |
| #8174 |
[FEA] support json to struct function |
| #8195 |
Bump mockito to 3.12.4 |
| #8193 |
Increase databricks cluster autotermination to 6.5 hours [skip ci] |
| #8182 |
Support STRING order-by columns for RANGE window functions |
| #8167 |
Add oom retry handling to GpuGenerateExec.doGenerate path |
| #8183 |
Disable asserts for non-empty nulls |
| #8177 |
Fix 340 shim of GpuCreateDataSourceTableAsSelectCommand and shim GpuDataSource for 3.4.0 |
| #8159 |
Verify CPU fallback class when creating HIVE table [Databricks] |
| #8180 |
Follow-up for ORC Decimal read failure (#8172) |
| #8172 |
Fix ORC decimal read when precision/scale changes |
| #7227 |
Fix PCBS integration tests for Spark-3.4 |
| #8175 |
Restore test_substring_column |
| #8162 |
Support Java 17 for packaging |
| #8169 |
Fix AnsiCastShim for 330db |
| #8168 |
[DOC] Updating profiling/qualification docs for usability improvements [skip ci] |
| #8144 |
Add 340 shim for GpuInsertIntoHiveTable |
| #8143 |
Add handling for SplitAndRetryOOM in nextCbFromGatherer |
| #8102 |
Rewrite two tests from AnsiCastOpSuite in Python and make compatible with Spark 3.4.0 |
| #8152 |
Fix Spark-3.4 test failure in AdaptiveQueryExecSuite |
| #8154 |
Use repo1.maven.org/maven2 instead of default apache central url |
| #8150 |
xfail test_substring_column |
| #8128 |
Fix CastOpSuite failures with Spark 3.4 |
| #8145 |
Fix nz timestamp unit tests |
| #8146 |
Set version of slf4j for Spark 3.4.0 |
| #8058 |
Add retry to BatchByKeyIterator |
| #8142 |
Enable ParquetWriterSuite test 'sorted partitioned write' on Spark 3.4.0 |
| #8035 |
[FEA] support StringTranslate function |
| #8136 |
Add GPU support for KnownNullable expression (Spark 3.4.0) |
| #8096 |
Add OOM retry handling for existence joins |
| #8139 |
Fix auto merge conflict 8138 [skip ci] |
| #8135 |
Fix Orc writer test failure with Spark 3.4 |
| #8129 |
Fix compile error with Spark 3.4.0 release and bump to use 3.4.0 release JAR |
| #8093 |
Add cuda12 build support [skip ci] |
| #8108 |
Make Arm methods static |
| #8060 |
Support repetitions in regexp choice expressions |
| #8081 |
Re-enable empty repetition near end-of-line anchor for rlike, regexp_extract and regexp_replace |
| #8075 |
Update some integration tests so that they are compatible with Spark 3.4.0 |
| #8063 |
Update docker to support integration tests against JDK17 [skip ci] |
| #8047 |
Enable line/string anchors in choice |
| #7996 |
Sub-partitioning supports repartitioning the input data multiple times |
| #8009 |
Add in some more retry blocks |
| #8051 |
MINOR: Improve assertion error in assert_py4j_exception |
| #8020 |
[FEA] Add Spark 3.3.3-SNAPSHOT to shims |
| #8034 |
Fix the check for dedicated per-shim files [skip ci] |
| #7978 |
Update JNI and private deps version to 23.06.0-SNAPSHOT |
| #7965 |
Remove stale references to the pre-shimplify dirs |
| #7948 |
Init plugin version 23.06.0-SNAPSHOT |