Skip to content
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.

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

Add release notes for 0.285 #21500

Closed
wants to merge 3 commits into from

Conversation

wanglinsong
Copy link
Member

@wanglinsong wanglinsong commented Dec 8, 2023

Missing Release Notes

Avinash Jain

Beinan

Bikramjeet Vig

JiamingMai

Jimmy Lu

Ke

Kevin Wilfong

Lyublena Antova

Pratyush Verma

Sotirios Delimanolis

Zac Blanco

kedia,Akanksha

xiaoxmeng

Extracted Release Notes

  • Add views support in iceberg connector #19872 (Author: pratyakshsharma): Add views support in iceberg connector
    • Support view creation via Iceberg connector.
  • Match hive support for type widening in the hive connector #19983 (Author: Miguel Blanco Godón): Match hive support for type widening in the hive connector
    • Added type widening support to the parquet column readers (int, bigint, and float) to be able to perform type conversions between inmediate numerical types.
    • Updated docs accordingly (hive connector - schema evolution).
    • Changes in ParquetPageSourceFactory so that the type-check does not fail with the new widenings supported.
  • [native] Add JWT processing when using http(s) comms #20290 (Author: Christian Zentgraf): [native] Add JWT processing when using http(s) comms
    • Add support for internal authentication using JWT. It can be configured using configs "internal-communication.jwt.enabled=[true/false]", "internal-communication.shared-secret=" and "internal-communication.jwt.expiration-seconds=".
  • Fix iceberg table creation using glue metastore #20699 (Author: pratyakshsharma): Fix iceberg table creation using glue metastore
    • Fix iceberg table creation using glue metastore.
  • [Iceberg] Implement ANALYZE when Hive is configured #20720 (Author: Zac Blanco): [Iceberg] Implement ANALYZE when Hive is configured
    • Iceberg connector now supports ANALYZE when configured with a Hive catalog and the table is unpartitioned.
  • Set initial number of tasks for scaled writer with HBO #20901 (Author: feilong-liu): Set initial number of tasks for scaled writer with HBO
    • Add optimization for scaled writers with HBO, it can improve latency for query with scaled writer enabled. It's controlled by session property enable_hbo_for_scaled_writer and default to false.
  • Heap Memory Based Worker Flag to stop processing new split when in low memory #20946 (Author: Swapnil Tailor): Heap Memory Based Worker Flag to stop processing new split when in low memory
    • Add config to enable memory based split processing slow down. This can be enabled via task.memory-based-slowdown-threshold.
  • Support Parquet writer versions V1 and V2 #20957 (Author: Mahadevuni Naveen Kumar): Support Parquet writer versions V1 and V2
    • Support for Parquet writer versions V1 and V2 is now added for Hive and Iceberg catalogs. It can be toggled using session property parquet_writer_version and config property hive.parquet.writer.version. Valid values for these properties are PARQUET_1_0 and PARQUET_2_0. Default is PARQUET_2_0. E.g., set session parquet_writer_version=PARQUET_1_0 / hive.parquet.writer.version=PARQUET_1_0.
  • Track whether an optimization decision was cost-based #20990 (Author: Lyublena Antova): Track whether an optimization decision was cost-based
    • Print information about cost-based optimizers and the source of stats they use (CBO/HBO) in explain plans when session property verbose_optimizer_info_enabled=true.
  • [pinot connector] Fix pinot single quote literal push down issue #21020 (Author: Xiang Fu): [pinot connector] Fix pinot single quote literal push down issue
    • Fix pinot single quote literal push down issue.
  • Support deleting whole partitions in Iceberg table #21048 (Author: wangd): Support deleting whole partitions in Iceberg table
    • The Iceberg connector now supports DELETE FROM <table> [where <filter>] statements to delete one or more entire partitions.
  • Remove redundant cast to varchar in join clause #21050 (Author: feilong-liu): Remove redundant cast to varchar in join clause
    • Add an optimization, which removes the redundant cast to varchar expression in join condition. The optimizer is controlled by session property remove_redundant_cast_to_varchar_in_join and enabled by default.
  • Add session property to set input table matching threshold #21063 (Author: feilong-liu): Add session property to set input table matching threshold
    • Add a session property history_input_table_statistics_matching_threshold to set the threshold for input statistics match in HBO.
  • Disable timestamp with time zone type for ddl #21096 (Author: Ajay Gupte): Disable timestamp with time zone type for ddl
    • Timestamp with time zone data type is not allowed in create table and alter table statements.
  • Enhance join reordering to work with non-simple equi-join clauses #21153 (Author: Anant Aneja): Enhance join reordering to work with non-simple equi-join clauses
    • Enhance join-reordering to work with non-simple equi-join predicates.
  • Track result sizes of partial aggregate evaluation in HBO #21160 (Author: Lyublena Antova): Track result sizes of partial aggregate evaluation in HBO
    • Extends HBO to track statistics of execution partial aggregation nodes.
    • Introduces a new flag use_partial_aggregation_history, which controls whether or not partial aggregation histories are used in deciding whether to split aggregates into partial and final.
    • Change PushPartialAggregationThroughExchange optimization to use partial aggregation statistics when available, and when use_partial_aggregation_history=true.
    • When partial aggregation statistics are used in PushPartialAggregationThroughExchange, the optimization also applies to multi-key aggregations (as opposed to single-key only when CBO is used).
  • add REST API to docs index #21183 (Author: Steve Burnett): add REST API to docs index
    • Added links for the REST API resource pages to the Presto documentation.
  • Simplify union node with empty input #21186 (Author: feilong-liu): Simplify union node with empty input
    • Fix the simplify plan with empty input rule to also work for union node with empty input but more than one non-empty input.
  • Add null stats for join probe #21203 (Author: feilong-liu): Add null stats for join probe
    • Track null probe keys in join and use it to enable outer join null salt in history based optimization.
  • Support indicating partition transform when add column to Iceberg table #21206 (Author: wangd): Support indicating partition transform when add column to Iceberg table
    • The Iceberg connector now supports ALTER TABLE <table> ADD COLUMN <column> [WITH (partitioning = '<transform_func>')] statements to indicate partition transform when add column to table.
  • Fix logging of plan optimizer #21226 (Author: feilong-liu): Fix logging of plan optimizer
    • Fix logging of optimizer triggering, eliminating false positives where optimizer is logged as triggered but actually not. The PlanOptimizer now return both result plan and whether the plan changed or not.
  • Update Iceberg from 1.3.1 to 1.4.1 #21230 (Author: Eduard Tudenhoefner): Update Iceberg from 1.3.1 to 1.4.1
    • Update Iceberg version from 1.3.1 to 1.4.1.
  • Worker Low Heap Memory Task Killer #21254 (Author: Swapnil Tailor): Worker Low Heap Memory Task Killer
    • Add High Memory Task Killer which troggers if worker is running out of memory and garbage collection not able to reclaim enough memory. The feature is enabled by experimental.task.high-memory-task-killer-enabled. There are two strategies for the task killer which can be set via experiemental.task.high-memory-task-killer-strategy. (default: FREE_MEMORY_ON_FULL_GC, FREE_MEMORY_ON_FREQUENT_FULL_GC).
    • Free memory on full garbage collection strategy triggers task killer if worker is running low memory and full GC is not able to reclaim enough memory. Low memory threhold is set by experimental.task.high-memory-task-killer-heap-memory-threshold and Full GC reclaim memory threhold is set by experimental.task.high-memory-task-killer-reclaim-memory-threshold.
    • Free memory on frequent full garbage collection strategy triggers task killer if worker is having frequent garbage collection and enough memory is not reclaimed during the garbage collections. Frequent garbage collection is identified via experimental.task.high-memory-task-killer-frequent-full-gc-duration-threshold and reclaim memory threshold is configured via experimental.task.high-memory-task-killer-reclaim-memory-threshold.
  • Convert values node followed by false filter to empty values node #21275 (Author: feilong-liu): Convert values node followed by false filter to empty values node
    • Enable optimization of values node followed by false filter to empty values node.
  • Enable timestamp type for SHOW STATS command #21286 (Author: Ajay Gupte): Enable timestamp type for SHOW STATS command
    • SHOW STATS command will support timestamp column based on user session's time zone.
  • Improve latency for HBO optimizer #21297 (Author: feilong-liu): Improve latency for HBO optimizer
    • Improve latency of history based optimizer.
  • Check consistency of param N for min_byN/max_byN #21306 (Author: mmorgan98): Check consistency of param N for min_byN/max_byN
    • MinbyN/maxbyN's third argument (N) is now checked to be unique.
  • Skip history optimizer query registration if previous registration timeout #21308 (Author: feilong-liu): Skip history optimizer query registration if previous registration timeout
    • Improve history optimizer latency for queries which timeout during query registration in history optimizer.
  • Solve critical vulnerability of Presto UI from @babel/traverse npm package #21322 (Author: Yihong Wang): Solve critical vulnerability of Presto UI from @babel/traverse npm package
    • Update Babel and related npm packages to solve critical vulnerability: GHSA-67hx-6x53-jw92 The Babel packages are used to generate the JavaScript files for Presto UI. The update prevents the build servers or developers' machines from running arbitrary code while building the JavaScript files.
  • Fix Typo in High Memory Task Manager Config #21327 (Author: Swapnil Tailor): Fix Typo in High Memory Task Manager Config
    • Fix typo in high memory task killer strategy config experiemental.task.high-memory-task-killer-strategy to experimental.task.high-memory-task-killer-stragtegy.
  • Update Joda-Time to 2.12.5 #21329 (Author: Yihong Wang): Update Joda-Time to 2.12.5
    • Update Joda-Time to 2.12.5 to use 2023c tzdata. Note: a corresponding update to the Java runtime should also be made to ensure consistent timezone data. For example, Oracle JDK 8u381, tzdata-java-2023c rpm for OpenJDK, or use Timezone Updater Tool to apply 2023c tzdata to existing JVM.
  • Fix High Memory Task Killer to kill highest memory consuming query #21349 (Author: Swapnil Tailor): Fix High Memory Task Killer to kill highest memory consuming query
    • Fix High Memory Task killer to kill highest memory consuming query.

All Commits

  • f89299a Upgrade Alluxio to 304 (Beinan)
  • 27c6d9e Track result sizes of partial aggregate evaluation in HBO (Lyublena Antova)
  • 2229fc4 [Iceberg] Add native NDV read and write support (Zac Blanco)
  • fb85091 update deploy-docker.rst (Steve Burnett)
  • f2f6718 Update query config to conform to presto standards (Bikramjeet Vig)
  • b2fc43c [native]Fix duplicate root memory pool in case of background arbitration (xiaoxmeng)
  • 2c5bdb4 Update Drift dependency to use Header transport by default (Sotirios Delimanolis)
  • 279cd59 [native] Share HTTP connection pools between HTTP clients (Jimmy Lu)
  • 1e2a5ae [native] Set connection time out so connection error can be retried timely (Jimmy Lu)
  • 4b7cc47 [native] Advance Velox (Masha Basmanova)
  • 7c51fd2 [native] Streamline default number of threads in HTTP thread pools. (Sergey Pershin)
  • 4f9bc23 Bypass metastore locking in Glue (pratyakshsharma)
  • 81850dc [native] Convert input field names to lower case in plan conversion (Kevin Wilfong)
  • d1c5d83 [native] Add caching of parsed Types (Kevin Wilfong)
  • 6c48ec5 Solve critical vulnerability of Presto UI (Yihong Wang)
  • 1ca0f98 [native] Clean old tasks based on termination time (Kevin Wilfong)
  • 820b29f Check consistency of param N for min_byN/max_byN (mmorgan98)
  • 67d04e0 [docs] Add documentation on @NullablePosition (Zac Blanco)
  • 40d652d Add docs for Prestodb to MySQL type mapping in MySQL Connector (kedia,Akanksha)
  • ec00815 [native] Advance Velox (Kevin Wilfong)
  • 8815076 Fix latency for HBO optimizer (feilong-liu)
  • dae8a28 Skip HBO registration if previous registration timeout (feilong-liu)
  • d39ce0f Remove invalid query ID cache from HBO cache manager (feilong-liu)
  • 278c285 [docs] Add advanced usage for java functions impls (Zac Blanco)
  • 8282585 Fix High Memory Task Killer (Swapnil Tailor)
  • 7eb6188 [native] Add static mapping of native query configs (Bikramjeet Vig)
  • 08b771c [native] Add serde parameter to remote function server configuration (Pedro Pedreira)
  • d35917c Update Joda-Time to 2.12.5 (Yihong Wang)
  • ed06dfb x (Swapnil Tailor)
  • 641c269 Fix Typo in High Memory Task Manager Config (Swapnil Tailor)
  • 43f996d Allow Type in agg function metadata constructors (Zac Blanco)
  • 9a2e540 Add support for user provided execution strategy (Chandrashekhar Kumar Singh)
  • 02889cd SQL statements support simple literal, as well as Unicode usage (kedia,Akanksha)
  • e4d8fc1 [native] Ensure all ExchangeSource instances are created via make_shared(). (Sergey Pershin)
  • 3512441 [native pos] Extend logging in NativeExecutionProcess (Andrii Rosa)
  • 0d350c4 [native pos] Trigger coredump when native process becomes unresponsive (Andrii Rosa)
  • b20986f [native pos] Use RequestErrorTracker to retry task update requests (Andrii Rosa)
  • b0542ce [native pos] Recover Thread#interrupted flag (Andrii Rosa)
  • 5530cb6 [native pos] Safely publish NativeExecutionProcess#process (Andrii Rosa)
  • e8108d1 [native] Rename TestNativeProbabilityFunctionQueries to TestPresto... (Michael Shang)
  • 4e27cd0 Fix comments (aditi-pandit)
  • 42f0462 [native] Fix result for empty global grouping sets (aditi-pandit)
  • 32f76b9 [native] Make PrestoServer own the common thread pools (Jialiang Tan)
  • 16970a0 [native] Add native-execution-enabled flag to NativeQueryRunnerUtils.getNativeWorkerSystemProperties (aditi-pandit)
  • f0bd6a5 [native] add a test case for text reader (Michael Shang)
  • 8a0da75 [native]Change spiller executor to cpu executor (xiaoxmeng)
  • 4199cad [native] Advance Velox (aditi-pandit)
  • e9b4ed4 Add test case for deletion in Iceberg table with partition evolution (wangd)
  • d1183bc Disallow dropping partition column in Iceberg table explicitly (wangd)
  • 541d7ec [native]Advance velox version and code cleanup (xiaoxmeng)
  • 1e25021 Correct return type of cume_dist() in docs (Jonathan Hehir)
  • 8b56666 Enable timestamp type for SHOW STATS command (Ajay Gupte)
  • 3b0fb82 Upgrade parquet to 1.13.1 (wypb)
  • 2e26dea Update AWS SDK v1 to 1.12.560 (kedia,Akanksha)
  • bd3cf42 Worker Low Heap Memory Task Killer (Swapnil Tailor)
  • 14e44e4 [native] Advance velox version (aditi-pandit)
  • b1d33ab Make the data directory used by QueryRunners in sync (yingsu00)
  • 66aab8b Convert values node followed by false filter to empty values node (feilong-liu)
  • a4d40df Use in-memory history provider with HiveQueryRunner (Lyublena Antova)
  • a8a8814 Update Iceberg from 1.3.1 to 1.4.1 (Eduard Tudenhoefner)
  • f2ff4d1 [Iceberg] Clean up TableStatisticsMaker dead code (Zac Blanco)
  • 98b9d02 Automatically label doc-only PRs (Tim Meehan)
  • 28302cd [native pos] Unregister shuffle handle on ShuffleWriter#stop (Andrii Rosa)
  • 53a46d9 [native pos] Rely on Spark to decide if stage writes to shuffle (Andrii Rosa)
  • 2172b36 [native pos] Disable optimized-partition-update-serialization (Andrii Rosa)
  • 696abe8 [native] Mitigate partitioning incompatibility for system tables (Andrii Rosa)
  • d213e21 [native] Include native workers in the total number of cluster nodes (Andrii Rosa)
  • d2df443 Fix bug in recursive assignment resolver (Anant Aneja)
  • 7e89851 test (Jialiang Tan)
  • f1cd030 Increase timeout when starting test query runner (Michael Shang)
  • 6376fa0 [native] Offload work from io thread pool to driver thread pool (Jialiang Tan)
  • b138fb5 [native] Categorize memory abort exception as insufficient resource (Jialiang Tan)
  • d209e37 Updating documentation of MAP_TOP_N_KEYS (Avinash Jain)
  • 409404a Remove duplicate code for get Iceberg table of different catalogType (wangd)
  • efdd8e0 Enhance join reordering to work with non-simple equi-join clauses (Anant Aneja)
  • 061c340 [native] Pass serde parameters down to insert table handle (Jialiang Tan)
  • 83b20c0 [native]Advance velox version (xiaoxmeng)
  • 7c6bc6e [native] Translate array agg config to velox (Pranjal Shankhdhar)
  • 2a18649 Add null stats for join probe (feilong-liu)
  • 54edca6 Address review comments (James Xu)
  • 69bed7f Describe Lead/Lag + Null Offset behavior more clearly (James Xu)
  • b67703a Support indicating partition transform when add column to Iceberg table (wangd)
  • 4b4b6df Fix the content of Hive security configuration page (pratyakshsharma)
  • 4a93220 [native] Drop redundant sorting keys from Window and TopNRowNumber nodes (Masha Basmanova)
  • a627d9b [native] Advance Velox (Masha Basmanova)
  • 82a3bc8 Fix logging of plan optimizer (feilong-liu)
  • 9bce03a Disable timestamp with time zone type on create & alter table for iceberg tables. (Ajay Gupte)
  • a00fdc6 [native] Add decimal data test for checksum (wypb)
  • d4d3ad6 Reduce memory usage of OptimizedPartitionedOutputOperator (Pranjal Shankhdhar)
  • 4f3b8ed Publish query plan when a query is running in AQE (Chandrashekhar Kumar Singh)
  • e9adf8c Add PrestoSparkExecutionContext for capturing different runtine info for logging (Arjun Gupta)
  • b6a639b Expand documentation for reduce_agg (Masha Basmanova)
  • 4c8704c Add HANA connector (JiamingMai)
  • 4d52f63 Silently drop tables with missing filesystem metadata (kiersten-stokes)
  • 7bfbf52 Make setup-ubuntu.sh executable (James Xu)
  • a87bbc3 Fix Typo in PageBuilder (kedia,Akanksha)
  • ff83f1c [native] Wrap generation of OkResponse's HTTP message body in try/catch. (Sergey Pershin)
  • 580f2e3 [Native] Remove redundant plan conversion for TableWriteNode (Ke)
  • fe56f79 [native] Add google::InstallFailureSignalHandler() (Deepak Majeti)
  • 4ecd792 Fix Iceberg table partitioned by columns of VarbinaryType (wangd)
  • fa8ed14 [native] Avoid moving Lambda's arguments in toVeloxExpr(). (Sergey Pershin)
  • 36186c9 [Native] Use velox's OutputBufferManager instead of old PartitionedOutputBufferManager (Ke)
  • 2b2bf2e [Native] Advance velox version (Ke)
  • cb589d4 Simplify union node with empty input (feilong-liu)
  • d117ff8 Build catalog-specific properties for IcebergQueryRunner w static method (kiersten-stokes)
  • 16cb531 [native] Document enable-old-task-cleanup and old-task-cleanup-ms configs (Pratyush Verma)
  • 5e30a43 Fix Iceberg table partitioned by columns of TimestampType (wangd)
  • 8957e66 add REST API to docs index (Steve Burnett)
  • 6242b0d Improve layout of table for statistics in docs (kedia,Akanksha)
  • a53e4b1 Allow both Java system properties and environment variables for HiveExternalWorkerQueryRunner (yingsu00)
  • 1b1f861 [native pos] Disable old task cleanup routine (Pratyush Verma)
  • 411ea36 [native pos] Process PartitionAndSerialize in batches to reduce memory pressure (Pratyush Verma)
  • 70ecc39 Add cost based decision log for scaled writer optimizer (feilong-liu)
  • 4225f39 [native] Allow exchange payload copy to happen in driver thread (Jialiang Tan)
  • abc7491 [native-pos] Disable AsyncDataCache (thus prefetching) (Jimmy Lu)
  • 3d3ee49 Update Avro version 1.11.3 (kedia,Akanksha)
  • 8caad94 Use LinkedBlockingQueue to queue requests when all threads are busy (Nikhil Collooru)
  • ceffc7e [Native] Test fetching from finished velox task for scaled writer (Ke)
  • 1f43398 [Native] Advance velox version (Ke)
  • c2b7f2c [native] Replace DWRF writer creation with common factory creation (Christian Zentgraf)
  • 71d5640 Fix stream ordering in ORC writer (Sergii Druzkin)
  • 14f866e [Native] Add e2e test for scaled writer (Ke)
  • 53f9760 [native] Expose stats on 'Tasks with a stuck operator'. (Sergey Pershin)
  • 85e914d [Native] Enhance bucketed table writer test (Ke)
  • e59a393 Set cache blob size to fix flaky action (Zac Blanco)
  • ef0b199 [native] QueryRunner env variables as default config source (Zac Blanco)
  • afca9b5 [Native] move native writer related test to dedicated class (Ke)
  • 3ab3e09 Fix memory accounting in InMemoryHashAggregationBuilder (Pratyush Verma)
  • 2a77f69 [Native] cosmetic change to use destination instead of bufferId (Ke)
  • c2bc8d1 List all input types for geometric_mean in doc (James Xu)
  • 6661a55 [native]Add stats counter for the max spill level exceeded count (xiaoxmeng)
  • 6791ea4 Fix typo in TaskExecutor (wangd)
  • 9378300 [native]Add to retry get data error on memory allocation failure (xiaoxmeng)
  • c6292d0 [native] Set sane defaults for PRESTO_SERVER and DATA_DIR for NativeQueryRunner (Zac Blanco)
  • 3a738e9 [native] Advance Velox. (Sergey Pershin)
  • 91efd78 Throw when trying to interpret a non-scalar function (Sreeni Viswanadha)
  • a16e1f7 Adjust the column width of the task list table (Yihong Wang)
  • 626b7e7 Fix NPE in AbstractVarcharType.writeSlice (Sreeni Viswanadha)
  • 0e52fc3 [native] Refactor TypeSignatureTypeConverter and test (Masha Basmanova)
  • ca11fd6 [Iceberg] read and combine statistics from the HMS (Zac Blanco)
  • 45bcbcf [Iceberg] Write Presto statistics to HMS (Zac Blanco)
  • 6a582e8 [Native] Fix GROUPING SETS with multiple aliased columns (aditi-pandit)
  • bb13af3 Extend payload join optimization for cases without map columns + bug fixes (Lyublena Antova)
  • 0892870 Functions for extracting plan id information from json query plans (Lyublena Antova)
  • 8d9dfa9 Remove redundant cast in join clause (feilong-liu)
  • 627165a Revert "Fix memory reservation accounting problem about InMemoryHashAggregationBuilder inside SpillableHashAggregationBuilder." (Pratyush Verma)
  • adb705b Revert "Add unit test for memory limit in spillable hash aggregation operator when trigger rehash." (Pratyush Verma)
  • 0717944 Use the 3rd number of a TaskId as the task number (Yihong Wang)
  • deafa02 [native]Send completed get data result if velox task has finished (xiaoxmeng)
  • c4df72a [native] Populate AggregationNode::Aggregate::rawInputTypes (Masha Basmanova)
  • da7b0f2 [native] Add support for parsing FUNCTION types (Masha Basmanova)
  • c7818f1 [native]Add non-reclaimable operator memory reclaim attempts (xiaoxmeng)
  • 5a08d8a Fix QueryContextManagerTest.defaultSessionProperties (aditi-pandit)
  • 035f523 [Native] Advance Velox version (aditi-pandit)
  • 1216027 [native] Add spill max file size to session property (Jialiang Tan)
  • aa39d9c Add session property to set input table matching threshold (feilong-liu)
  • bb433ab Support deleting whole partitions in Iceberg table (wangd)
  • 18ac248 Use HBO to set number of tasks for scaled writers (feilong-liu)
  • 1e249e1 Track number of tasks for scaled writer in HBO (feilong-liu)
  • 9104d71 Add numWrittenFiles in RuntimeStats for TableWriter (Ke)
  • 5d0ae9e [native pos] Fail fast on the driver if broadcast size exceeds the limit (Chandrashekhar Kumar Singh)
  • 72763e6 [native pos] Report estimated size and row count of broadcast data (Chandrashekhar Kumar Singh)
  • 627da72 Fix RowExpressionTest to be forward compatible with variant.toJson changes. (Krishna Pai)
  • b8c2241 [native] Add query memory session override (Jialiang Tan)
  • 3ca1339 [native] Use QueryConfig's default value for base default (Jialiang Tan)
  • 01c962e Update Low memory Monitor feature config to experimental (Swapnil Tailor)
  • 6bceeae [native] Add native_aggregation_spill_all session property (Jialiang Tan)
  • 4114f97 [native] Advance Velox version (Jialiang Tan)
  • 384aca5 Update partition spec when rename partition column in Iceberg table (wangd)
  • ef8ea9f [native][circleci] Fix native build failure (Michael Shang)
  • ce8acd1 Support sorting in the task list (Yihong Wang)
  • e795320 Use React table instead of reactalbe (Yihong Wang)
  • 53e07a1 Track whether an optimization decision was cost-based (Lyublena Antova)
  • 4583949 Allow hive pushdown filter for Analyze on partitions (Karteekmurthys)
  • 2436043 Improve documentation of Presto Resource Groups (kedia,Akanksha)
  • 6c9b457 [natie]Fix spill query config default setting (xiaoxmeng)
  • ca3012c [TEST][native][circleci] test on build parallemism (Michael Shang)
  • 138c6cc [native][circle] lower parallelism to avoid an oom issue (Michael Shang)
  • a23819a fixing pinot single quote literal push down issue (Xiang Fu)
  • 9d121df Add support for array_least_frequent UDF (Avinash Jain)
  • ed0c6e7 Skip Javadoc generation for Maven checks (Tim Meehan)
  • 13952e2 [native] Stop using PlanBuilder::finalAggregation with resultTypes API (Masha Basmanova)
  • 21ab1ea [native] Add PrestoQueryRunner (Masha Basmanova)
  • 09885f6 [Native] Add debug config for native engine execution (Bikramjeet Vig)
  • dab588a [native]advance velox version and update the code (xiaoxmeng)
  • d1ede7c [native pos] Enable TestPrestoSparkNativeGeneralQueries.testOrderBy/Subqueries/TopN (Masha Basmanova)
  • 13ab3dc Fix merge conflict in TestHiveTypeWidening (Tim Meehan)
  • ac62c07 [native] Advance velox version (Bikramjeet Vig)
  • 0010d55 Heap Memory Usage Based Worker Split Processing Slow Down (Swapnil Tailor)
  • 154ceeb Add views support in iceberg connector (pratyakshsharma)
  • a8d9b52 Add TestIcebergHiveMetadataListing.java class (pratyakshsharma)
  • 76b25a6 Move buildInitialPrivilegeSet to MetastoreUtil (pratyakshsharma)
  • 5434a36 Support Parquet writer versions V1 and V2 (Mahadevuni Naveen Kumar)
  • 92a4592 [native] Add list all task endpoint to server operation (Jialiang Tan)
  • ed5727c [native] Add JWT processing when using http(s) comms (Christian Zentgraf)
  • 3736f69 [native]Improve spill directory failure handling (xiaoxmeng)
  • e4d6044 Remove subfield pruning for lambda for parametric functions (Ruslan Mardugalliamov)
  • 7a6f376 Add subfield pruning from lambdas (Ruslan Mardugalliamov)
  • aef4860 Add NoneSubfield subfield path element (Ruslan Mardugalliamov)
  • aaf917d Add type widening support to numeric types (Miguel Blanco Godón)
  • accd109 [Native] Trim native prefix for system session property (Ke)

@wanglinsong wanglinsong requested a review from a team as a code owner December 8, 2023 01:36
Copy link

github-actions bot commented Dec 8, 2023

Codenotify: Notifying subscribers in CODENOTIFY files for diff e718f7a...db0570f.

Notify File(s)
@steveburnett presto-docs/src/main/sphinx/connector/hive.rst
presto-docs/src/main/sphinx/connector/iceberg.rst
presto-docs/src/main/sphinx/release.rst
presto-docs/src/main/sphinx/release/release-0.285.rst

@github-actions github-actions bot added the docs label Dec 8, 2023
@wanglinsong
Copy link
Member Author

@tdcmeehan @steveburnett please coordinate with the code contributor to review the release notes.

Copy link
Collaborator

@majetideepak majetideepak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@wanglinsong some comments. I will review the issues you linked in the PR.
I will also be coordinating with the code contributors for missing release notes.


Prestissimo (native Execution) Changes
______________________________________
* Add support for internal authentication using JWT. It can be configured using configs "internal-communication.jwt.enabled=[true/false]", "internal-communication.shared-secret=<shared-secret-value>" and "internal-communication.jwt.expiration-seconds=<value in seconds>".
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's add a link to the JWT documentation.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@nmahadevuni Please update.

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@czentgr Can you update the Prestissimo doc ?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The option is documented at
../develop/presto-native.rst. This contains a list of options for prestissimo.
The actual link with anchor should be
../develop/presto-native.html#jwt-authentication-support.

Do I need to do anything else?

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That should be it.

presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved

Hive Changes
____________
* Changes in ParquetPageSourceFactory so that the type-check does not fail with the new widenings supported.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We need to fix this statement. It is not clear what changes were made and what new widenings mean.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@aaneja can you update this?

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The general change this is part of is already mentioned ' Added type widening support to the parquet column readers (int, bigint, and float) to be able to perform type conversions between inmediate numerical types.'

This release note for a class level change is not needed, IMO

Hive Changes
____________
* Changes in ParquetPageSourceFactory so that the type-check does not fail with the new widenings supported.
* Support for Parquet writer versions V1 and V2 is now added for Hive and Iceberg catalogs. It can be toggled using session property `parquet_writer_version` and config property `hive.parquet.writer.version`. Valid values for these properties are PARQUET_1_0 and PARQUET_2_0. Default is PARQUET_2_0. E.g., set session parquet_writer_version=PARQUET_1_0 / hive.parquet.writer.version=PARQUET_1_0.
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Support Parquet writer versions V1 and V2 in Hive and Iceberg catalogs.
Let's add a link to the document for the config.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@nmahadevuni can you update the hive/iceberg connector docs?

presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
Copy link
Member Author

@wanglinsong wanglinsong left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Updates per review.

presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved
Hive Changes
____________
* Changes in ParquetPageSourceFactory so that the type-check does not fail with the new widenings supported.
* Support for Parquet writer versions V1 and V2 is now added for Hive and Iceberg catalogs. It can be toggled using session property `parquet_writer_version` and config property `hive.parquet.writer.version`. Valid values for these properties are PARQUET_1_0 and PARQUET_2_0. Default is PARQUET_2_0. E.g., set session parquet_writer_version=PARQUET_1_0 / hive.parquet.writer.version=PARQUET_1_0.
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@nmahadevuni can you update the hive/iceberg connector docs?

presto-docs/src/main/sphinx/release/release-0.285.rst Outdated Show resolved Hide resolved

Prestissimo (native Execution) Changes
______________________________________
* Add support for internal authentication using JWT. It can be configured using configs "internal-communication.jwt.enabled=[true/false]", "internal-communication.shared-secret=<shared-secret-value>" and "internal-communication.jwt.expiration-seconds=<value in seconds>".
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@nmahadevuni Please update.


Hive Changes
____________
* Changes in ParquetPageSourceFactory so that the type-check does not fail with the new widenings supported.
Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@aaneja can you update this?

@tdcmeehan
Copy link
Contributor

Closing in favor of #21579

@tdcmeehan tdcmeehan closed this Dec 22, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
Status: Done
Development

Successfully merging this pull request may close these issues.

6 participants