This is an automated email from the ASF dual-hosted git repository.
mblow pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/asterixdb.git
commit bbfa4becf18ce9c60dc6c57ae3f97c5c4b6ccd99
Merge: f9c6fe50c6 e15750431c
Author: Michael Blow <michael.b...@couchbase.com>
AuthorDate: Fri Sep 19 19:46:33 2025 -0400
Merge branch 'gerrit/phoenix' into 'master'
* [ASTERIXDB-3641] Sampling query taking long time to run
* [ASTERIXDB-3636][STO] Fix buffer reservations understimation
* [NO ISSUE][*HYR][MISC] += Lazy helper for deferred initialization
* [NO ISSUE][RT] Notify joblet cleanup if joblet creation fails
* [NO ISSUE][*DB][EXT] Avoid serialization of JsonFactory
* [ASTERIXDB-3641] Sampling query taking long time to run
* [NO ISSUE][*DB][RT] Whisper logging for active partition,txn block msgs
* [NO ISSUE][*DB][NET] Upgrade Netty to 4.1.125.Final
* [ASTERIXDB-3644] Callback for result consumption
* [NO ISSUE] Split countn_01 into multiple queries
* [NO ISSUE][HYR][STO] FileMapManager performance improvements
* [NO ISSUE][HYR][STO] Add missing serialVersionUID
* [NO ISSUE] Fix result for pushdown test
* [ASTERIXDB-3646][COMP] Missing fields when selecting all fields with
window function
Change-Id: I24dcea00986058def1eaac1cf317147c5b3e1196
.../active/message/ActivePartitionMessage.java | 5 +
.../apache/asterix/optimizer/rules/cbo/Stats.java | 52 ++++++++--
.../optimizer/rules/pushdown/PushdownContext.java | 2 +-
.../api/http/server/QueryResultApiServlet.java | 12 ++-
.../apache/asterix/app/nc/NCAppRuntimeContext.java | 11 +--
.../apache/asterix/test/common/TestExecutor.java | 7 ++
.../aggregate/countn_01/countn_01.3.query.sqlpp | 18 +---
.../aggregate/countn_01/countn_01.5.query.sqlpp | 2 +-
...tn_01.5.query.sqlpp => countn_01.6.query.sqlpp} | 14 +--
...tn_01.5.query.sqlpp => countn_01.7.query.sqlpp} | 0
.../async-deferred/AsyncDeferredQueries.xml | 2 +-
.../window_subquery/window_query.001.ddl.sqlpp} | 25 ++---
.../window_subquery/window_query.002.update.sqlpp} | 35 ++++---
.../window_subquery/window_query.003.query.sqlpp} | 20 ++--
.../window_subquery/window_query.004.query.sqlpp} | 20 ++--
.../results/aggregate/countn_01/countn_01.1.adm | 5 +-
.../results/aggregate/countn_01/countn_01.3.adm | 2 +-
.../results/aggregate/countn_01/countn_01.4.adm | 1 +
.../countn_01/{countn_01.3.adm => countn_01.5.adm} | 0
.../api/cluster_state_1/cluster_state_1.1.regexadm | 4 -
.../cluster_state_1_full.1.regexadm | 4 -
.../cluster_state_1_less.1.regexadm | 4 -
.../same-datasource-diff-scan.004.plan | 64 ++++++------
.../same-datasource-diff-scan.005.plan | 96 +++++++++---------
.../same-datasource-diff-scan.006.plan | 98 +++++++++----------
.../same-datasource-diff-scan.007.plan | 66 ++++++-------
.../pushdown/window_subquery/window_query.002.adm | 1 +
.../pushdown/window_subquery/window_query.003.adm | 1 +
.../same-datasource-diff-scan.004.plan | 72 +++++++-------
.../same-datasource-diff-scan.005.plan | 106 ++++++++++----------
.../same-datasource-diff-scan.006.plan | 108 ++++++++++-----------
.../same-datasource-diff-scan.007.plan | 64 ++++++------
.../same-datasource-diff-scan.004.plan | 72 +++++++-------
.../same-datasource-diff-scan.005.plan | 106 ++++++++++----------
.../same-datasource-diff-scan.006.plan | 108 ++++++++++-----------
.../same-datasource-diff-scan.007.plan | 64 ++++++------
.../src/test/resources/runtimets/sqlpp_queries.xml | 5 +
...tractParquetDeltaBinaryPackingValuesWriter.java | 2 +
.../bytes/encoder/ParquetDeltaByteArrayWriter.java | 4 +
.../ParquetDeltaLengthByteArrayValuesWriter.java | 3 +
.../ParquetPlainFixedLengthValuesWriter.java | 2 +
.../ParquetPlainVariableLengthValuesWriter.java | 2 +
.../operation/lsm/flush/FlushColumnMetadata.java | 25 +++--
.../lsm/flush/FlushColumnTupleWriter.java | 12 +++
.../lsm/flush/NoWriteFlushColumnMetadata.java | 45 +++++++--
.../lsm/merge/MergeColumnTupleWriter.java | 9 ++
.../lsm/merge/MergeColumnWriteMetadata.java | 13 ++-
.../asterix/column/values/IColumnValuesWriter.java | 5 +
.../values/writer/AbstractColumnValuesWriter.java | 2 +
.../values/writer/BooleanColumnValuesWriter.java | 9 ++
.../values/writer/DoubleColumnValuesWriter.java | 9 ++
.../values/writer/FloatColumnValuesWriter.java | 9 ++
.../values/writer/LongColumnValuesWriter.java | 15 ++-
.../values/writer/NoOpColumnValuesWriter.java | 9 ++
.../writer/NullMissingColumnValuesWriter.java | 9 ++
.../values/writer/StringColumnValuesWriter.java | 12 +++
.../values/writer/UUIDColumnValuesWriter.java | 11 ++-
.../values/writer/DummyColumnValuesWriter.java | 5 +
.../asterix/common/config/StorageProperties.java | 33 ++++---
.../parser/factory/JSONDataParserFactory.java | 6 +-
.../asterix/runtime/message/TxnIdBlockRequest.java | 5 +
asterixdb/pom.xml | 2 +-
...4j-persistence-extension--1.3.0_MIT_License.txt | 3 +-
.../algebricks/algebricks-core/pom.xml | 2 +-
.../hyracks/api/result/IResultDirectory.java | 2 +
.../api/result/IResultPartitionManager.java | 4 +-
.../hyracks/api/result/ResultDirectoryRecord.java | 4 +
.../apache/hyracks/api/result/ResultJobRecord.java | 23 ++++-
.../java/org/apache/hyracks/api/util/Lazy.java | 40 ++++----
.../hyracks/client/result/ResultDirectory.java | 5 +
.../client/result/ResultDirectoryRemoteProxy.java | 5 +
hyracks-fullstack/hyracks/hyracks-cloud/pom.xml | 2 +-
.../hyracks/control/cc/ClusterControllerIPCI.java | 7 ++
.../control/cc/result/IResultDirectoryService.java | 2 +
.../control/cc/result/ResultDirectoryService.java | 12 +++
.../cc/work/ReportResultPartitionConsumedWork.java | 57 +++++++++++
.../control/common/base/IClusterController.java | 2 +
.../hyracks/control/common/ipc/CCNCFunctions.java | 34 +++++++
.../common/ipc/ClusterControllerRemoteProxy.java | 7 ++
.../control/nc/result/ResultPartitionManager.java | 15 ++-
.../control/nc/result/ResultPartitionReader.java | 4 +-
.../nc/work/FailedJobletCreationCleanupWork.java | 61 ++++++++++++
.../hyracks/control/nc/work/StartTasksWork.java | 6 +-
.../hyracks/hyracks-data/hyracks-data-std/pom.xml | 2 +-
.../hyracks/hyracks-dataflow-common/pom.xml | 2 +-
.../examples/btree/helper/RuntimeContext.java | 3 +-
.../hyracks/hyracks-storage-am-btree/pom.xml | 2 +-
.../hyracks/hyracks-storage-am-common/pom.xml | 2 +-
.../hyracks-storage-am-lsm-btree-column/pom.xml | 2 +-
.../column/api/AbstractColumnTupleWriter.java | 6 ++
.../column/impls/btree/ColumnBTreeBulkloader.java | 19 ++--
.../hyracks/hyracks-storage-am-lsm-common/pom.xml | 2 +-
.../am/lsm/common/impls/VirtualBufferCache.java | 33 +++----
.../hyracks/hyracks-storage-common/pom.xml | 4 +
.../common/buffercache/ColumnBufferPool.java | 99 ++++++++++---------
.../common/buffercache/FreeColumnBufferPool.java | 17 ++++
.../common/buffercache/IColumnBufferPool.java | 4 +-
.../storage/common/file/FileMapManager.java | 54 +++++++++--
.../storage/common/file/IFileMapManager.java | 19 ++++
.../common/file/SynchronizedFileMapManager.java | 87 +++++++++++++++++
.../support/TestStorageManagerComponentHolder.java | 3 +-
.../storage/common/ColumnBufferPoolTest.java | 16 +--
hyracks-fullstack/hyracks/hyracks-util/pom.xml | 2 +-
hyracks-fullstack/pom.xml | 4 +-
104 files changed, 1421 insertions(+), 792 deletions(-)
diff --cc
asterixdb/asterix-algebra/src/main/java/org/apache/asterix/optimizer/rules/cbo/Stats.java
index 7ffffd53b7,78cf365daa..aa73da13e0
---
a/asterixdb/asterix-algebra/src/main/java/org/apache/asterix/optimizer/rules/cbo/Stats.java
+++
b/asterixdb/asterix-algebra/src/main/java/org/apache/asterix/optimizer/rules/cbo/Stats.java
@@@ -175,13 -173,17 +175,14 @@@ public class Stats
boolean unnestOp1 =
joinEnum.findUnnestOp(joinEnum.leafInputs.get(idx1 - 1));
boolean unnestOp2 =
joinEnum.findUnnestOp(joinEnum.leafInputs.get(idx2 - 1));
boolean unnestOp = unnestOp1 || unnestOp2;
+ boolean okOp = acceptableOp(joinExpr.getFunctionIdentifier());
- Index.SampleIndexDetails idxDetails1 = (Index.SampleIndexDetails)
index1.getIndexDetails();
- Index.SampleIndexDetails idxDetails2 = (Index.SampleIndexDetails)
index2.getIndexDetails();
- if (((idxDetails1.getSourceCardinality() <
idxDetails1.getSampleCardinalityTarget())
- || (idxDetails2.getSourceCardinality() <
idxDetails2.getSampleCardinalityTarget())
- ||
(!(joinExpr.getFunctionIdentifier().equals(AlgebricksBuiltinFunctions.EQ)))
- || exprUsedVars.size() > 2) && !unnestOp && okOp) { //*
if there are more than 2 variables, it is not a simple join like r.a op s.a
+ ILogicalOperator leafInput1 = joinEnum.leafInputs.get(idx1 - 1);
+ ILogicalOperator leafInput2 = joinEnum.leafInputs.get(idx2 - 1);
+ LogicalVariable var1 = exprUsedVars.get(0);
+ LogicalVariable var2 = exprUsedVars.get(1);
+ // If there are more than 2 variables, it is not a simple join
like r.a op s.a
- if (!unnestOp && (exprUsedVars.size() > 2
++ if (okOp && !unnestOp && (exprUsedVars.size() > 2
+ || isJoinSelFromSamplesApplicable(leafInput1, leafInput2,
index1, index2, var1, var2))) {
double sels =
findJoinSelFromSamples(joinEnum.leafInputs.get(idx1 - 1),
joinEnum.leafInputs.get(idx2 - 1), index1, index2,
joinExpr, jOp);
if (sels == 0.0) {
@@@ -195,28 -197,45 +196,67 @@@
}
}
+ private boolean isJoinSelFromSamplesApplicable(ILogicalOperator
leafInput1, ILogicalOperator leafInput2,
+ Index index1, Index index2, LogicalVariable var1, LogicalVariable
var2) throws AlgebricksException {
+ Index.SampleIndexDetails details1 = (Index.SampleIndexDetails)
index1.getIndexDetails();
+ Index.SampleIndexDetails details2 = (Index.SampleIndexDetails)
index2.getIndexDetails();
+ if (details1.getSourceCardinality() >=
details1.getSampleCardinalityTarget()
+ && details2.getSourceCardinality() >=
details2.getSampleCardinalityTarget()) {
+ return false;
+ }
+ double numDistinct1 = computeNumDistinct(leafInput1, var1, index1);
+ if (numDistinct1 < 0) {
+ return false;
+ }
+ double avgNumRowsPerValue1 = details1.getSourceCardinality() /
numDistinct1;
+ double numDistinct2 = computeNumDistinct(leafInput2, var2, index2);
+ if (numDistinct2 < 0) {
+ return false;
+ }
+ double avgNumRowsPerValue2 = details2.getSourceCardinality() /
numDistinct2;
+ return avgNumRowsPerValue1 * avgNumRowsPerValue2 *
Math.min(numDistinct1, numDistinct2) <= Math
+ .max(Math.max(details1.getSourceCardinality(),
details2.getSourceCardinality()), 750000);
+ }
+
+ private double computeNumDistinct(ILogicalOperator leafInput,
LogicalVariable var, Index index)
+ throws AlgebricksException {
+ List<List<IAObject>> result = runSamplingQueryDistinct(this.optCtx,
leafInput, var, index);
+ if (result == null) {
+ return -1; // Negative value indicates failure
+ }
+ double numDistincts = findPredicateCardinality(result, true);
+ Index.SampleIndexDetails details = (Index.SampleIndexDetails)
index.getIndexDetails();
+ if (numDistincts == 0) {
+ numDistincts = details.getSourceCardinality(); // All values are
nulls
+ }
+ if (numDistincts == 0) {
+ numDistincts = 1; // Sample is empty
+ }
+ return numDistincts;
+ }
+
+ private boolean acceptableOp(FunctionIdentifier functionIdentifier) {
+ if (functionIdentifier.equals(AlgebricksBuiltinFunctions.NEQ)) {
+ return true;
+ }
+ if (functionIdentifier.equals(AlgebricksBuiltinFunctions.LT)) {
+ return true;
+ }
+ if (functionIdentifier.equals(AlgebricksBuiltinFunctions.GT)) {
+ return true;
+ }
+ if (functionIdentifier.equals(AlgebricksBuiltinFunctions.GE)) {
+ return true;
+ }
+ if (functionIdentifier.equals(AlgebricksBuiltinFunctions.LE)) {
+ return true;
+ }
+ if (functionIdentifier.equals(BuiltinFunctions.IF_MISSING_OR_NULL)) {
// added this for q16 in CH2
+ return true;
+ }
+ return false;
+ }
+
private double naiveJoinSelectivity(List<LogicalVariable> exprUsedVars,
double card1, double card2, int idx1,
int idx2, boolean unnestOp1, boolean unnestOp2) throws
AlgebricksException {
ILogicalOperator leafInput;
