database/doris
2
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

[fix](planner) Fix table sample not take effect if exist conjunct #29814

Browse Source
This commit is contained in:
Xinyi Zou
2024-01-11 04:30:58 +08:00
committed by yiguolei
parent 4a98db2d47
commit a94343c5f9
3 changed files with 26 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
fe/fe-core/src/main/java/org/apache/doris/nereids/glue/translator/PhysicalPlanTranslator.java
View File

@ -646,7 +646,7 @@ public class PhysicalPlanTranslator extends DefaultPlanVisitor<PlanFragment, Pla
BaseTableRef tableRef = new BaseTableRef(ref, olapTable, tableName);
tupleDescriptor.setRef(tableRef);
olapScanNode.setSelectedPartitionIds(olapScan.getSelectedPartitionIds());
olapScanNode.setSampleTabletIds(olapScan.getSelectedTabletIds()); // TODO
olapScanNode.setSampleTabletIds(olapScan.getSelectedTabletIds());
if (olapScan.getTableSample().isPresent()) {
olapScanNode.setTableSample(new TableSample(olapScan.getTableSample().get().isPercent,
olapScan.getTableSample().get().sampleValue, olapScan.getTableSample().get().seek));

23
fe/fe-core/src/main/java/org/apache/doris/planner/OlapScanNode.java
View File

@ -999,6 +999,7 @@ public class OlapScanNode extends ScanNode {
// 3. Sampling partition. If Seek is specified, the partition will be the same for each sampling.
long hitRows = 0; // The number of rows hit by the tablet
Set<Long> hitTabletIds = Sets.newHashSet();
long partitionSeek = tableSample.getSeek() != -1
? tableSample.getSeek() : (long) (new SecureRandom().nextDouble() * selectedPartitionList.size());
for (int i = 0; i < selectedPartitionList.size(); i++) {
@ -1024,16 +1025,24 @@ public class OlapScanNode extends ScanNode {
? tableSample.getSeek() : (long) (new SecureRandom().nextDouble() * tablets.size());
for (int j = 0; j < tablets.size(); j++) {
int seekTid = (int) ((j + tabletSeek) % tablets.size());
Tablet tablet = tablets.get(seekTid);
if (sampleTabletIds.size() != 0 && !sampleTabletIds.contains(tablet.getId())) {
// After PruneOlapScanTablet, sampleTabletIds.size() != 0,
// continue sampling only in sampleTabletIds.
// If it is percentage sample, the number of sampled rows is a percentage of the
// total number of rows, and It is not related to sampleTabletI after PruneOlapScanTablet.
continue;
}
long tabletRowCount;
if (!FeConstants.runningUnitTest) {
tabletRowCount = tablets.get(seekTid).getRowCount(true);
tabletRowCount = tablet.getRowCount(true);
} else {
tabletRowCount = selectedTable.getRowCount() / tablets.size();
}
if (tabletRowCount == 0) {
continue;
}
sampleTabletIds.add(tablets.get(seekTid).getId());
hitTabletIds.add(tablet.getId());
sampleRows -= tabletRowCount;
hitRows += tabletRowCount;
if (sampleRows <= 0) {
@ -1044,7 +1053,15 @@ public class OlapScanNode extends ScanNode {
break;
}
}
LOG.debug("after computeSampleTabletIds, hitRows {}, selectedRows {}", hitRows, selectedRows);
if (sampleTabletIds.size() != 0) {
sampleTabletIds.retainAll(hitTabletIds);
LOG.debug("after computeSampleTabletIds, hitRows {}, totalRows {}, selectedTablets {}, sampleRows {}",
hitRows, selectedRows, sampleTabletIds.size(), totalSampleRows);
} else {
sampleTabletIds = hitTabletIds;
LOG.debug("after computeSampleTabletIds, hitRows {}, selectedRows {}, sampleRows {}", hitRows, selectedRows,
totalSampleRows);
}
}
public boolean isFromPrepareStmt() {

5
fe/fe-core/src/test/java/org/apache/doris/analysis/SelectStmtTest.java
View File

@ -961,6 +961,11 @@ public class SelectStmtTest {
OriginalPlanner planner16 = (OriginalPlanner) dorisAssert.query(sql16).internalExecuteOneAndGetPlan();
Set<Long> sampleTabletIds16 = ((OlapScanNode) planner16.getScanNodes().get(0)).getSampleTabletIds();
Assert.assertEquals(1, sampleTabletIds16.size());
String sql17 = "SELECT * FROM db1.table1 TABLESAMPLE(15 PERCENT) where siteid != 0";
OriginalPlanner planner17 = (OriginalPlanner) dorisAssert.query(sql17).internalExecuteOneAndGetPlan();
Set<Long> sampleTabletIds17 = ((OlapScanNode) planner17.getScanNodes().get(0)).getSampleTabletIds();
Assert.assertEquals(2, sampleTabletIds17.size());
FeConstants.runningUnitTest = false;
}