ravipesala commented on a change in pull request #3177:
[CARBONDATA-3337][CARBONDATA-3306] Distributed index server
URL: https://github.com/apache/carbondata/pull/3177#discussion_r279427052
##########
File path:
core/src/main/java/org/apache/carbondata/core/datamap/DataMapUtil.java
##########
@@ -91,43 +102,108 @@ public static DataMapJob getDataMapJob(Configuration
configuration) throws IOExc
* @param carbonTable
* @throws IOException
*/
- public static void executeDataMapJobForClearingDataMaps(CarbonTable
carbonTable)
+ private static void executeDataMapJobForClearingSegmentDataMaps(DataMapJob
dataMapJob,
+ CarbonTable carbonTable) throws IOException {
+ SegmentStatusManager.ValidAndInvalidSegmentsInfo
validAndInvalidSegmentsInfo =
+ getValidAndInvalidSegments(carbonTable,
FileFactory.getConfiguration());
+ List<String> invalidSegment = new ArrayList<>();
+ for (Segment segment : validAndInvalidSegmentsInfo.getInvalidSegments()) {
+ invalidSegment.add(segment.getSegmentNo());
+ }
+ DistributableDataMapFormat dataMapFormat = new
DistributableDataMapFormat(carbonTable,
+ validAndInvalidSegmentsInfo.getValidSegments(), invalidSegment, true,
+ DataMapLevel.FG);
+ dataMapJob.execute(dataMapFormat);
+ }
+
+ public static void executeClearDataMapJob(CarbonTable carbonTable, String
jobClassName)
throws IOException {
- String dataMapJobClassName =
"org.apache.carbondata.spark.rdd.SparkDataMapJob";
- DataMapJob dataMapJob = (DataMapJob) createDataMapJob(dataMapJobClassName);
+ DataMapJob dataMapJob = (DataMapJob) createDataMapJob(jobClassName);
if (dataMapJob == null) {
return;
}
- String className =
"org.apache.carbondata.core.datamap.DistributableDataMapFormat";
- SegmentStatusManager.ValidAndInvalidSegmentsInfo
validAndInvalidSegmentsInfo =
- getValidAndInvalidSegments(carbonTable,
FileFactory.getConfiguration());
- List<Segment> validSegments =
validAndInvalidSegmentsInfo.getValidSegments();
- List<Segment> invalidSegments =
validAndInvalidSegmentsInfo.getInvalidSegments();
- DataMapExprWrapper dataMapExprWrapper = null;
- if (DataMapStoreManager.getInstance().getAllDataMap(carbonTable).size() >
0) {
- DataMapChooser dataMapChooser = new DataMapChooser(carbonTable);
- dataMapExprWrapper = dataMapChooser.getAllDataMapsForClear(carbonTable);
- } else {
- return;
+ executeDataMapJobForClearingSegmentDataMaps(dataMapJob, carbonTable);
+ }
+
+ public static DataMapJob getEmbeddedJob() {
+ String className = "org.apache.carbondata.indexserver.EmbeddedDataMapJob";
+ DataMapJob dataMapJob = (DataMapJob)
DataMapUtil.createDataMapJob(className);
+ if (dataMapJob == null) {
+ throw new ExceptionInInitializerError("Unable to create
EmbeddedDataMapJob");
}
- DistributableDataMapFormat dataMapFormat =
- createDataMapJob(carbonTable, dataMapExprWrapper, validSegments,
invalidSegments, null,
- className, true);
- dataMapJob.execute(dataMapFormat, null);
+ return dataMapJob;
}
- private static DistributableDataMapFormat createDataMapJob(CarbonTable
carbonTable,
- DataMapExprWrapper dataMapExprWrapper, List<Segment> validsegments,
- List<Segment> invalidSegments, List<PartitionSpec> partitionsToPrune,
String clsName,
- boolean isJobToClearDataMaps) {
- try {
- Constructor<?> cons =
Class.forName(clsName).getDeclaredConstructors()[0];
- return (DistributableDataMapFormat) cons
- .newInstance(carbonTable, dataMapExprWrapper, validsegments,
invalidSegments,
- partitionsToPrune, isJobToClearDataMaps);
- } catch (Exception e) {
- throw new RuntimeException(e);
+ public static List<ExtendedBlocklet> executeDataMapJob(CarbonTable
carbonTable,
+ FilterResolverIntf resolver, Configuration configuration, DataMapJob
dataMapJob,
+ List<PartitionSpec> partitionsToPrune, List<Segment> validSegments,
+ List<Segment> invalidSegments, DataMapLevel level) throws IOException {
+ return executeDataMapJob(carbonTable, resolver, configuration,
dataMapJob,partitionsToPrune,
+ validSegments, invalidSegments, level, false);
+ }
+
+ /**
+ * Prune the segments from the already pruned blocklets.
+ */
+ public static void pruneSegments(List<Segment> segments,
List<ExtendedBlocklet> prunedBlocklets) {
+ Set<Segment> validSegments = new HashSet<>();
+ for (ExtendedBlocklet blocklet : prunedBlocklets) {
+ // Clear the old pruned index files if any present
+ blocklet.getSegment().getFilteredIndexShardNames().clear();
+ // Set the pruned index file to the segment
+ // for further pruning.
+ String shardName = CarbonTablePath.getShardName(blocklet.getFilePath());
+ blocklet.getSegment().setFilteredIndexShardName(shardName);
+ validSegments.add(blocklet.getSegment());
+ }
+ segments.retainAll(validSegments);
+ }
+
+ static List<ExtendedBlocklet> pruneDataMaps(CarbonTable table,
+ FilterResolverIntf filterResolverIntf, List<Segment> segmentsToLoad,
+ List<PartitionSpec> partitions, List<ExtendedBlocklet> blocklets) throws
IOException {
+ List<ExtendedBlocklet> cgDataMaps = pruneDataMaps(table,
filterResolverIntf, segmentsToLoad,
+ partitions, blocklets,
+ DataMapLevel.CG);
+ return pruneDataMaps(table, filterResolverIntf, segmentsToLoad,
+ partitions, cgDataMaps,
+ DataMapLevel.FG);
+ }
+
+ static List<ExtendedBlocklet> pruneDataMaps(CarbonTable table,
+ FilterResolverIntf filterResolverIntf, List<Segment> segmentsToLoad,
+ List<PartitionSpec> partitions, List<ExtendedBlocklet> blocklets,
DataMapLevel dataMapLevel)
+ throws IOException {
+ DataMapExprWrapper dataMapExprWrapper =
+ new DataMapChooser(table).chooseDataMap(dataMapLevel,
filterResolverIntf);
+ if (dataMapExprWrapper != null) {
+ List<ExtendedBlocklet> extendedBlocklets = new ArrayList<>();
+ // Prune segments from already pruned blocklets
+ for (DataMapDistributableWrapper wrapper : dataMapExprWrapper
+ .toDistributable(segmentsToLoad)) {
+ TableDataMap defaultDataMap = DataMapStoreManager.getInstance()
Review comment:
it is not defaultDataMap, please change the name
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
With regards,
Apache Git Services
