ajantha-bhat commented on a change in pull request #3688: [CARBONDATA-3765]
Refactor Index Metadata for CG and FG Indexes
URL: https://github.com/apache/carbondata/pull/3688#discussion_r406066860
##########
File path:
integration/spark/src/main/scala/org/apache/spark/sql/execution/command/index/ShowIndexesCommand.scala
##########
@@ -54,69 +62,87 @@ case class ShowIndexesCommand(
override def processData(sparkSession: SparkSession): Seq[Row] = {
val carbonTable = CarbonEnv.getCarbonTable(dbNameOp,
tableName)(sparkSession)
setAuditTable(carbonTable)
- getFileIndexInfo(carbonTable) ++ getSIInfo(sparkSession, carbonTable)
- }
-
- // get info for 'index datamap'
- private def getFileIndexInfo(carbonTable: CarbonTable): Seq[Row] = {
- val indexes =
DataMapStoreManager.getInstance().getDataMapSchemasOfTable(carbonTable).asScala
- if (indexes != null && indexes.nonEmpty) {
- indexes.map { index =>
- Row(
- index.getDataMapName,
- index.getProviderName,
- index.getIndexColumns.mkString(","),
- index.getPropertiesAsString,
- index.getStatus.name(),
- index.getSyncStatus
- )
- }
- } else {
- Seq.empty
- }
+ getIndexInfo(sparkSession, carbonTable)
}
- // get info for SI
- private def getSIInfo(sparkSession: SparkSession, carbonTable: CarbonTable):
Seq[Row] = {
+ private def getIndexInfo(sparkSession: SparkSession, carbonTable:
CarbonTable): Seq[Row] = {
CarbonInternalMetastore.refreshIndexInfo(
carbonTable.getDatabaseName, tableName, carbonTable)(sparkSession)
- val indexesMap = CarbonInternalScalaUtil.getIndexesMap(carbonTable)
- if (null == indexesMap) {
- throw new Exception("Secondary index information is not loaded in main
table")
- }
- val indexTableMap = indexesMap.asScala
- if (indexTableMap.nonEmpty) {
- val indexList = indexTableMap.map { indexInfo =>
- try {
- val isSITableEnabled = sparkSession.sessionState.catalog
- .getTableMetadata(TableIdentifier(indexInfo._1,
dbNameOp)).storage.properties
- .getOrElse("isSITableEnabled", "true").equalsIgnoreCase("true")
- if (isSITableEnabled) {
- (indexInfo._1, indexInfo._2.asScala.mkString(","), "enabled")
- } else {
- (indexInfo._1, indexInfo._2.asScala.mkString(","), "disabled")
+ val indexesMap = CarbonIndexUtil.getIndexesMap(carbonTable)
+ if (null != indexesMap) {
+ val indexTableMap = indexesMap.asScala
+ if (indexTableMap.nonEmpty) {
+ val secondaryIndex =
indexTableMap.get(CarbonIndexProvider.SI.getIndexProviderName)
+ var finalIndexList: Seq[(String, String, String, String, String,
String)] = Seq.empty
+
+ if (secondaryIndex.isDefined && null != secondaryIndex.get) {
+ val siIterator = secondaryIndex.get.entrySet().iterator()
+ while (siIterator.hasNext) {
+ val indexInfo = siIterator.next()
+ try {
+ val isSITableEnabled = sparkSession.sessionState.catalog
+ .getTableMetadata(TableIdentifier(indexInfo.getKey,
dbNameOp)).storage.properties
+ .getOrElse("isSITableEnabled", "true").equalsIgnoreCase("true")
+ if (isSITableEnabled) {
+ finalIndexList = finalIndexList :+
+ (indexInfo.getKey, "carbondata",
indexInfo.getValue
+ .get(CarbonCommonConstants.INDEX_COLUMNS),
"NA", "enabled", "NA")
+ } else {
+ finalIndexList = finalIndexList :+
+ (indexInfo.getKey, "carbondata",
indexInfo.getValue
+ .get(CarbonCommonConstants
+ .INDEX_COLUMNS), "NA", "disabled", "NA")
+ }
+ } catch {
+ case ex: Exception =>
+ LOGGER.error(s"Access storage properties from hive failed for
index table: ${
+ indexInfo.getKey
+ }")
+ finalIndexList = finalIndexList :+
+ (indexInfo.getKey, "carbondata",
indexInfo.getValue
+ .get(CarbonCommonConstants.INDEX_COLUMNS),
"NA", "UNKNOWN", "NA")
+ }
+ }
+ }
+
+ indexesMap.asScala
+ .filter(map =>
!map._1.equalsIgnoreCase(CarbonIndexProvider.SI.getIndexProviderName))
Review comment:
No. Now you have map1(map2(map3)))
SI map - map2(map3)
Cgfgmap - map2(map3)
Instead of 3 level map it is two level map. It saves default map location
size of map1.
More over size, no need of lookup and code may be cleaner. You can check.
----------------------------------------------------------------
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:
[email protected]
With regards,
Apache Git Services
