anishshri-db commented on code in PR #43961:
URL: https://github.com/apache/spark/pull/43961#discussion_r1439533439
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDB.scala:
##########
@@ -604,17 +734,56 @@ class RocksDB(
}
private def getDBProperty(property: String): Long = {
- db.getProperty(property).toLong
+ if (useColumnFamilies) {
+ // get cumulative sum across all available column families
+ assert(!colFamilyNameToHandleMap.isEmpty)
+ colFamilyNameToHandleMap
+ .values
+ .map(handle => db.getProperty(handle, property).toLong)
+ .sum
+ } else {
+ db.getProperty(property).toLong
+ }
}
private def openDB(): Unit = {
assert(db == null)
- db = NativeRocksDB.open(dbOptions, workingDir.toString)
- logInfo(s"Opened DB with conf ${conf}")
+ if (useColumnFamilies) {
+ val colFamilies = NativeRocksDB.listColumnFamilies(dbOptions,
workingDir.toString)
+
+ var colFamilyDescriptors: Seq[ColumnFamilyDescriptor] =
Seq.empty[ColumnFamilyDescriptor]
+ // populate the list of available col family descriptors
+ colFamilies.asScala.toList.foreach { family =>
+ val descriptor = new ColumnFamilyDescriptor(family,
columnFamilyOptions)
+ colFamilyDescriptors = colFamilyDescriptors :+ descriptor
Review Comment:
Seems like this is linear since it relies/uses underlying immutable
collection impl. Switched to `ArrayBuffer` instead which should give us
amortized constant time for append
--
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.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
