Github user chenghao-intel commented on a diff in the pull request:
https://github.com/apache/spark/pull/11249#discussion_r53397085
--- Diff:
streaming/src/main/scala/org/apache/spark/streaming/StreamingContext.scala ---
@@ -269,6 +270,39 @@ class StreamingContext private[streaming] (
RDDOperationScope.withScope(sc, name, allowNesting = false,
ignoreParent = false)(body)
}
+ private[streaming] val recoverableAccuNameToId: mutable.Map[String,
Long] = {
+ if (isCheckpointPresent) {
+ // accumulators created by StreamingContext must provide name, so
it's safe to call name.get
+ mutable.Map(_cp.trackedAccs.map(accCP => (accCP.name, -1L)).toSeq:
_*)
+ } else {
+ mutable.Map.empty
+ }
+ }
+
+ /**
+ * Different from accumulator in SparkContext, it will first try to
recover from Checkpoint
+ * if it exist.
+ *
+ * @param initialValue initial value of accumulator. It will be
ignored when recovering from
+ * checkpoint
+ * @param name name is required as identity to find
corresponding accumulator.
+ */
+ def accumulator[T](initialValue: T, name: String)(implicit param:
AccumulatorParam[T])
--- End diff --
getOrCreateRecoverableAccumulator[T](createFunc: () => T, name:
String)(...)? Which also explicitly to tell developer that only accumulators
created via this API can be recoverable?
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
