Github user kayousterhout commented on a diff in the pull request:
https://github.com/apache/spark/pull/1309#discussion_r14695528
--- Diff: core/src/main/scala/org/apache/spark/Accumulators.scala ---
@@ -51,6 +51,13 @@ class Accumulable[R, T] (
Accumulators.register(this, true)
+ /** A name for this accumulator / accumulable for display in Spark's UI.
+ * Note that names must be unique within a SparkContext. */
+ def name: String = s"accumulator_$id"
--- End diff --
It seems weird to me that the ID and name both need to be unique. While I
realize that it's confusing and probably should be avoided to have two
accumulators with the same name, can you just index accumulators by their ID in
the UI to avoid the need for these to be unique? It just seems like this could
lead to hard to diagnose bugs where someone creates two accumulators with the
same name that appear as the same in the UI (even though in practice they're
different).
---
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 infrastruct...@apache.org or file a JIRA ticket
with INFRA.
---
