Github user srowen commented on a diff in the pull request:
https://github.com/apache/spark/pull/4701#discussion_r25074705
--- Diff: core/src/main/scala/org/apache/spark/rdd/JdbcRDD.scala ---
@@ -64,8 +64,8 @@ class JdbcRDD[T: ClassTag](
// bounds are inclusive, hence the + 1 here and - 1 on end
val length = 1 + upperBound - lowerBound
(0 until numPartitions).map(i => {
- val start = lowerBound + ((i * length) / numPartitions).toLong
- val end = lowerBound + (((i + 1) * length) / numPartitions).toLong -
1
+ val start = lowerBound + ((BigDecimal(i) * length) /
numPartitions).toLong
--- End diff --
It does although the difference is +/-1 for all partitions but the last.
Yea I get your point about the final partition but isn't it possible that its
end is >= upperBound? I haven't thought it through, I admit. A version with
`BigInteger` here is definitely an improvement so I'd support that. I suppose
I'm still concerned this fails in exactly the corner case you're trying to fix
-- very large `length`.
---
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]
