Jason Lowe

One consideration that recently popped up on the Spark side is differentiating errors that are recoverable by reattempting the task vs. errors that are fatal to the process. For example,...

@firestarman do you have the stacktrace where we're trying to partition an empty table? There's checks in the plugin where we have avoided trying to shuffle empty tables, curious what...

> The stacktrace that Mithun posted above should be somewhat relevant to the partitioning issue. Ah, thanks! I didn't look far enough back in the issue, missed that. > I'm...

Bummer, losing a prior CI status just for a headline change is really unfortunate. That may be worse, since changing headlines occurs more often than changing the PR base. Thinking...

Note that for the CHAR type, casting to a string requires stripping the trailing whitespaces from the value to match the CPU behavior. See https://github.com/NVIDIA/spark-rapids/pull/6188#discussion_r935683053. It would be nice if...

We fallback to the CPU for max width decimals in #6398. Lowering priority as the plugin now produces correct values for pmod, albeit via falling back for this corner case.

Note that if we end up needing to increase the default reserve memory to better handle complex regex kernels being launched with large thread stack space requirements, this may impact...

No, I don't think it would be necessary. If the regex kernels don't have an appreciably larger footprint than any other random libcudf kernel then I don't see the need...

@andygrove FYI I added #4511 to the list, since I think we need to improve the current situation where regex kernels can fail with a confusing OOM error due to...

I filed https://github.com/rapidsai/cudf/issues/10778 to request cuio functionality for this.