[BUG] SplitAndRetryOOM query78 at 100TB with spark.executor.cores=64 #9204
Labels
bug
Something isn't working
reliability
Features to improve reliability or bugs that severly impact the reliability of the plugin
I am trying out some scaling tests with spark.executor.cores=64 and I ran into an SplitAndRetryOOM with NDS query78.
The place this is happening it looks to be the actual aggregation, as we should have concatenated tables before calling into it. The unfortunate thing is that we may need to split things up right after we concatenate them:
The text was updated successfully, but these errors were encountered: