1) The Knuth variance is an Impala built-in. Internally, Impala can handle aggregate functions with different intermediate and output types.   Basically, the only reason you are not allowed to create UDAs with different intermediate/output types is because we have not enabled the feature in sermantic analysis. For us, enabling the feature is the easy part. Adding extensive testing is the hard part.   If you are curious, the check for preventing you from creating such UDAs is in: ./fe/src/main/java/com/cloudera/impala/analysis/CreateUdaStmt.java lines 137 following   2) Like I said, enabling the feature is not hard, but does involve a non-trivial QA effort, so I cannot promise a concrete release at this point. I'd recommend keeping an eye on that JIRA for updates to the target version.     ... View more

Hi fil,   I'm afraid there is no way to force the encoding for a particular column. It is questionable whether you'd get any significant benefits from a manual choice.   Alex ... View more

Hi fil,   I'm afraid Impala currently does not take advantage of index pages when reading/writing data.   Index pages are part of the Parquet spec, so in theory, you shoud be able to write Parquet files with index pages (via some other tool) and have them be readable by Impala - but Impala will ignore the index pages.   We do plan on taking advantage of index pages and min/max/etc values, but we do not have a concrete target date for that feature yet.   Alex   ... View more

Looks like you are running into this issue: https://issues.cloudera.org/browse/IMPALA-2048   I'd suggest you give the workaround a try.   We've identified the issue and fixed it. ... View more