Gian Merlino

All tests passed except `standard-its / (Compile=openjdk8, Run=openjdk8, Cluster Build On K8s) ITNestedQueryPushDownTest integration test`, which has been weird in other PRs as well, and I don't believe would be...

Hi @fmarelli, It looks likely that there is a jackson version mismatch of some kind, perhaps between two different jackson jars. Have you updated the pom.xml at all? Is it...

Yeah, 0.10.1 would be better, I just haven't updated that project in a while :) You might need to downgrade Jackson even farther. Druid is built & tested using version...

Hi @Janesh this repository is just an example of how you might use internal unstable Druid APIs, and it isn't officially supported. If you are looking for something more stable...

Hi @Janesh, none of the query stuff in druid-processing is considered a stable API. They will work, but if you use them then you will often need to fix compile...

@Janesh happy to help

Hi @fmarelli, I agree in principle although currently we are focusing our efforts on supporting JSON/HTTP clients and on the SQL layer. If only we had infinite resources :) In...

fwiw, Druid currently (0.9.0) only supports the "query" dataSource type for "groupBy" queries, so it would probably be of limited usefulness to plywood as it does make other kinds of...

> > When decompressing a 64MB block using a 4KB chunk size, this leads to thousands of allocations and deallocations totaling GBs of memory. > > Is that a real...

> I agree with @wgtmac's concern about the expected size. For compression/decompression we are targeting the page size. The page size is limited by two configs, `parquet.page.size` and `parquet.page.row.count.limit`. (See...