Add support for ORC reader about velox HOT 12 OPEN

@yingsu00 I have started to understand the code. I am already up to speed with developer docs on Velox. But still wrapping my head around the implementation details.

In terms of time commitment, I should easily be able to spend ~4-5 hours daily going up to 10-12 hours/day if required. However, being new to the code base, I may need some guidance to speed up.

from velox.

@pedroerp Now I mainly use online use cases to do some performance tests. First, I will convert the format to parquet format. If the performance is good, I will develop orc format support, because online Hive tables are mainly in orc format.

from velox.

@yingsu00 I would like to attempt this and get it done. Still pretty new to the library but I hope to be able to contribute with guidance. Let me know if this is okay.

from velox.

@ankur4369 Thank you for offering help! ORC reader is crucial and performance critical. Some functionality may already be supported in DWRF reader but many are not. It will take several months to make it work and performant. How's your time commitment on it?

@mbasmanova @pedroerp What's Meta's plan on this? Will XLDB folks work on ORC reader?

from velox.

@ankur4369 It would be very nice if you could help add support for the ORC format. ORC is similar to DWRF and we have support for DWRF already. I hope it would be just a handful of tweaks to add support for ORC. We should be able to help by answering questions, brainstorming and reviewing the PRs. I'm CC-ing Zhenyuan and Orri who should be able to provide guidance as well.

I suggest to first figure out how to make ORC work without worrying too much about performance. Perf optimizations can come after basic support is in place.

CC: @zzhao0 @oerling

from velox.

@ankur4369 Is there any update on this?

from velox.

Is this still relevant? If so, what is blocking it? Is there anything you can do to help move it forward?

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs.

from velox.

@mbasmanova @pedroerp Does Velox have any plan to support ORC reader? I saw the above message, seems this feature didn't finish, right?

from velox.

@wpleonardo It would be nice to add support for ORC. Would you like to help build that?

from velox.

Re-opening since ORC support is still relevant.

from velox.

Hi, what's the status now?

from velox.

I don't know of anyone actively working on this. @wubiaoi, is this something you would be interested in contributing?

from velox.