Re: Suggestion for KeyRangeLookupTree

[Sivabalan <n....@gmail.com>]

sorry for the delay. I gave you contributor's permission.

On Mon, 20 Jun 2022 at 07:19, Unbelievable Miracle <ma...@gmail.com>
wrote:

> Hi~I had created Jira https://issues.apache.org/jira/browse/HUDI-4284. I
> am
> working on it now.
> Please give me the contributor permission, my JIRA ID : miracle  Thank you.
>
> Sivabalan <n....@gmail.com> 于2022年6月18日周六 03:28写道：
>
> > +1 sounds fair. Please abstract it out so that anything can be plugged
> in.
> > Btw, we are moving towards leveraging col stats and bloom filter in
> > metadata table for our index. So, this may not be required in the long
> > term. But should be useful for now. Please go ahead.
> >
> > On Sat, 11 Jun 2022 at 12:48, sagar sumit <sa...@gmail.com>
> wrote:
> >
> > > +1
> > > This is a good proposal.
> > > However, I wonder how likely it is to hit correctly ordered record keys
> > > with shuffling.
> > > Anyway, red-black tree would be useful in the worst case and I
> > > encourage you to create a patch for review.
> > >
> > > Regards,
> > > Sagar
> > >
> > > On Fri, Jun 10, 2022 at 9:20 PM 陆游 <ma...@gmail.com> wrote:
> > >
> > > > Hi Community，
> > > > I found that the existing KeyRangeLookupTree implementation is Binary
> > > > Sorting Tree.
> > > > Although it is shuffled before insertion, it may still cause uneven
> > > > distribution. I want to implement it as a Red Black Tree. What do you
> > > > think?
> > > > Thanks!
> > > >
> > >
> >
> >
> > --
> > Regards,
> > -Sivabalan
> >
>


-- 
Regards,
-Sivabalan