secondary index consistency guarantee

[金砖 <ji...@wacai.com>]

hi, everyone:
I read from document there are 3 levels of consistency.
http://phoenix.apache.org/secondary_indexing.html


In "Disallow table writes until mutable index is consistent", index will be rebuilt in the background in the event of a commit failure.


Does the rebuilt procedure will be performed incrementally , the same with "Disable mutable indexes on write failure until consistency restored" , and all configurations will be the same ?


Or index will be rebuilt from the beginning ? 


If the latter case, when index table is huge, whole cluster will be overload, and unavailable.

Re: secondary index consistency guarantee

[James Taylor <ja...@apache.org>]

Hello,
The index rebuild process is incremental, based on the time stamp at which
the index write failure originally occurred.
Thanks,
James

On Wed, May 31, 2017 at 2:22 AM 金砖 <ji...@wacai.com> wrote:

> hi, everyone:
>
>  I read from document there are 3 levels of consistency.
>
> http://phoenix.apache.org/secondary_indexing.html
>
> In "Disallow table writes until mutable index is consistent",  index will
> be rebuilt in the background in the event of a commit failure.
>
> Does the rebuilt procedure will be performed incrementally , the same with
> "Disable mutable indexes on write failure until consistency restored" , and
>  all configurations will be the same ?
>
> Or index will be rebuilt from the beginning ?
>
> If the latter case, when index table is huge, whole cluster will be
> overload, and unavailable.
>
>