Hey Joel,

Thanks much for the review! I have updated the KIP to address all your
comments. Please find the diff here
<https://cwiki.apache.org/confluence/pages/diffpagesbyversion.action?pageId=67638408&selectedPageVersions=26&selectedPageVersions=27>
.

Regards,
Dong

On Sat, Jun 3, 2017 at 11:43 AM, Joel Koshy <jjkosh...@gmail.com> wrote:

> +1
>
> Few additional comments (most of which we discussed offline):
>
>    -
>
>    This was summarized in the “discuss” thread, but it is worth recording
>    in the KIP itself that the LEO in DescribeDirsResponse is useful to
> measure
>    progress of the move.
>    -
>
>    num.replica.move.threads defaults to # log directories - perhaps note
>    that we typically expect a 1-1 mapping to disks for this to work well.
>    -
>
>    Can you clarify in the KIP whether intra.broker.throttled.rate is
>    per-broker or per-thread?
>    -
>
>    Reassignment JSON: can log_dirs be made optional? i.e., its absence
>    would mean “any”
>    -
>
>    Can you also explicitly state somewhere that “any” translates to
>    round-robin assignment today?
>
>
> On Mon, Apr 3, 2017 at 9:49 AM, Dong Lin <lindon...@gmail.com> wrote:
>
> > Hi all,
> >
> > It seems that there is no further concern with the KIP-113. We would like
> > to start the voting process. The KIP can be found at
> > *https://cwiki.apache.org/confluence/display/KAFKA/KIP-
> > 113%3A+Support+replicas+movement+between+log+directories
> > <https://cwiki.apache.org/confluence/display/KAFKA/KIP-
> > 113%3A+Support+replicas+movement+between+log+directories>.*
> >
> > Thanks,
> > Dong
> >
>

Reply via email to