On Thu, Dec 21, 2017 at 10:33 AM, Mike Drob <[email protected]> wrote:
> I'm not super comfortable with that... > > If it has a spotty record like you suggest, then I don't want to be ironing > out issues with it so close to beta/release. There's already enough to iron > out and we're so close to the end that I don't want to risk destabilizing > at this point... > > Fair point. Let me undo it. I just came across this nice doc on region normalizer [1] by our brother Romil Choksi (Can we get this added to the refguide? I could copy/paste it if folks don't mind). Reading it, it seems like it is easy doing an enable and a one-off run so it is a low-barrier trying it. We need to talk this tool up when folks are in situation where they have ill-filled regions and want to merge adjacents. Meantime, try it a few times ... S 1. https://community.hortonworks.com/articles/54987/hbase-region-normalizer.html > Mike > > On Thu, Dec 21, 2017 at 11:50 AM, Stack <[email protected]> wrote: > > > There's been a request to enable region normalization by default. Sounds > > reasonable to me. Any objections? I think normalization has spotted > record > > so far. Enabling it we can try and iron out and issues with it before we > do > > the hbase 2.0.0 RC. If not possible, can disable before RC. > > > > Thanks, > > S > > > > On Mon, Dec 18, 2017 at 2:34 PM, Stack <[email protected]> wrote: > > > > > (I thought I'd already posted a DISCUSSION on defaults for 2.0.0 but > > can't > > > find it...) > > > > > > Dear All: > > > > > > I'm trying to get some eyeballs/thoughts on changes you'd like seen in > > > hbase defaults for hbase-2.0.0. We have a an ISSUE and some good > > discussion > > > already up at HBASE-19148. > > > > > > A good case is being made for enabling balancing by table as default. > > > > > > Guanghao Zhang has already put in place more sensible retry/timeout > > > numbers. > > > > > > Anything else we should change? Shout here or up on the issue. > > > > > > Thanks, > > > S > > > > > > > > > > > >
