On Sat, Jan 18, 2014 at 8:58 AM, Stack <[email protected]> wrote: > > On Fri, Jan 17, 2014 at 2:14 PM, Ted Dunning <[email protected]> > > wrote: > > > > > > > > That comment indicates a lack of understanding of ZK, not a bug in ZK. > > > > > > Mighty Ted Dunning, I'd just like to say that the quote is badly out of > context and misrepresents the clueful gentleman quoted (The comment > actually leads off with the root problem being HBase's USE of ZK, NOT ZK). >
Fair correction. I don't think I was the one who removed the context, but I should have been more sensitive to its lack. Corrections always welcome. Corrections from Stack pretty much always accepted without question. > >> >> Having the ability to know exact deltas would help make HBase region > >> >> assignment more robust. > > Nor am I sure what this is about. Those actually working on making HBase > region assignment more 'robust' have not asked for the above; i.e. have ZK > make fundamental changes in the way it operates, though it is working as > advertised, just to 'fix' a downstream project's misuse. > See previous general principle. This is much more compatible with what I remember of the strategies being used in the HBase Master.
