I'm checking now, but I think there was some kind of network problem during the time.
I was also asking for future reference =-) On Mon, Sep 7, 2015 at 4:42 PM, Ted Yu <[email protected]> wrote: > For the region with corrupt hfile, regardless how many times it is > attempted to be opened, failure to open would result, right ? > > > > > On Sep 7, 2015, at 1:36 PM, Hbase Janitor <[email protected]> > wrote: > > > > Is there any way to make region in transition try again if the initial > try > > failed? > > > > On Mon, Sep 7, 2015 at 3:22 PM, Hbase Janitor <[email protected]> > > wrote: > > > >> looks like I have a corrupt hfile, probably related to the replication > >> problems i'm having now. > >> > >> Thanks for the help! > >> > >>> On Mon, Sep 7, 2015 at 3:16 PM, Ted Yu <[email protected]> wrote: > >>> > >>> After timeout, the master would re-assign those regions. > >>> > >>> Can you check region server log to see why region open failed ? > >>> Was there any HFile which couldn't be opened ? > >>> > >>> Pastebinning log snippet if needed. > >>> > >>> Cheers > >>> > >>> On Mon, Sep 7, 2015 at 11:24 AM, Hbase Janitor <[email protected] > > > >>> wrote: > >>> > >>>> Hi, > >>>> > >>>> I have another problem, on my hbase 1.0 cluster I have a lot of > regions > >>>> that seem to be stuck in transition. I saw a exception attempting to > >>>> connect to zookeeper around the time they were hung. > >>>> > >>>> Is there a way to force them to retry? > >> > >> >
