hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michele Giusto <m.giu...@reply.it>
Subject Re: RegionServer continues lease expired
Date Tue, 10 Mar 2015 11:56:13 GMT
Nick Dimiduk <ndimiduk@...> writes:

> 
> In AWS you should use internal host names, not external host names. 
The
> reason being is that when a host resolves it's own name, it is 
resolved to
> the internal name, which means that's how it is referenced when 
entries are
> added to meta. If you use internal host names, /etc/hosts entries are 
not
> required.
> 
> -n
> 
> On Mon, Mar 9, 2015 at 2:37 AM, Michele Giusto <m.giusto@...> wrote:
> 
> > Hi Ted, hi Nick,
> > thanks for helping. I was not that confident that the problem was 
the
> > hostname, however that was it. The node was recognized with its FQDN
> > because when we added the new node we forgot to modify /etc/hosts
> > neither on the old nodes nor on the new one. Now I have fixed all
> > /etc/hosts and it is working.
> >
> > Any idea on the reason why we were getting that problem? At the end 
of
> > the day, all the nodes were communicating as expected (DNS is up,
> > /etc/hosts settings are probably superfluous in my case) and other
> > services (hdfs, map-reduce, Impala, ...) were not reporting problem.
> >
> >
> > Thanks again,
> > Michele
> >
> >
> >
> >
> >
> 

Hi Nick,
thanks for answering, we are using internal host names so things should 
have worked also before my edits to /etc/hosts, I suppose. However I 
have also found another wrong configuration on Impala and fixed it, now 
everything is working and probably the HBase problem was not the real 
reason behind my never ending queries.


Bye,
Michele




Mime
View raw message