Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Zookeeper >> mail # user >> is it possible to run multiple servers on the same machine and have data replication working?


Copy link to this message
-
Re: is it possible to run multiple servers on the same machine and have data replication working?
You also have to take care of the "clientPort".
You have to use different clientPorts per instance. Otherwise, the instance
can not start up because the port is in use.
/Hartmut

2012/9/1 Peter Litvak <[EMAIL PROTECTED]>

> Well, I actually had to create three server.x entries in the config since
> it wouldn't start with just two.
> Maybe the problem is in that even though I have three entries I'm starting
> only two server instances.
> So basically config has (as suggested by getting started guide):
> server.1=localhost:2887:3887
> server.2=localhost:2888:3888
> server.3=localhost:2889:3889
>
> I also have two different data dirs with myid (1 and 2).
> And then I start two servers using this config.
> I then have a program that starts one client connected to the first
> server which creates a node and immediately start the another client that
> connects to the second server and checks for an existence of that node and
> it fails.
>
> On Sat, Sep 1, 2012 at 4:51 PM, Jordan Zimmerman <
> [EMAIL PROTECTED]
> > wrote:
>
> > Make sure both instances are in both config files. Also make sure the
> > myid files exist.
> >
> > ===================> > Jordan Zimmerman
> >
> > On Sep 1, 2012, at 1:46 PM, Peter Litvak <[EMAIL PROTECTED]> wrote:
> >
> > > Hmm, then something else is wrong, I don't get data replication across
> > the
> > > servers. I'm wondering what could be the problem.
> > > On Sep 1, 2012 4:42 PM, "Hartmut Lang" <[EMAIL PROTECTED]> wrote:
> > >
> > >> Hi Peter,
> > >>
> > >> running multiple instances on the same machine is no problem.
> > >> You just have to use different config-files (zoo.cfg), and set
> different
> > >> data-dirs in them (dataDir=...).
> > >>
> > >> Hartmut
> > >>
> > >>
> > >> 2012/8/31 Patrick Hunt <[EMAIL PROTECTED]>
> > >>
> > >>> Take a look at this: https://github.com/phunt/zkconf If you run it
> > >>> with --count=x (see the examples) it will generate the proper
> configs,
> > >>> as well as a start script to run all locally.
> > >>>
> > >>> Patrick
> > >>>
> > >>> On Fri, Aug 31, 2012 at 1:05 PM, Peter Litvak <
> [EMAIL PROTECTED]>
> > >>> wrote:
> > >>>> Hi All,
> > >>>>
> > >>>> There is an explanation on how to run multiple instances of the
> > >> zookeeper
> > >>>> server on the same machine (in getting started) but that seems
> doesn't
> > >>>> allow data replication between those servers.
> > >>>> Is there a way to achieve a data replication among servers running
> on
> > >> the
> > >>>> same machine?
> > >>>>
> > >>>> I suspect that the answer is NO, just wanted to make sure.
> > >>>>
> > >>>> Thank you,
> > >>>> Peter
> > >>>
> > >>
> >
>