lucene-solr-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Erick Erickson <>
Subject Re: SolrCloud
Date Mon, 02 Apr 2012 12:32:55 GMT
No, you don't have to run zookeeper on each replica. Zookeeper
is a repository for your system (cluster) information. It knows
about each replica, but ZK does not need to run on each shard.

You can run one zookeeper instance for your entire cluster, no matter
how many shards/replicas you have.

Here's a good place to get started understanding ZK:

Internally, SolrCloud uses ZooKeeper to understand what to do
with update and search requests. In effect, it asks ZK
"How many shards are there and what is the address of each
leader"? and "does the right thing" with the results...

My suggestion is that you pretty much forget ZK exists until
you get a bit more comfortable with SolrCloud. Run it embedded
in a single instance (and do NOT shut that instance down!).

>From there, you should see SolrCloud "just work" and it'll at least
get you started.


On Mon, Apr 2, 2012 at 1:59 AM, asia <> wrote:
> Thanks for replying,
> So if i will make a replica of each shard,then should I use zookeeper for
> every shards and replica or only for the replica.! more question i want to
> ask is that I am using solr in tomcat and eclipse environment using
> I am a bit confuse as to how to use zookeeper in it along with tomcat.I have
> downloaded zookeeper jar files also but need little help in it.
> -Asia
> --
> View this message in context:
> Sent from the Solr - User mailing list archive at

View raw message