ZooKeeper Observers

ZooKeeper Observers#

Observers: Scaling ZooKeeper Without Hurting Write Performance#

  • Observers are non-voting members of an ensemble which only hear the results of votes, not the agreement protocol that leads up to them.

  • Therefore they can fail, or be disconnected from the cluster, without harming the availability of the ZooKeeper service.

How to use Observers#

Firstly, in the config file of every node that is to be an Observer, you must place this line:

peerType=observer

Secondly, in every server config file, you must add :observer to the server definition line of each Observer. For example:

server.1:localhost:2181:3181:observer

This tells every other server that server.1 is an Observer, and that they should not expect it to vote.

$ bin/zkCli.sh -server localhost:2181

Example use cases#

  • As a datacenter bridge

  • As a link to a message bus

Tips#

zookeeper-3.4.10-observer/
├── conf/
│    ├── java.env
│    ├── zoo-2181.cfg
│    ├── zoo-2182.cfg
│    ├── zoo-2183.cfg
│    └── zoo-2188.cfg
├── data-2181/
│    └── myid
├── data-2182/
│    └── myid
├── data-2183/
│    └── myid
├── data-2188/
│    └── myid
├── logs-2181
├── logs-2182
├── logs-2183
└── logs-2188

cd /usr/local/zookeeper-3.4.10-observer
bin/zkServer.sh start conf/zoo-2181.cfg
bin/zkServer.sh start conf/zoo-2182.cfg
bin/zkServer.sh start conf/zoo-2183.cfg
bin/zkServer.sh start conf/zoo-2188.cfg
bin/zkServer.sh status conf/zoo-2181.cfg
bin/zkServer.sh status conf/zoo-2182.cfg
bin/zkServer.sh status conf/zoo-2183.cfg
bin/zkServer.sh status conf/zoo-2188.cfg
bin/zkCli.sh -server localhost:2188

References#

Contents