SAP Hybris cluster : usage and limitations

What is a cluster?

In a computer system, a cluster is a group of servers (nodes) that act like a single system and enable high availability and, in some cases, load balancing and parallel processing.

Click here for more details

How to configure an SAP Hybris cluster?

SAP Hybris cluster can be configured in three ways:

+ JGroups communication toolkit
+ UDP multicast
+ UDP unicast

To stay simple, I have chosen the JGroups method!

Step 1: Activate the cluster mode in of each node
# enable cluster mode in the current node
clustermode =true
Step 2: Set JGroups as the communication’s protocol in the SAP Hybris cluster ( of each node)
# use jgroups as cluster communication method 
Step 3: Set cluster id

You can set cluster id of SAP Hybris nodes in two different ways :

  • Manually by setting in (different value per node)
# the identifier of the cluster (integer value)<cluster_id_integer>
  • Automatically by autodiscovery option (5.5 SAP Hybris version and above)
# enable auto discovery mode (introduced on the 5.5 version) 

By using the second option SAP Hybris will calculate the cluster id and save it under the table “CLNodeInfos”, so if you try the following request in HAC (Hybris Administration Console)

select * from CLNodeInfos

You will get something like that :


Step 4: Set the JGroups’s communication protocol
  • Set the JGroupsBroadcastMethod as the communication listener between nodes,
# The Jgroups' hybris implementation listener 

JGroupsBroadcastMethod class is the SAP Hybris’ implementation of JGroups listener

  • The SAP Hybris implementation of JGroups support two transport protocols (TCP, UDP)

for TCP, you should put the following lines in your

# <node_ip> : is the ip address of the SAP hybris node
# communication port, by defaut it is 7800 as set by SAP hybris

And for UDP (default mode)

# the SAP hybris default upd multicast port

The default jgroups-udp.xml / jgroups-tcp.xml paths are


you can set any path in the properties cluster.broadcast.method.jgroups.configuration under, for example

# The custom path of the jgroups-tcp configuration file 

Below an example of settings using Jgroups/TCP transport protocol/autodiscovery


How to configure an SAP Hybris cluster locally?

Make sure to unroll all the above steps and that the following tomcat setting was changed between nodes :

# http port 
# https port 
# ajp port 
# jmx port 
# jmx server port 

then run ‘ant server’ on each node under /bin/platform

ant server


– Once autodiscovery is activated, no way to control the interval ids (for example it is not possible to start autodiscovery from the id 5 and above)

– You can have duplicated node ids by using autodiscovery due to network communication issues

– By using autodiscovery, you can not exclude nodes from task execution since the id calculation is done dynamically

Tagged , ,

Leave a Reply

Your email address will not be published. Required fields are marked *