To set up clusters of engines, you store shared files on a system and set multiple
engines to use those files.
About this task
To set up blueprint design servers and engines in a clustered configuration, you install them on
separate systems and connect them to the same database and network storage. You can install as many
blueprint design servers and engines as you need. You can also install blueprint design servers and
engines as cold-standby servers to serve as disaster recovery units. You can install a blueprint
design server and an engine on each cluster node, or you can install blueprint design servers and
engines on separate nodes. If you install an engine and a blueprint designer on the same node,
ensure that your operating system is compatible with the engine.
You do not need to install the same number of blueprint design servers and engines. Most
high-availability setups need more engines than blueprint design servers.
Then, you configure a load balancer to distribute the traffic to the blueprint design servers.
Instead of accessing the blueprint design servers and engines directly, users access the load
balancer URL. To the users, that URL appears to host a single instance of the blueprint design
server or engine.
A complete high-availability system of blueprint designers and engines includes the following systems:
- One or more blueprint design server nodes
- One or more engine nodes
- A database for the engine nodes and the RabbitMQ service
- A database for the blueprint design server nodes
- A shared file system for the blueprint design server nodes
- A load balancer
You can install each of these systems on a separate server, or you can combine some of them to
reduce the number of overall nodes. For example, you can use the same database for the engine nodes
and the blueprint design server nodes. You can also install one or more of the databases, shared
file system, and load balancer on the same server. However, for greatest reliability, install the
load balancer, databases, and shared file system on separate servers. For example, the following
topology diagram includes three blueprint design server nodes, three engine nodes, and a single
server that hosts the databases, RabbitMQ, and shared file system.
Note: These instructions host
the RabbitMQ service, the databases for the engine and Keystone service on a single server, and the
shared file system on a single server. For a truly high-availability setup and to facilitate
disaster recovery, you must ensure that these systems are stable and reliable. For example, you can
configure automatic backups and have a cold standby system
ready.
You can set up a cluster of blueprint design servers and engines independently of the IBM UrbanCode Deploy server.
You do not necessarily have to have a cluster of IBM UrbanCode Deploy
servers.
What to do next
Add engines to the cluster. See
Adding engines to clusters.