Clustering Requirements
General Considerations
Databases should be set up by a database administrator (DBA) who has previous experience with relational databases in the HA configuration (such as, PostgreSQL, Microsoft SQL Server, and InfluxDB). The required knowledge includes installation, optimization, and high availability clustering.
Additional performance tuning might be necessary in a production environment.
You may need to edit the commands and settings for optimal performance in a production environment. The installation and configuration steps outlined are examples and are intended for a QA or sandbox environment only.
All failover configurations must be fully tested and validated before being used in production.
Supported Operating Systems
For information about the ThingWorx Foundation version to implement, see System Requirements.
Virtual IP Addresses
Virtual IP addresses are required for the following:
Users and assets to ThingWorx Foundation and Connection Servers
ThingWorx Foundation to databases
Hardware Requirements
Complete hardware redundancy is to be used in a ThingWorx HA configuration.
Each instance of an application should be running on separate hardware to avoid single points of failure at the hardware level. For example, ThingWorx servers (physical, virtual, or cloud-based) should not operate on the same physical hardware.
This requirement is expected for all applications in the ThingWorx HA configuration (ThingWorx Foundation, PostgreSQL, ZooKeeper, and Ignite) to mitigate the risk of hardware failures.
Redundant routers, switches, and power supplies are assumed.
PostgreSQL Prerequisites
Pgpool-II and PostgreSQL DB is installed in RHEL or Ubuntu environments.
Two or more DB host servers are running a supported version of PostgreSQL, whereas three is recommended.
Two servers are running Pgpool-II 3.7.<latest> with watchdog configured. However, you can use other HA configurations that do not use Pgpool-II, such as PostgreSQL on VMs.
Microsoft SQL Server Prerequisites
Two or more DB host servers are running a supported version of Microsoft SQL Server.
Microsoft SQL Server is configured to operate through one of the following Microsoft HA methodologies:
Always on failover cluster instances
Always on availability groups
InfluxDB Prerequisites
Two or more meta nodes, whereas three is recommended for most use cases.
Two or more data nodes. It is recommended to have an even number of data nodes.
Was this helpful?