ProfileUnity Cluster Sizing Suggestions

With larger environments, it’s important to consider the implications of different architectures and system types when building a ProfileUnity Console cluster.  Some of the key things to note are how many geographic locations will be home to machines running the ProfileUnity Client Tools and what type of environment is being delivered.  For example, single-user Desktop OSes or multi-user Server OSes with the RDSH Server Role enabled.

Cluster sizing examples

VDI / Desktop-only environment:  (6.8.3 client-tools and newer only)

VDI Max Login Count** (The max number of users logging in over the busiest 15-minute timespan)	0 – 50,000 Logins	50,000 – 200,000 Logins
ProfileUnity Console Specs (Based on worst-case scenario, only 2 nodes up)	3 Nodes 2 vCPU 4-6GB RAM	3 Nodes 4 vCPU 6-8GB RAM

RDSH Server-only (i.e., XenApp) environment:  (Including all running VDI pre-6.8.3 client-tools)

Running RDSH Server Count	0 – 10,000 Machines	10,000 – 20,000 Machines
ProfileUnity Console Specs (Based on worst-case scenario, only 2 nodes up)	3 Nodes 2 vCPU 4-8GB RAM	3 Nodes 4 vCPU 8-16GB RAM

*All cluster sizing suggestions are based on the worst-case scenario of only 2 nodes in operation.

**The VDI suggestions aren’t based on the total user or machine count, but rather the maximum number of logins that might happen within a 15-minute window.  The safest bet is to use the max named or concurrent user count as the "worst-case login-storm" situation.

***Make sure if you have more than 7000 machines connecting that each console node has a system environment variable named ERL_MAX_PORTS set to 20000, then restart the RabbitMQ service.

For additional information, please see this Liquidware Community post.