What can I do to optimize overall rXg performance?
December 20, 2023
1. Collapse unused/redundant policies. If the enforcements are the same for multiple policies, consider collapsing those into one policy.
Infrastructure devices are a good example. Each time you add a new infrastructure device, a new policy is created. All of the infrastructure devices should be collapsed into a single management policy.
2. Size subnets to an appropriate size. Don't use /16s when a /24 will be more than enough.
3. Infrastructure/Non-Account-Group policies should be associated with a bandwidth queue that is per group and has per IP disabled.
To check how effective these changes are, you can look at the total number of PF Queues before and after making these changes in Instruments >> Traffic Rates >> PF Queues (Record Count at the bottom)
Infrastructure devices are a good example. Each time you add a new infrastructure device, a new policy is created. All of the infrastructure devices should be collapsed into a single management policy.
2. Size subnets to an appropriate size. Don't use /16s when a /24 will be more than enough.
3. Infrastructure/Non-Account-Group policies should be associated with a bandwidth queue that is per group and has per IP disabled.
To check how effective these changes are, you can look at the total number of PF Queues before and after making these changes in Instruments >> Traffic Rates >> PF Queues (Record Count at the bottom)
