VM-Series on Amazon Web Services Performance and Capacity
Many factors, such as AWS instance size,
maximum packets per second supported, number of cores used, and
AWS placement group, can affect performance.
We recommend additional testing within your environment
to ensure that you meet your performance and capacity requirements.
For a complete listing of all VM-Series features and capacities,
see compare VM-Series firewalls.
The performance and capacities listed in the following table have
been generated under these test conditions:
Recommended AWS Instance types (c5/m5/c5n/m5n) and sizes
that use the AWS Nitro Hypervisor with Enhanced Networking Adapter
(ENA). Additionally, SR-IOV and DPDK are enabled and AWS placement
groups are configured.
Firewall throughput is measured with App-ID™ technology features
enabled utilizing 64 KB HTTP 1.1 transactions.
IPsec VPN is measured with App-ID™ enabled and performance
is tested between a pair of VM-Series firewall instances in a placement
group deployed within the same availability zone and region. The
performance will vary based on AWS instance type and connectivity
topology (for example, connecting from on-premises hardware to VM-Series on
AWS, or from VM-Series in an AWS VPC to an AWS VGW in another VPC,
or VM-Series to VM-Series across regions).
Threat Prevention throughput is measured with App-ID, User-ID,
IPS, antivirus, and anti-spyware features enabled, utilizing 64
KB HTTP 1.1 transactions.
**Refers to recommended AWS instance size of a supported AWS
instance type based on CPU cores, memory, network interfaces and
pricing. For example, m5.xlarge instance (with 2 vCPUs, 16GB memory,
4ENIs at its price is recommended to support VM-300 model for a
range of common use cases.
***IPsec VPN Throughput is tested on the VM-100 with 1 tunnel,
VM-300 with 2 tunnels, VM-500 with 6 tunnels, and VM-700 with 12
tunnels.
