Not too long ago I saw a post on twitter referencing Citrix ADC (Netscaler) sizing recommendations based upon the License edition. This wasn’t meant to ‘size’ the ADC to support a number of sessions, it was to notify you that based upon your license/bandwidth edition, there are CPU/Memory restraints. So even if you give the ADC more CPU, it won’t necessarily take advantage of that addition.

The Citrix article explaining this is located here: https://support.citrix.com/article/CTX139485

Notice that if you have a VPX10 or VPX200, they won’t use more than 2vCPUs. In order for the VPX1000 to use more than 2 vCPUs you need assign it at least 4GB of Memory.

Now what happens if we head over to the AWS Marketplace to deploy some ADC instances. In this case I chose a VPX200 Advanced edition, but the same settings come up no matter what bandwidth/edition.

As you can see the “*Vendor Recommended” instance size is the m4.xLarge instance, which if you hover over it, tells you it’s a 4vCPU x 16GB. Well that is certainly interesting… If we double check the chart above you’ll see that VPX200 can’t utilize more than 2vCPUs. Had we deployed this we’d be paying for an extra 2vCPUs that we can’t even use, costing you hundreds or even thousands of dollards if your deploying multiple instances.

A better option would be to choose the m4.Large instance size. This instance size half the price of the m4.xLarge.

In the Cloud world saving every little bit helps. This does apply to the on prem world as well. Be sure to also take a look at the Citrix ADC Data-Sheet (thanks @Mbp_Netscaler), which helps you understand the sizing as well (below screenshot).

Citrix Azure AWS ADC sizing

Do yourself a favor, don’t get duped… Double check you have the right instance sized deployed.

Leave a Reply

Your email address will not be published.