There is sizing, and then there is right sizing

When it comes to firewall upgrades or refreshes, you need to know your:

Current traffic volumes and mixture
Session counts
Number of users
Current and future Internet or private connection pipes
Current and future throughput requirements (VPN, SSL Inspection and decryption, etc...)
Current and future unified threat management (UTM) needs
Interface requirements
Number of IP devices
LAG requirements
Future growth

I am sure you can come up with more variables and compare it to the vendors specification sheet in order to come up with an accurate size or model of a security appliance. However, nothing beats a real/live test with a traffic generator.

When I usually size a box based on the customer's requirements, I also factor in real life examples, previous firewall implementations and what the Internet/application traffic trends. I have heard so many "oh we are not going to use such and such application" or "we will never use SSL inspection".

So when your SE comes back with a bigger box, don't immediately shoot him/her down and think he is trying to up sale you.

Comments

MPLS vs VPN (Internet Connection) and power

This topic has been covered extensively by experts. What has not been covered in my opinion, is the underlying and fundamental change of transport infrastructure and specially power. The traditional WAN transport mechanisms are solid in terms of power normalization all through the last mile. With the new (or not so new) shift towards commercially available Internet connections (namely DSL and Cable), customers need to watch out for excessive power coming through those lines and the respective modems and into their edge devices. There are surge protectors out there with "ethernet in/out" ports which could be used to mitigate this problem. Happy conversation out there...

“If you want to make beautiful music, you must play the black and the white notes together.” ― Richard M. Nixon

Does your product integrate with other security products? At this point, you should hate product silos (point products) as much as I do. I understand and respect “divide and conquer” or “best of breed” strategies. I also understand having different security vendors at different layers of the network could possibly prevent an incident better (one vendor might not see/catch a vulnerability but another might have a signature or way of catching it). But isn’t it time to ask your vendors how and if they can work with each other? So what we (vendors) are competitors? If by integrating with each other, we are able to increase the return on investment (ROI) for the customer, then why not (I know it might sound naive and unaware, but could you just imagine ). Vendors have application programming interfaces (APIs) for interaction with other platforms. However, customers have to have application development resources to write code for these APIs, and that's not answ...

SD-WAN? Let's talk

I am sure at this point, you all have caught up on with this SD-WAN craze. Some early vendors are now bought by industry giants. Some have really cool cloud provisioning and management dashboards. Some claims that your edge has never been this simple to set up before. However, there are still a few concerns I have: VPN technology is a mature one by now. There are vendors out there who can offload IPsec tasks and processes onto a hardware based processor. Would these new "SD-WAN" providers give me the same throughput/granularity? How scalable is the solution? Applications - There are vendors out there who did WAN Optimization for living and have a very rich application signature database. What about the pure play vendors? Same goes for routing. can you handle complex routing? can you offload or optimize those touting decisions? What if I already have that infrastructure (VPN/Routing) in place? would your solution be a tunnel in ...

Systems Engineering Life

Search This Blog