Hi All, complete newbie, trying to learn the f5 lingo. We recently purchased a pair of i5000 series. Each physical box will be deployed in seperate geographical locations. I plan on carving each box up so that i have 2 Big IPs on each physical box at each location. Digging around trying to find configuration guidelines and best practices is a little tricky, when i'm not sure i have the right terminology :). Any help will be appreaciated.
i didn't migrate from this load balancer to F5 before, but you can check the below main points when you are migrating from any vendor to F5.
The configuration on F5 is very simple and straigh forward as a load balancer. It is divided into three main objects as below:
Good luck in your migration.
I would really advise getting a consultant in for something like this. If not for the whole project, then for doing some knowledge transfer and best practices and such. Sure people will help you here as you see, but still having someone around (in front or as backup) who has done it before it very valuable.
Absolutely agree, and yes we do have a consultant lined up.
I'm trying to be a little proactive and do some research on my own which will allow me to ask the right question and make appropriate decisions when the need arises.
It's hard to google stuff when you don't know what it's called... 🙂
Totally clear, Mohamed Salah helped you with some load balancing terms.
As for the carving up, that can be done with route domains and partitions or with vCMP. first is more like have separate routing instances, the second is more like running complete virtual machines. something to look into to have an idea what is most useful for you.
@DeenaFirst off I do not recommend doing this yourself if you have never configured an BIG-IP before because it's a lot of ground to cover. If you must configure it yourself you should be able to use the following article. One big piece to keep in mind is if you will be deploying your BIG-IP in path or out of path, by this I mean in order for your client to reach the BIG-IP and ultimately the servers does it have to go through the BIG-IP or no? If you deploy this out of path (one arm mode) this will require SNAT so if you have to preserve the client IP for the server to see that will no longer be possible.
Aside from that a gotcha after setup is making sure you record your master key otherwise your backups are useless if a failure ever occurs. I do not recommend carving out your BIG-IP into multiple route domains or partitions unless absolutely necessary because it adds unnecessary complexity, especially if you are new to BIG-IPs. The following are the sections of the BIG-IPs and a general description of what is configured in each section.
Local Traffic: This is where you configure load balancing
Network: This is where you configure all device IPs except the management interface IP, VLANs, and routing
System: This is where you configure almost everything for the management of the device such as SNMP, DNS, NTP, SSH, code upgrades, licensing, and so on.
This next article is a study guide for the BIG-IP but it might help you understand the device a bit better as well.