Azure Networking Options — Core Cloud Services
Time to give Azure some attention.
First, let’s really think about why these — AWS, Azure — are the two biggest cloud services providers in the world:
They’ve been doing internet things for a long time.
Amazon launched in 1995, a virtual bookstore.
Microsoft, well, you know.
They’ve lived, breathed, and had a monumental impact on creating the infrastructure we use today,
that they’re selling to us today. Of course, the Store of Everything and the Company of Everything would encourage us to put everything in their hands.
Also: Azure has a lot less silly names for modules. Important. I appreciate straightforwardness.
I said ‘a lot less’, not ‘100% sensible names’
Microsoft has a clear set of Azure Fundamentals that anyone can interact with. Let’s talk about networking basics, basically to say, again, “Hi, employers, I have an interest in this and will ask questions to better my knowledge.”
- Virtual Networking
- Azure Load Balancer
- Traffic Manager
Virtual Networking is almost how we use it on our individual machines — It’s a logically isolated network that can be subnetted. Users interact directly with the web tier — the interface we see in a browser.
Create a Virtual Machine
You can manage a group of load balanced VMs with Scale Sets — When you need more VMs, they’re automatically made, and deleted in response to demand.
Azure Spot Instances offers unused resources at a discounted rate. All of these things are explained if you mouse over the Information toggle on the appropriate pages.
Not the greatest idea, but we can leave it up for testing. In a production environment, use the Advanced controls in the Networking tab to create rule limits.
Speaking of which:
I picked Azure Load Balancer for now, as it supports multiple traffic types and instances (Port forwarding, internal Azure traffic, ect). Just HTTP traffic can be used with Application Gateway.
Load Balancing is essentially, backups under the same IP address. A cloud version of HSRP (Hot Standby Routing Protocol) with Cisco routers.
If there are 3 Azure VMs, and one goes down, the others pick up the slack. If there’s a sudden influx of traffic, the load is distributed evenly among all three. The user does not care if VM A, B, or C is processing their request.
You can also set up extensions to install after the VM is set up. Less fiddling around with apt-get commands.
Because I’m in the Sandbox option of Azure Fundamentals Learning, I can’t push it to production. The module is supposed to be pushing a Wordpress app, but I take the time to explore instead, as I did that yesterday!
Traffic Manager means “We don’t want things to go slowly!”, which high availability doesn’t really solve. It works in conjunction with regions, directing traffic to the region closest to the user, to avoid slowdown.
Don’t Forget to check out my projects page for more Azure.
Originally published at https://www.runtcpip.com on December 17, 2019.
