Another benefit of using Azure Infrastructure-as-a-Service is that it's simple to take advantage of Microsoft's advanced cloud networking services. When organizations need highly available VMs, load balancers come into play.
When organizations used to build server farms on premises, it was complicated to manage hardware load balancers to distribute traffic between multiple servers and to rebalance if one failed. Enabling’s engineers spent many long nights dealing with F5 and Cisco hardware load balancers when setting up and troubleshooting Skype and Lync server pools. Managing a load balancer took deep networking skills, especially with complicated platforms like F5 and Cisco.
When building a highly available servers in Azure, customers have two choices:
1) provision a virtual load balancer from a third party like F5, Kemp, or even Cisco
2) provision virtual load balancers provided by Microsoft
Microsoft offers a basic load balancer capability for free, and a standard load balancer service for a fee. Data about the differences is available at https://docs.microsoft.com/en-us/azure/load-balancer/load-balancer-overview?wt.mc_id=docsmsft-twitter
Discussion of load balancing and high visibility measure is part of Enabling’s IaaS Planning and Design Service. Virtual load balancers have been trending on premises too, so when moving to Azure, we haven’t seen customers object to using Microsoft's virtual capability.
Why are some organizations still electing to pay separately for a third party's load balancer within Azure? Sometimes they have an existing licensing contract which makes it advantageous to continue to use Kemp or F5, or in rare cases there is complicated hybrid setup where some active servers on-premises are load balanced with partner active servers that run in Azure.
To ask our engineers if Azure’s load balancers can meet your needs, contact firstname.lastname@example.org