Saturday, November 14, 2020

Network engineering continued ...

 This is a continuation of the earlier posts starting with this one: http://ravinote.blogspot.com/2020/09/best-practice-from-networking.html

  1. The virtual machines for the individual customers are sticky. Customers don’t usually release their resource and even identify it by their name or ip address for their daily work.  They host applications, services and automations on their virtual machines and often cannot let go of their virtual machine unless files and programs have a migration path to another compute resource. Typically, they do not take this step to create regular backups and keep moving the resource.  


  1. While container platforms for Platform-as-a-service (PaaS) have enabled software to be deployed without any recognition of the host and frequently rotated from one host to another, the end user's adoption of PaaS platform depend on the production readiness of the applications and services the force for PaaS adoption has made little or no changes to the use and proliferation of virtual machines by individual users 


  2. With the move towards serverless computing, networking has become somewhat more ubiquitous and taken for granted as hosts and resources disappear behind business functions. There are no more leases and IT resources required as compute is setup and torn down on demand automatically. The shift between deep partitioned modular cloud applications to serverless computing is a sliding scale and applications tend to leverage what's best for their business logic and not by available networks even for private cloud.

    Software defined networking has made it easy to setup and tear down networks at scale. Ip addresses and port numbers that were once sticky are no longer the case. When the deployments are on PaaS or on Kubernetes, the networking best practice is already leveraged enabling the applications to free up concerns on resources and to do more for the business.

No comments:

Post a Comment