As most of us know, not all Cloud platforms are alike. You can be one of multiple installations sharing a service as co-tenant (software as a service). You may be purchasing Infrastructure as a Service, having full capacity of your installation. Even then, you most likely do not have sole access to a server; more likely you have a virtual machine sharing resources with other Virtual Machines.
When managing your own hardware it is important to maintain performance statistics, something we remind you about frequently. Those statistics let you know how the utilization of your server is progressing. That allows you to make an educated guess about future needs. It also allows you to evaluate what is going on when performance seems to drag.
It seems to me that these kinds of statistics are just as meaningful, or maybe even more so in a Cloud environment. In a cloud environment, the performance you get is often impacted by activities outside of your systems, because you are sharing hardware with other services. Unless you are sharing resources on a massively parallel scale such as Azure Blob Services, it is likely you will experience some impact by other users some time.
Are you getting what you pay for? Do you need to consider a higher cost option for Cloud services where higher levels of resources are guaranteed? How will you know if you don’t monitor your Cloud services?
What are your thoughts? Leave a comment, or drop an Email to btaylor@sswug.org.
Cheers,
Ben