Map a workload onto IaaS, PaaS and SaaS and cost it both ways

short · 40 min · Objective 2.5

Task

Take one real workload -- a small company's file server and email -- and decide who manages each layer under IaaS, PaaS and SaaS, then compare five years of owning a server against renting the equivalent. The output is a responsibility matrix and a cost comparison, the two documents a move to the cloud is decided on.

Steps

  1. Write lab/cloud/responsibility.csv with header layer,on_premises,iaas,paas,saas, one row for each of facility, hardware, virtualisation, operating system, runtime, application, data and access, with each cell customer or provider.
  2. Record in lab/cloud/fit.txt which model fits the file server and which fits email, with a reason for each.
  3. Price a server able to run the file server: purchase price, support contract, power and a share of rack space, over five years. Record the items in lab/cloud/capex.csv with header item,year,cost.
  4. Price the equivalent IaaS virtual machine and storage for five years, including an estimate of data transferred out. Record the items in lab/cloud/opex.csv with the same header.
  5. Record in lab/cloud/decision.txt which is cheaper over five years, and one thing that would change the answer.

Verify

These checks run in a POSIX shell: Terminal on macOS or Linux, and on Windows Git Bash (it comes with Git for Windows) or WSL. A stock Windows PowerShell or Command Prompt has no awk or grep, so there the first line fails.

awk -F, 'NR>1 {n++} END {print n" layer(s)"}' lab/cloud/responsibility.csv
awk -F, '$1=="data" || $1=="access" {print $1": saas="$5}' lab/cloud/responsibility.csv
awk -F, 'NR>1 {s+=$3} END {print "capex total "s}' lab/cloud/capex.csv
awk -F, 'NR>1 {s+=$3} END {print "opex total "s}' lab/cloud/opex.csv
grep -Eic 'egress|transfer|utilisation|switch off|scale' lab/cloud/decision.txt

Eight layers are covered, and data and access remain the customer's even under SaaS -- the shared responsibility model in one line. The two totals give your comparison, and the decision should name what moves it: an always-on server usually favours owning, while a workload that can be switched off or scaled down favours renting.

Notes

Real comparisons also count staff time, which the cloud reduces for hardware and not at all for the operating system under IaaS, and the cost of the move itself.

This is an independent study companion for CompTIA Server+ SK0-005 and is not produced by or endorsed by CompTIA.