What the cloud is

01The problem

Tallybook is a Lagos start-up whose app lets small businesses send invoices and get paid. It has no server room. Everything runs on a public cloud provider: the servers, the database, the file storage, the network.

The finance director has two worries. The cloud bill, paid in dollars, has grown every month, and nobody can explain exactly why. And at the end of August, on the busiest invoicing days of the month, the app slowed to a crawl for nearly an hour, twice. This course is the review Tallybook needs: what it's running, what it's paying for, what it's wasting, and what it would take to make it reliable.

02The concept

Cloud computing means renting computing resources from a provider, paying for what you use, usually by the hour or second, instead of buying hardware.

Service models

Regions and zones

A region is a geographic area (Tallybook uses Cape Town, the closest to Lagos). Each region has several availability zones: separate data centres with their own power and networking. Spreading across zones protects against one failing (lesson 7).

Shared responsibility

The provider secures the buildings, hardware and its own services. You are responsible for what you put on them: who has access, which data is public, software updates on your servers, and your settings. Most cloud breaches are customer misconfigurations, not provider failures.

Elasticity

The big promise: resources can grow and shrink with demand, and you stop paying when you stop using them. That only helps if you actually use it (lessons 4 to 6).

03Example

Tallybook's inventory of cloud resources:

import pandas as pd

base = "https://academy.cloudtechanalytics.com/datasets/cloud/"
resources = pd.read_csv(base + "resources.csv")

print(len(resources), "resources")
pd.crosstab(resources["type"], resources["status"], margins=True)

133 resources
status         attached  available  running  stopped  unattached  All
type
bucket                0          6        0        0           0    6
database              0          0        2        0           0    2
disk                 38          0        0        0           8   46
load_balancer         0          0        2        0           0    2
public_ip             0          0        0        0           3    3
snapshot              0         36        0        0           0   36
vm                    0          0       34        4           0   38
All                  38         42       38        4          11  133

Virtual machines (VMs) are servers; each has a disk attached. Snapshots are point-in-time copies of disks; buckets are file storage. Notice the statuses: some VMs are stopped, and some disks and public IP addresses are attached to nothing. Now look at the VMs by environment and size:

vms = resources[resources["type"] == "vm"]
vms.pivot_table(index="environment", columns="size", values="resource_id", aggfunc="count", fill_value=0)

size         large  medium  xlarge
environment
development      5      12       1
production       9       0       4
staging          0       6       0

Production is what customers use. Staging is a copy for testing releases; development machines are for engineers. Some VMs have no environment at all, which means nobody recorded what they're for.

04Walkthrough

1. Run the cells. List the VMs with no environment or no team. What do their names suggest?
2. For each resource type, decide which service model it belongs to.
3. Find the region of every resource. Why might Tallybook not use a region in Europe?
4. Read the shared responsibility table and list three things Tallybook is responsible for.

05Practice

06Check your understanding