Cloud and Linux

Cloud concepts 4 topics

• Entry check: are you ready for the cloud?
  A forty-five minute check of the Linux, ssh and networking skills this course builds on, with a pointer to what to revisit for each one.
• What the cloud is: regions, instances and images
  The building blocks of a cloud (instances, images, volumes, rules, names), regions and zones, types and images, what billing and stopping mean, shared responsibility…
• Who may do what: profiles, roles and tags
  Accounts, profiles and permissions, least privilege, roles for servers, tags for owners and cost, the audit trail, and two safety habits.
• Size and cost: the meter that runs while you sleep
  What a cloud bill measures, right-sizing, three pricing models, the simulated clock, the common leaks and a weekly cost habit, with every price illustrative.

Launching and connecting 4 topics

• Key pairs and the first connection
  Why keys replace passwords, making and importing a key pair, launching with a key, opening port 22 to your own address, host keys, the connect-info command and a table…
• Security groups: who may connect
  What a security group is, rules and sources, groups as sources, design rules, testing from outside, live changes, and a diagnosis path for it works for me but not for…
• The metadata service: a machine that knows itself
  The metadata address, what it answers and why, role credentials, the forged-request risk, session tokens, and why secrets never go in user-data.
• cloud-init: a server that builds itself
  What cloud-init does at first boot, cloud-config and script forms, a first file, launching with it, quiet partial failure, validation with the schema check, and when to…

Operating in the cloud 4 topics

• Volumes and snapshots: keeping the data
  Root and data volumes, attaching, making a file system and mounting, zones, snapshots and crash consistency, restoring to a new volume, growing in two steps, copies in…
• Images and the life of an instance
  Images and golden images, image versus cloud-init, keeping images honest, the life of an instance and its address, elastic addresses, protection, and replacing rather…
• Load balancers, health checks and DNS
  A load balancer with targets, port and health path, how the health check decides who gets traffic, reading cx lb health, draining a server for maintenance, and DNS…
• Watching: metrics, alarms, events and cost
  Reading metrics as trends, alarms that turn numbers into messages and what makes a good one, the event log as the audit trail (and one profile per person), and the cost…

Infrastructure as code 4 topics

• Infrastructure as code: the account in a file
  A stack file (stack, groups, volumes, instances, dns), validation with cx validate, errors for tabs, capitals and unknown keys, apply that makes the account match and is…
• Plan before apply: changes you can read
  Reading a plan (+, ~, -, !), changes that cause an outage or need a replacement, removal with --prune only for the stack's own resources, --dry-run, --detailed-exitcode…
• Drift, tags and clean-up
  Drift between the file and the account and how a plan shows it, tags for owner and environment, the leftovers that cost money for nothing (detached volumes, idle…
• Designing for failure: zones, regions and recovery
  The sizes of failure and their defences, spreading across zones behind a balancer, copying snapshots to another region, RPO and RTO with arithmetic, and rehearsing the…

Infrastructure behind the cloud 4 topics

• What a cloud is made of: hosts, networks and disks
  The parts of a cloud (regions and zones, the control plane, compute hosts and the hypervisor, the virtual network and security groups, block storage and images), a card…
• Identity, keys and the metadata path: who may do what
  The parts of identity in the cloud (profile, roles and permissions, key pair, security group, metadata service, event log), a card for each, and the path from a person…
• Describing and changing infrastructure: code, plans and drift
  The parts of infrastructure as code (stack file, version control, plan, apply, the account, tags), a card for each, and one change through the cycle of describe, review…
• Running at scale: load, names, watching and cost
  The running parts of a service (load balancer, DNS, metrics, alarms, event log, cost report), a card for each, one day of the booking service read as a single picture of…

Interview ready 2 topics

• Interview ready: accounts, instances, access and data
  The questions people really ask about regions and zones, instances, security groups, access, volumes and snapshots, and cost, with model answers, a routine for timed…
• Interview ready: code, load balancing and resilience
  The questions people really ask about infrastructure as code, plans, drift, load balancers, DNS, RPO and RTO and zone failure, with model answers, strong and weak…

The booking service in the cloud

A stack file that describes the booking service (security group, servers in two zones, tags, a DNS record), validated, planned and applied, with a load balancer and a health check, and the commands that prove it works.

The access and data pack

Profiles with least rights, a security group that allows ssh from the office only, a data volume with a snapshot schedule and a copy in another region, and a restore test.

The cloud map and handover

A card for each part, a failure design with RPO and RTO, a drift check, a cost report with owners for every resource and a one-page handover.

Why take this course

• Launch, secure and operate servers in the cloud
• Know what is in force, not just what you typed
• Keep the bill and the risk under control

You will build

• The booking service in the cloud
• An access and data pack
• A cloud map and handover

What you can do afterwards

• Run a small cloud account with confidence
• Explain your choices, their risks and their cost
• Answer cloud interview questions with specifics