⚙️ Data Ops

The terminal, the scheduler, and the database that has to stay up — Linux glue, running PostgreSQL in production, and keeping pipelines reliable.

No lessons published yet — this is the roadmap, so you can see exactly what the course will cover before it lands.

Before you start: Learn SQL — Series 1. Series 2 is PostgreSQL under load — transactions, locks, MVCC. You need to be able to write and read the queries you are watching block each other.

Short on time? The 34 episodes marked core are the shortest path to being productive at work. Start there and pick up the rest whenever a job asks for them.

Series 1 · Linux for Data

do real data work from the terminal, with shell tools alone

01
Linux — Why Data Work Lives on This OS, and at a Prompt
corebeginnercoming soon
02
The File System — Paths, ls, cp, mv, find and Where Your Data Lives
corebeginnercoming soon
03
ps, df, systemctl & journalctl — Reading the Machine Your Job Runs On
corebeginnercoming soon
04
cat, less, head, tail & wc — Look at Data Without a Database
corebeginnercoming soon
05
Pipes & Redirection — Compose Tiny Tools into a Pipeline
corebeginnercoming soon
06
cut, grep, sort & uniq — Slice Columns and Rows on the Command Line
corebeginnercoming soon
07
sed & awk — Reshape CSV and TSV Text Streams
corebeginnercoming soon
08
ssh, scp & rsync — Move Data Between Machines
beginnercoming soon
09
Connection Failures — Host, Port, Firewall, Credentials and TLS
corebeginnercoming soon
10
Permissions, Users & Environment Variables — Secrets a Job Can Read
corebeginnercoming soon
11
Docker — Images, Containers, and the Commands You Actually Type
corebeginnercoming soon
12
VS Code — Remote-SSH, Dev Containers, and Editing Where the Data Is
corebeginnercoming soon
13
code-server — VS Code in a Browser, SSH FS, and AI Chat on the Box
beginnercoming soon
14
cron & systemd Timers — Turn a Script into a Nightly Job
corebeginnercoming soon
15
File Backups — tar, rsync Snapshots, Rotation, and Proving a Restore Works
corebeginnercoming soon
16
Git — Version Control for the Scripts and Config That Run Your Data
corebeginnercoming soon
17
Bash Glue — Variables, Loops, Exit Codes & Logging for ETL
corebeginnercoming soon
18
VPS & Cloud Compute — Getting a Linux Box, and Sizing It for Data
corebeginnercoming soon
19
S3 Object Storage — Backblaze, R2, rclone, and the Offsite Copy
corebeginnercoming soon
20
NGINX — One Public Door for Everything Listening on localhost
corebeginnercoming soon
21
Let's Encrypt — HTTPS in Ten Minutes, and the Renewal That Fails Silently
corebeginnercoming soon
22
Project — A Nightly Job Written Only With Shell Tools
corebeginner#projectcoming soon

Series 2 · Running PostgreSQL in Production

run PostgreSQL yourself — backups, locks, migrations, users

01
Transactions & ACID — BEGIN, COMMIT, ROLLBACK and What They Guarantee
coreintermediatecoming soon
02
Isolation Levels: The ANSI Four — Dirty Reads, Phantoms & What Postgres Really Does
coreintermediatecoming soon
03
MVCC, Bloat & VACUUM — Why Postgres Keeps Old Row Versions
coreintermediatecoming soon
04
Locks & Deadlocks — Why Your UPDATE Is Hanging
coreintermediatecoming soon
05
Connections & Pooling — PgBouncer and the Limit You Will Hit
coreintermediatecoming soon
06
Configuration & Memory — shared_buffers, work_mem and Defaults Worth Changing
intermediatecoming soon
07
Backup & Restore — pg_dump, PITR, and Proving It Actually Restores
coreintermediatecoming soon
08
Schema Migrations Without Downtime — Locks, Defaults & the Safe Order
coreintermediatecoming soon
09
Roles, Grants & Row-Level Security — Least Privilege for Data Access
coreintermediatecoming soon
10
Replication & Upgrades — Read Replicas, Failover and Major-Version Moves
intermediatecoming soon
11
Project — Migrate a Live Table Without Blocking Anyone
coreintermediate#projectcoming soon

Series 3 · Reliability

keep jobs running, and find out fast when one stops

01
Scheduling at Scale — Choosing Between cron, systemd Timers and an Orchestrator
coreadvancedcoming soon
02
Rundeck — What It Is, What Data Teams Use It For, and Running One in Docker
coreadvancedcoming soon
03
Jenkins — What It Is, Why Data Jobs End Up on It, and Running One in Docker
advancedcoming soon
04
Logging & Structured Output — Make a Failure Readable at 3 a.m.
coreadvancedcoming soon
05
Monitoring & Metrics — Watch Postgres, Disk and Job Health
advancedcoming soon
06
Alerting & On-Call — Page a Human Only When It Matters
advanced#google-srecoming soon
07
Freshness & Volume Checks — Catch a Pipeline That Silently Stopped
coreadvancedcoming soon
08
Incident Response & Postmortems — Diagnose, Fix, Write It Down
advanced#google-srecoming soon
09
Capacity, Storage & Retention — Disk, Partitions and TTL
advancedcoming soon
10
Cost Control — Compute, Storage and Where the Money Goes
advancedcoming soon
11
Project — Put a Northwind Pipeline On Call
coreadvanced#projectcoming soon

Where this leads

Continuous Delivery — starts from Series 1 · 43, and Series 3.
Kubernetes for Data Workloads — starts from Series 1, and Series 1 · 37 in particular.
Storage for Data — starts from Series 1.
Terraform & Pulumi — starts from Series 1, and Series 1 · 46 and Series 1 · 47 in particular.
Virtualization for Data — starts from Series 1.

Not here, on purpose

Writing and tuning the queries themselves — covered in Learn SQL, Series 3 — EXPLAIN, indexes and the query-tuning project
Row-level data quality, dedup & conformance — covered in Learn ETL & Data Pipelines — here we watch whether the pipeline ran at all
Schema design & normalization — covered in Learn Data Modeling; this track only migrates a schema safely
Kubernetes, Terraform and infrastructure as code — not covered; this track is the machine, the database and the job — the things a data person is actually handed. In this course, Series 1 · 46 is how you obtain one machine and size it for data work, and Series 1 · 47 is the object storage beside it. Both stop there
Python as a language for writing the jobs — covered in Python for Data. In this course, Series 1 is the shell tools you reach for when there is no Python on the box
Building the pipeline's logic — covered in ETL & Data Pipelines — here it already exists and has to keep running
Continuous integration, deployment pipelines and shipping a change safely — covered in Continuous Delivery, which owns the whole subject. In this course, Series 3 · 04 is the CI server used as a job runner, which is the opposite subject
Expressing a pipeline as a DAG of dependent steps — covered in ETL & Data Pipelines, Series 2 · 45, which is how a pipeline expresses its own dependencies. In this course, Series 3 · 00 is choosing what runs it and keeping that thing alive

⚙️ Data Ops

Series 1 · Linux for Data

Linux — Why Data Work Lives on This OS, and at a Prompt

The File System — Paths, ls, cp, mv, find and Where Your Data Lives

ps, df, systemctl & journalctl — Reading the Machine Your Job Runs On

cat, less, head, tail & wc — Look at Data Without a Database

Pipes & Redirection — Compose Tiny Tools into a Pipeline

cut, grep, sort & uniq — Slice Columns and Rows on the Command Line

sed & awk — Reshape CSV and TSV Text Streams

ssh, scp & rsync — Move Data Between Machines

Connection Failures — Host, Port, Firewall, Credentials and TLS

Permissions, Users & Environment Variables — Secrets a Job Can Read

Docker — Images, Containers, and the Commands You Actually Type

VS Code — Remote-SSH, Dev Containers, and Editing Where the Data Is

code-server — VS Code in a Browser, SSH FS, and AI Chat on the Box

cron & systemd Timers — Turn a Script into a Nightly Job

File Backups — tar, rsync Snapshots, Rotation, and Proving a Restore Works

Git — Version Control for the Scripts and Config That Run Your Data

Bash Glue — Variables, Loops, Exit Codes & Logging for ETL

VPS & Cloud Compute — Getting a Linux Box, and Sizing It for Data

S3 Object Storage — Backblaze, R2, rclone, and the Offsite Copy

NGINX — One Public Door for Everything Listening on localhost

Let's Encrypt — HTTPS in Ten Minutes, and the Renewal That Fails Silently

Project — A Nightly Job Written Only With Shell Tools

Series 2 · Running PostgreSQL in Production

Transactions & ACID — BEGIN, COMMIT, ROLLBACK and What They Guarantee

Isolation Levels: The ANSI Four — Dirty Reads, Phantoms & What Postgres Really Does

MVCC, Bloat & VACUUM — Why Postgres Keeps Old Row Versions

Locks & Deadlocks — Why Your UPDATE Is Hanging

Connections & Pooling — PgBouncer and the Limit You Will Hit

Configuration & Memory — shared_buffers, work_mem and Defaults Worth Changing

Backup & Restore — pg_dump, PITR, and Proving It Actually Restores

Schema Migrations Without Downtime — Locks, Defaults & the Safe Order

Roles, Grants & Row-Level Security — Least Privilege for Data Access

Replication & Upgrades — Read Replicas, Failover and Major-Version Moves

Project — Migrate a Live Table Without Blocking Anyone

Series 3 · Reliability

Scheduling at Scale — Choosing Between cron, systemd Timers and an Orchestrator

Rundeck — What It Is, What Data Teams Use It For, and Running One in Docker

Jenkins — What It Is, Why Data Jobs End Up on It, and Running One in Docker

Logging & Structured Output — Make a Failure Readable at 3 a.m.

Monitoring & Metrics — Watch Postgres, Disk and Job Health

Alerting & On-Call — Page a Human Only When It Matters

Freshness & Volume Checks — Catch a Pipeline That Silently Stopped

Incident Response & Postmortems — Diagnose, Fix, Write It Down

Capacity, Storage & Retention — Disk, Partitions and TTL

Cost Control — Compute, Storage and Where the Money Goes

Project — Put a Northwind Pipeline On Call

Where this leads

Not here, on purpose