Volumes and Persistent Storage

Overview

Persist data with named volumes, use bind mounts for live code, know when tmpfs fits, and back up a volume.

Container filesystems are ephemeral. Volumes survive container removal; bind mounts map host paths; tmpfs keeps data in memory.

This is a core tutorial in Module 7 · Volumes & Storage of the REBASH Academy Docker for Cloud & DevOps Engineers series — written for Cloud, DevOps, Platform, and SRE engineers.

Prerequisites

• Dockerfile Best Practices and Multi-Stage Builds

Learning Objectives

By the end of this tutorial, you will be able to:

• Create and mount a named volume
• Contrast volume vs bind mount
• Use tmpfs for scratch data
• Backup/restore a volume with a helper container

Architecture

This topic’s control points and relationships are shown below.

Theory

What

Containers are ephemeral by default: the writable layer disappears when the container is removed. Volumes, bind mounts, and tmpfs mounts persist or isolate data differently. Named volumes are managed by Docker; bind mounts map host paths; tmpfs keeps data in memory.

Why

Databases, queues, and CI caches need durable or shared storage. Bind-mounting source code enables live development. Choosing the wrong mount type causes permission errors, data loss on recreate, or accidental exposure of host files.

How it works

Declare mounts with docker run -v / --mount or Compose volumes:. Named volumes live under Docker’s volume directory and survive container recreation. Bind mounts use an absolute host path — ideal for code, risky for production config if paths differ. tmpfs is useful for scratch space and sensitive temporary files that must not touch disk. Volume drivers can back remote storage; local is the default. Permissions inside the container depend on user IDs (UID/GID) matching ownership on the volume.

Type	Use
Named volume	Databases, durable app data
Bind mount	Dev source, host configs
tmpfs	Secrets scratch, caches (lost on stop)

Key concepts

• Lifecycle — volumes persist after docker rm unless removed explicitly
• UID mapping — non-root containers often need chowned volumes
• Backup — treat volumes as stateful; plan snapshots
• Compose namespacing — project prefixes on volume names

Common pitfalls

• Storing production data only in the container writable layer
• Bind-mounting /var/run/docker.sock without understanding host takeover risk
• Permission denied after switching USER in the image
• Deleting volumes with docker volume prune without checking labels

Hands-on Lab

Objective

Persist data in a named Docker volume across container recreation, optionally mirror a host bind mount, and capture docker volume inspect evidence.

Prerequisites

• Docker Engine or Docker Desktop
• Write access under ~/rebash-docker/module-07/host-data

Lab environment

Workspace: ~/rebash-docker/module-07/host-data

Named volume rebash-mod07-data and bind path ./bind-data are created during the lab.

Terminal

mkdir -p ~/rebash-docker/module-07/host-data && cd ~/rebash-docker/module-07/host-data

Real-world scenario

A stateful sidecar writes cache files that must survive container upgrades. You prove a named volume keeps data after docker rm, add a bind mount for config files the platform team edits on the host, and attach inspect output for the change record.

Step-by-step tasks

Task 1 – Write to a named volume and recreate the container

Terminal

cd ~/rebash-docker/module-07/host-data
docker volume create rebash-mod07-data | tee volume-create.txt
docker run --rm --name rebash-mod07-writer -v rebash-mod07-data:/data alpine:3.20 \
  sh -c 'echo rebash-persist-v1 > /data/cache.txt && cat /data/cache.txt' | tee volume-write-v1.txt
docker run --rm -v rebash-mod07-data:/data alpine:3.20 cat /data/cache.txt | tee volume-read-after.txt
grep -q 'rebash-persist-v1' volume-read-after.txt

Expected output

Second container reads rebash-persist-v1 without the first container running.

Task 2 – Inspect the named volume

Terminal

cd ~/rebash-docker/module-07/host-data
docker volume inspect rebash-mod07-data --format 'Name={{ "{{" }}.Name{{ "}}" }} Mountpoint={{ "{{" }}.Mountpoint{{ "}}" }}' | tee volume-inspect.txt
grep -q 'Name=rebash-mod07-data' volume-inspect.txt

Expected output

volume-inspect.txt shows the volume name and host mountpoint path.

Task 3 – Bind mount host directory and read back

Create bind-data/config.txt:

config.txt

bind-mount-ok

Run with bind mount:

Terminal

cd ~/rebash-docker/module-07/host-data
docker run --rm -v "$PWD/bind-data:/config:ro" alpine:3.20 cat /config/config.txt | tee bind-read.txt
grep -q 'bind-mount-ok' bind-read.txt

Expected output

bind-read.txt prints bind-mount-ok from the host file.

Validation steps

• Data in rebash-mod07-data survives container removal
• volume-inspect.txt records name and mountpoint
• Bind-mounted config.txt is readable inside a container

Common errors and fixes

Error	Cause	Fix
Volume not found	Typo in volume name	docker volume ls and match rebash-mod07-data
cat: /data/cache.txt: No such file	Wrong mount path	Mount to /data consistently
Permission denied on bind mount	SELinux/AppArmor	On enforcing hosts use :Z suffix cautiously in lab only

Challenge exercise

Append a second line to the volume from a new container, then pack evidence:

Terminal

cd ~/rebash-docker/module-07/host-data
docker run --rm -v rebash-mod07-data:/data alpine:3.20 \
  sh -c 'echo rebash-persist-v2 >> /data/cache.txt && wc -l /data/cache.txt' | tee volume-write-v2.txt
grep -q '2' volume-write-v2.txt

Expected output

volume-write-v2.txt shows line count 2.

Learning outcomes

• Used a named volume to persist data across ephemeral containers
• Inspected volume metadata with docker volume inspect
• Contrasted named volumes with a read-only bind mount for host-managed config

Cleanup

Terminal

cd ~/rebash-docker/module-07/host-data
docker volume rm rebash-mod07-data 2>/dev/null || true
rm -rf bind-data 2>/dev/null || true

Validation

• Lab commands run under ~/rebash-docker/module-07/host-data/
• volume-read-after.txt and volume-inspect.txt prove persistence and inspect
• You can explain each Theory section in your own words
• You can describe one production failure mode for this topic

Code Walkthrough

Production practice for Volumes and Persistent Storage always combines:

1. Inspect before you change (status, plan, logs, dry-run)
2. Prefer reversible, documented changes (Git, IaC, drop-ins, version pins)
3. Capture evidence (command output, pipeline logs) for handovers
4. Prefer current tools and APIs over legacy shortcuts
5. Least privilege — escalate credentials only when required

Keep runbooks short enough to follow under pressure. Automate checks; keep humans for judgement.

Security Considerations

• Treat credentials and tokens for docker as privileged — never commit them
• Prefer short-lived auth (OIDC, roles, SSO) over long-lived keys
• Validate blast radius before apply/deploy/delete operations
• Restrict who can approve production changes
• Collect audit logs; limit who can read sensitive traces

Common Mistakes

Storing production data only in the container writable layer

Validate assumptions against the Theory section and official docs before changing production.

Bind-mounting /var/run/docker.sock without understanding host takeover risk

Lab shortcuts (open security groups, admin roles, skip approvals) must not ship unchanged.

Changing production without a rollback path

Always know how to revert (previous artefact, prior release, state rollback, DNS failback).

Best Practices

• Encode Volumes and Persistent Storage changes as code and review them in pull requests
• Pin versions (images, modules, actions, provider plugins)
• Separate environments with clear promotion gates
• Alert on symptoms with runbooks attached
• Destroy lab resources; tag everything with owner and expiry where possible

Troubleshooting

Symptom	Likely cause	Fix
Auth / permission denied	Wrong identity, policy, or scope	Check caller identity, roles, and least-privilege policies
Timeout / no route	Network, DNS, security group, or endpoint	Trace path, DNS, and allow-lists before retrying
Drift / unexpected plan	Manual change or wrong state/workspace	Reconcile desired vs actual; avoid click-ops on managed resources
Pipeline/job red	Flaky step, cache, or missing secret	Read failing step logs; bisect recent workflow/config changes
Cost spike	Idle load balancer, NAT, oversized compute	Inventory billable resources; stop/delete labs promptly

Summary

Volumes and Persistent Storage is essential for Cloud and DevOps engineers working with docker. Practise the lab until the inspection and change path is muscle memory, then continue the track.

Interview Questions

1. Named volume versus bind mount trade-offs?
2. What happens to a named volume on docker rm?
3. How do permissions problems show up with mounts?
4. Backup approach for volume data?
5. Security risks of bind-mounting docker.sock?

Sample answer — question 2

Confirm the mount in docker inspect and file paths inside the container.

Sample answer — question 4

Never mount docker.sock into untrusted containers.

Related Tutorials

• Course overview
• Docker Networking Fundamentals

References

• Manage data in Docker