Running Strapi?

Strapi stuck on v3, plugins broken, or admin down? We bring it back.

Strapi gives you a self-hosted headless CMS that you own, which is a feature until an upgrade breaks half the plugins and nobody has time to fix them. Most of the Strapi work we handle is older v3 and v4 projects that need to move forward without losing the content models the editors have built around.

  • Since 2015Fixing technology
  • FreeEstimates
  • US teamReal engineers

strapi-start.log

still broken

What it is

Strapi at a glance

Strapi is an open-source, Node.js headless CMS first released in 2015 by the team that would become Strapi Inc. Unlike Contentful or Sanity, Strapi runs on your infrastructure, backed by your own Postgres, MySQL, SQLite or MariaDB, and the admin panel is served from the same application.

The project has shipped three major versions: v3 (2020), v4 (2021) with the entity service API and rewritten admin, and v5 (2024) with the Document Service API and TypeScript throughout. Each version bumped how plugins are built, which is the single biggest source of upgrade pain. Strapi offers a hosted Cloud product, though the majority of production installs are self-hosted on AWS, DigitalOcean, Render or Heroku.

Type
Self-hosted headless CMS, Node.js
First released
2015
License
MIT for the core, enterprise features under a paid license
Maintained by
Strapi, Inc. and an open-source community
Typically used for
Self-hosted content APIs, marketing sites, mobile backends, internal tools, teams who want to own the data layer

The honest version

Strapi pros and cons

No technology is all upside. This is where Strapi earns its place and where it tends to cause trouble.

Where Strapi is strong

  • You host it, you own itThe code and the database both live on your infrastructure. There is no vendor lock-in and no surprise price change on a Tuesday morning.
  • A content-type builder editors can useNew types and fields are created through the admin without a developer, which cuts a lot of back-and-forth on small content changes.
  • REST and GraphQL APIs ship togetherBoth APIs are generated from the schema, with per-role permissions, so the front end can pick whichever fits the use case without extra wiring.
  • A healthy plugin ecosystemSEO, i18n, documentation, Cloudinary and S3 upload providers, and dozens more are a yarn install away, which saves a lot of basic plumbing.
  • Database of your choicePostgres, MySQL, MariaDB or SQLite for development. The data is readable with ordinary SQL tools, which is a relief when you need to debug a problem.
  • Free to startThe core is MIT and runs on any Node host. Small sites can go live without a license fee, which lowers the bar for a proof of concept.

Where Strapi bites

  • Major upgrades hurtv3 to v4 required rewriting plugins and content model migrations. v4 to v5 is another plugin API break. Teams on old versions sit on them because the upgrade is real work.
  • You own the opsBackups, patching, admin builds, database tuning, uploads directory, Redis for sessions. All of it is on the team running the server, and it is more work than teams expect from a CMS.
  • Admin panel builds are heavyThe admin is a webpack build that regularly needs more memory than a small production instance has. OOM at build time is a recurring complaint.
  • Plugins vary in qualityCommunity plugins span from well-maintained to abandoned. Choosing one for a core workflow and then finding it stopped working on v5 is a familiar story.
  • Uploads directory needs careThe default on-disk uploads folder is fine until the instance is redeployed and the files vanish. Most production installs end up on S3 or Cloudinary, which is one more thing to configure.
  • Enterprise features are gatedSSO, audit logs, content history and review workflows live in the paid tier. Larger teams end up paying anyway, which blunts the free-and-open headline.

Sound familiar?

Common Strapi problems we fix

The Strapi problems that bring people to us, and what you have once each one is fixed.

running Strapi v3.6

Stuck on Strapi v3 or an early v4

A project built on v3 before the entity service existed, or an early v4 install with plugins that no one has touched in years. Security patches have long since moved on.

Upgraded onto a supported version

Plugin 'users-permissions' failed to load

Plugins broken after an upgrade

A plugin that assumed the v3 or early v4 API, a service renamed between minor versions, or a yarn upgrade that caught a peer dependency mid-stream. The admin refuses to start.

Plugins compatible, admin healthy

KnexTimeoutError: pool

Database connections exhaust the pool

A busy custom controller leaking connections, a Postgres instance sized smaller than the traffic needs, or a connection pool left at the default while the app has eight workers.

Pool sized to real load

admin build: out of memory

The admin panel runs out of memory to build

A production container with 512 MB trying to run webpack through the Strapi admin build. We pre-build the admin in CI and ship the output, which fixes it for good.

Admin built in CI, container slim

uploads vanished after redeploy

Uploaded images disappeared

The default local uploads provider on an ephemeral container filesystem. Every deploy wipes the directory, and editors slowly lose faith in the system.

Uploads on S3 or compatible storage

JWT_SECRET missing

Admin cannot start in production

The .env was never fully populated when the app was deployed from a developer's laptop, so required secrets are missing and the admin exits on startup.

Secrets managed, startup deterministic

schema out of sync

Content types and database disagree

A content type added or edited in one environment and the schema file not committed, leaving staging and production disagreeing about what the model looks like.

Schema in Git, environments aligned

no backup in 90 days

No tested backup of content or uploads

A Postgres dump that is scheduled but never restored, and uploads that live on a disk nobody has ever imaged. A restore drill reveals both problems at the worst possible time.

Backups tested on a fresh machine

What we do

Strapi help, from rescue to upkeep

Call us for one broken plugin, or hand over the whole installation.

Rescue
Admin refusing to start, plugins failing to load, uploads lost. We get the CMS serving editors again and document what broke.
Version upgrades
v3 to v4 and v4 to v5, with plugins ported, content model migrated, tests run against a copy of the database before the real cutover.
Hosting review
Picking between self-hosted on AWS, DigitalOcean or Render, Strapi Cloud or something in between, with the cost and operational trade-offs written down.
Database and uploads
Postgres sized and tuned, uploads moved to S3 or Cloudinary, backups scheduled and restored on a copy so they are known to work.
Custom plugins and workflows
Admin customisations, lifecycle hooks, and bespoke APIs built against the service layer rather than hacked into controllers.
Managed Tech
Patches tracked, backups rehearsed on a schedule, admin builds run in CI, and someone to pick up the phone when the Node process restarts in a loop.

How it works

From first message to fixed

  1. Look Free estimate

    Tell us what's going wrong. We look at it, then tell you plainly what's wrong and what it will cost.

  2. Fix

    We fix the cause, not the symptom. If something is down, getting it back up comes first.

  3. Harden

    Backups, monitoring, updates and access, so the same problem can't quietly come back.

  4. Hand back or stay on

    You get a plain-language write-up. Or we stay on for hosting, upkeep and new work.

Often found together

Running Strapi alongside something else?

We work on the whole stack, so the problem can be wherever it actually is.

Questions

Strapi questions

Something else? Call+1 737 727 7327

We are on Strapi v3, how do we upgrade?

There is no in-place path from v3 to v5. We stand up a v4 or v5 instance alongside the existing one, define the content types that match the old schema, write a migration that moves data through the APIs rather than through the database, and move the front end over once the parity is proven. v3 keeps serving until the cutover and then it retires.

Why does our Strapi admin crash on startup?

The common causes are a missing JWT_SECRET or APP_KEYS in the environment, a plugin that failed to load and took the app down with it, or a schema file on disk that disagrees with the database. The server logs usually name the culprit, and most of the fixes are one line in the .env or one plugin pinned to a known-good version.

Should we self-host Strapi or move to Strapi Cloud?

Self-host when you want to control the database, run inside a specific network, or save on the Cloud subscription at a scale where ops time is cheaper than hosting fees. Move to Cloud when nobody on the team wants to look after Node, Postgres and the admin build any longer. The right answer depends on what your team already does for a living.

Are our Strapi uploads safe?

Not if they are on the default local provider on an ephemeral container filesystem. Every redeploy or reboot wipes them. We switch the upload provider to S3 or Cloudinary, migrate the existing files once, and confirm new uploads land in the bucket before anything is deleted locally.

Should we pick Strapi, Contentful or Sanity?

Strapi when owning the data layer and the SQL database matters. Contentful when the editor group is large and roles and environments are part of the ask. Sanity when the Studio should feel like part of the application codebase. We will say which one matches your team in plain language before any migration starts.

How much does Strapi help cost?

The initial quote is free. Share the Strapi version and what you need done, and we come back with a scope and a number before the first commit lands.

Free estimate

Strapi giving you trouble?

Tell us what you're running and what's going wrong. A person reads it and comes back with an estimate.

  • No charge to find out what it will take
  • If it's urgent, say so

Replies come from a person. By sending, you authorize us to email you back. No marketing.