FAQ

How does Kinto compare to other solutions?

Before we started building our own data storage service, we took a look at what was already out there. Our initial intent was to use and possibly extend an existing community project rather than reinventing the wheel.

However, since none of the existing solutions we tried was a perfect fit for the problems we needed to solve, notably regarding fine-grained permissions, we started our own stack using the experience we gained from building Firefox Sync.

What follows is a comparison table showing how Kinto stacks up compared to some other projects in this space.

Project

Kinto

Parse Server

Firebase

CouchDB

Kuzzle

Remote-Storage

Hoodie

BrowserFS

Offline-first client

Fine-grained permissions

~

1

Easy query mechanism

2

3

Conflict resolution

4

Validation

Revision history

File storage

Batch/bulk operations

Changes stream

Pluggable authentication

5

Pluggable storage / cache

Self-hostable

Decentralised discovery

6

Open source

Language

Python

Node.js

Erlang

Node.js

Node.js 7

Node.js

Node.js

1

Currently, user plugin in Hoodie auto-approves users, but they are working on it.

2

CouchDB uses Map/Reduce as a query mechanism, which isn’t easy to understand for newcomers.

3

Remote Storage allows “ls” on a folder, but items are not sorted or paginated.

4

Kinto uses the same mechanisms as Remote storage for conflict handling.

5

Remote Storage supports OAuth2.0 implicit grant flow.

6

Support for decentralised discovery is planned but not implemented yet.

7

Remote Storage doesn’t define any default implementation (as it is a procol) but makes it easy to start with JavaScript and Node.js.

You can also read a longer explanation of our choices and motivations behind the creation of Kinto on our blog.

Why the name «Kinto»?

«Kinto-Un» is the name of the flying nimbus of San Goku. It is a small personal cloud, that flies at high speed and that you can share with pure heart riders :)

I am seeing an Exception error, what’s wrong?

Have a look at the Troubleshooting section to see what to do.

Can I encrypt my data?

Kinto server stores any data you pass to it, whether it’s encrypted or not. We believe encryption should always be done on the client-side, and we make it easy to use encryption in our Kinto.js client.

Is there a package for my Operating System?

No, but it’s a great idea. Maintaining packages for several platforms is time-consuming and we’re a small team.

Currently we make sure it’s easy to run with Docker or Python pip.

We also have a single-click deployment on some cloud providers.

Important

If you’d like to help us out by maintaining packages for your favourite OS, we’d be delighted to collaborate with you!

Why did you chose to use Python rather than X?

We love Python because it’s a concise & expressive language with powerful data structures & easy to learn, so it was an obvious choice for the development team.

In addition, the Operations team at Mozilla is comfortable with deploying and managing Python applications in production.

However, Python is just an implementation detail per se. Kinto is defined by an HTTP API that could be implemented in any language.

Is it Web Scale?

YES™. Have a look at the /dev/null backend. ;-)

Can I store files inside Kinto?

Yes, using the https://github.com/Kinto/kinto-attachment plugin.

I want to add business logic to Kinto!

By default, Kinto has no domain-specific logic. When we need some, we usually start by writing a plugin.

Plugins can hook in many parts of the API. Events subscribers are the most frequently used hooks, and allow you to perform extra checks or operations, or even raise HTTP exceptions if necessary. Plugins can also add new URLs to the API etc.

If you eventually hit a point where you need even more logic on the server side, you can build your own Kinto-esque service using the REST resources abstractions from kinto.core. In this way, your service will inherit all the best practices and conventions that Kinto itself has, and you can seamlessly migrate.

Maybe Kinto is not what you need after all, don’t hesitate to start a conversation!

How does Kinto authenticate users?

Kinto authentication system is pluggable and controlled from settings.

By default it ships with a very simple (but limited) Basic Authentication policy, which distinguishes users using the value provided in the header. In other words, any combination of user:password will be accepted. Kinto will encrypt them and determine a unique user id from them.

See also:

Note

We’d be delighted to add more built-in authentication methods into Kinto. Please reach out if you’re interested!

How to disable the permissions system (for development)?

By default, only the creator of the object has permission to write into it.

During development, it can be convenient to give the permission to write to any user.

Just create the bucket (or the collection) with system.Everyone in the write principals. For example, using httpie:

echo '{"permissions": {"write": ["system.Everyone"]}}' | \
    http PUT http://localhost:8888/v1/buckets/a-bucket --auth user:pass

If two users modify the same collection offline, how does that conflict get resolved?

When using concurrency control request headers, the conflicting operation will be rejected by the server.

The application developer can implement custom conflict resolution strategies, using the two versions of the object, or the history of actions of that object.

Some helpers are provided in the https://github.com/Kinto/kinto.js client. The three provided conflict resolution strategies are:

  • SERVER_WINS: local changes are overridden by remote ones ;

  • CLIENT_WINS: remote changes are overridden by local ones ;

  • MANUAL (default): handle them on your own.

Then there is, of course, a convenient helper to handle conflict one by one.

Would you recommend Redis or PostgreSQL?

You can use both of them:

  • Redis is usually easier to install and run than PostgreSQL. But you will have a database running in memory which means your data should be smaller than your server RAM. Redis is great for the cache backend.

  • PostgreSQL is the recommended backend for storage and permission in production. Mainly because data integrity is guaranteed, thanks to «per-request» transactions. It’s also usually easier to backup and export data out of a PostgreSQL database.

Why PostgreSQL to store arbitrary JSON?

Kinto backends are pluggable.

We provide an implementation for PostgreSQL that relies on JSONB (version >=9.4). It is very performant, allows sorting/filtering on arbitrary JSON fields, the eco-system is rich and strong, and above all it is a rock-solid standard.

If you prefer MongoDB, RethinkDB or X, don’t hesitate to start a storage, permission or cache backend, we’ll be delighted to give you a hand!

Why did you chose to use Pyramid rather than X?

Flask or Django Rest Framework could have been very good candidates!

We chose the Pyramid framework because we like its flexibility and extensibility. Plus, we could leverage Cornice helpers, which bring HTTP best practices out-of-the-box.

What about aggregation/reporting around data, is Kinto ready for that?

This is not available from the main API — and probably never will.

However, this is something that can be done aside or on top of Kinto.

For example, you could use ElasticSearch. There is tutorial for that!

Also, if you use PostgreSQL for storage, you can create custom views in the database that can be consumed for custom reporting.

Say I wanted to move all my Kinto data out of the database, would the best way to be via the backend?

It really depends on how you setup things, and what kind of data is there. One really simple way is to use the HTTP API. But depending the access you have to the user’s data, it might or might not be the solution you’re looking for. If you have access to the server, then doing a dump would get you the data out, but it won’t be in any documented format (it will be in an internal representation).

Nevertheless you can use the Kinto HTTP API to sync two databases.