> Most relational database management systems do not support nested records, so tables are in first normal form by default. In particular, SQL does not have any facilities for creating or exploiting nested tables. [0]
“Not with that attitude.”
– frectonz
Ok, hear me out: what if we make something that takes a postgres database dir, tars it together and encodes it as a binary blob in SQLite?
We could have SQLite within postgres within sqlite within postgres! Is it practical or even slightly useful? Of course not - but it's SQL databases all the way down. Not that this is a good thing in itself.
No need for `tar`, there is "SQLite Archive" https://www.sqlite.org/sqlar.html
Take it one step further, the table-oriented database(tm) , embed clickhouse, MongoDB, Redis and PostgreSQL to ensure you have more flexibility than anyone can utilize efficiently. The one database to rule them all.
“We need to go deeper” (https://i.kym-cdn.com/photos/images/newsfeed/000/384/176/d2f...)
My next project
I think SQLite columns for SQLite would be superior to SQLite’s JSON columns whose operators are a whole ‘nother query language you need to learn and seem comparatively limited.
Agreed, the JSON search queries in Postgres are esoteric, to say the least.
But after spending some time with a mixed-schema table at even modest scale, I’m wondering how often a better design could have cut the whole problem off.
With Claude you barely had to learn the language this days as you just need to prompt, but SQLite column is an interesting idea.
I spent some time learning those queries for a project and when I grokked it they where very handy.
wouldn't that just be a foreign key to another table or, a list of keys or am i missing something?
What are the use cases for this? I can't imagine designing a database schemas to use this in a typical product. Is it intended for hybrid applications to back up local user data directly with their account info?
I can think of plenty.
The most interesting one for me is if you're running a SaaS product like Notion where your users create custom applications that manage their own small schema-based data tables.
Letting users create full custom PostgreSQL tables can get complex - do you want to manage tens of thousands of weird custom tables in a PostgreSQL schema somewhere?
I'd much rather manage tens of thousands of rows in a table where one of the columns is a BLOB with a little SQLite database in it.
> Letting users create full custom PostgreSQL tables can get complex - do you want to manage tens of thousands of weird custom tables in a PostgreSQL schema somewhere?
Yea, I'd be fine with that - postgres has the concept of databases and schemas within those databases. If you really want to build a product like that I'd suggest starting with per-tenant schemas that leverage table inheritance as appropriate. The permissions would be pretty easy to manage.
Though, in a lot of cases I've actually seen this done every client ends up with a dedicated server (or container - whatever tech you use to do it, something completely isolated from other instances) because user version management ends up being a huge issue. When you're building something that custom it's highly likely that version migrations need to be done with client oversight to ensure everything actually works.
I have yet to find an actual real world case where the inner-platform effect is the right solution. Usually when tools like that are selected the software ends up being so generic and flexible that's it's useless. Custom application/BI environment development relies on really judiciously telling users they can't have most features - with the hard part being figuring out which features are necessary and which ones you can cut to reduce bloat.
Notion has 100 million users, managing schema-per-tenant at our scale sounds like a complexity nightmare. We have 480+ identical schemas across 100+ Postgres hosts, and that already takes a lot of brainpower & engineering time to manage T_T
> managing schema-per-tenant at our scale sounds like a complexity nightmare.
The per-tenant schema could be the tenant's responsibility. Most non-technical users can handle the idea of tables & columns, assuming you leverage UI/UX patterns they are already familiar with.
Our UI looks like a table: https://www.notion.so/help/intro-to-databases
As long as we never add new features, never need to change how we map UI <-> Postgres DDL, and our users never make any mistakes when they change their tables, it could work without being a complexity nightmare
You manage a fleet of devices that need to get operating parameters regularly, but they're complicated and SQLite is a great mechanism for sending that.
So at the backend you have a postgres database that contains the device details etc as well as the operating parameters for that device.
You can update the operating parameters as part of a postgres transaction so either all the BLOBs are updated, or none.
Using /tmp on the postgres cluster (server) is a bit of a hack, it would be nicer to have memory based SQLite blobs.
In terms of security, you get Postgres row level security, so each SQLite value is protected in the same way as the rest of the row.
I’m thinking maybe you’d like to use litefs for multi-tenant dbs close to the tenant. But perhaps you’ll want a centralized billing/reports database under postgres as well.
So, instead of saving the client sqlite db of the org to cloud storage you save it to the centralized db column instead. Litefs probably doesn’t support it yet, but wouldn’t be too hard to add.
The top line of the README says: "Embed an SQLite database in your PostgreSQL table. AKA multitenancy has been solved."
But I'm still having trouble trying to grok the intricacies of it. In a sense, I guess it has well isolated individual SQLite DBs and you'd have to go out of your way to join over them. With that said, does PostgreSQL manage and pool all the writes correctly? So you don't need to worry about SQLite concurrency issues?
If by solving multinenancy they mean
CREATE TABLE tenants (
id BIGINT NOT NULL,
database SQLITE DEFAULT execute_sqlite(
empty_sqlite(),
'CREATE TABLE users (etc.)'
and all the other tables
for each tenant
)
);
Your other concerns are very real. Those sqlite dbs could become very large. I prefer the use case depicted in another reply: preparing sqlite dbs before shipping them to their own devices. Or maybe receiving them and performing analysis, maybe after having imported it in overall psql tables. Or similar scenarios in which all the db is read or written at once. Anyway, once we have a tool we start using it.
> then they don't need to make joins between sqlite dbs.
The extension could also provide custom index access methods (given that SQLite only supports a handful of column types in the first place.) That would allow you to incorporate the keys in the index heaps, as opposed to table heaps, boom, you get bitmap index scans for Joins, i.e. GIN but with a bit more redundancy.
Each of the columns is an instance of a SQLite database, so I assume (without looking at the source) that they properly multi-thread as needed.
So there's not cross-SQLite-database connections or multiple writers going on.
You could join over them, but not really in the way you're thinking.
Each of the columns that are databases would be updated when the functions execute.
You could do weird crap like INSERT/DELETE as part of a postgres level SELECT.
You may want to use the $$ way to declare strings for your examples. Something like:
-- Create a todo for "frectonz"
UPDATE people
SET database = execute_sqlite(
database,
$sqlite$INSERT INTO todos VALUES ('solve multitenancy')$sqlite$
WHERE name = 'frectonz';
oh nice, i didn't know this existed, thanks
The string between the dollar signs can only be closed by another set of dollar quotes with the same string between them. So it allows you to do quotes within quotes within quotes if necessary.
They /tmp file mechanism sounds like a bit of a hack, is that definitely necessary?
It may be possible to create a SQLite in-memory database instead and then load the binary blob data into it using the backup API or some kind of trick with VACUUM INTO.
I think the right approach would be to store the sqlite database as a varlena type that can be TOASTed and then "expanded" using the Expanded Datum API so that it's a live open database connection for the life of the transaction:
https://www.postgresql.org/docs/17/xtypes.html#XTYPES-TOAST
https://github.com/postgres/postgres/blob/master/src/include...
Thanks i will look into this more, /tmp stuff is most definitely a hack.
Someone, somewhere will eventually find a legitimate use case for it.
"Introducing pglite-fusion: Embed SQLite Databases Directly Into PostgreSQL Tables
What is pglite-fusion? pglite-fusion is a powerful PostgreSQL extension that bridges two popular database worlds by allowing you to embed SQLite databases directly into PostgreSQL tables. Each row can contain its own SQLite database using the new `SQLITE` column type—perfect for scenarios where you need highly localized or structured data at the row level.
How It Works - SQLITE Type: Define columns in PostgreSQL that hold embedded SQLite databases, stored as CBOR-encoded `Vec<u8>` objects. - Querying SQLite: Use the `query_sqlite` function to run SQL queries on the embedded SQLite databases. Results are returned as JSON-encoded arrays in a single row. - Updating SQLite: Use `execute_sqlite` to modify the SQLite database and retrieve updated instances.
Why You'll Love It - Seamlessly blend the simplicity of SQLite with the scalability of PostgreSQL. - Unlock new use cases like row-specific data isolation, nested data structures, or lightweight experimentation with different SQLite schemas. - Built on Rust using the pgrx framework for safety and performance.
Under the Hood pglite-fusion temporarily stores SQLite databases as files in `/tmp` to perform operations. SQLite processes these files to execute queries or updates and passes the results back into PostgreSQL.
Who Is It For? Developers and DBAs who want fine-grained control over data storage, hybrid database use cases, or just love the flexibility of combining PostgreSQL and SQLite in innovative ways.
This is likely great for serving SQLite data to frontends using the SQLite http-vfs.
Would be great combined with functions/triggers/views to mirror specific data/queries from Postgres as SQLite.
1NF crime against humanity?
I’m trying to think through when I’d reach for this over jsonb… I guess the fact that there’s an enforced schema? And that you could do aggregations on your SQLite db? Or maybe if you wanted to send the whole delete db to a client??
> enforced schema
I have bad news for you [0] about SQLite’s view on schema consistency.
Yo, dawg, I heard you like databases...
This is nuts. I can't think of a use for it, but I'm sure it's "a solution that will eventually find a problem".
That’s fun, but I think I'll just use an SQLite foreign data wrapper instead.
That's the sensible option but not the fun one.
Then you need one wrapper per database, with this approach you can have one database per row.
But can I have one row that holds all the databases?
yes you can
CREATE TABLE crime_against_humanity ( databases SQLITE[] );
Speed, anyone?
How long does it take to update a table of, say, 1k rows? 1m rows? Same when subqueries and joins are involved to calculate what's to be updated?
The current implementation is writing out the DB to `/tmp` then reading the resulting file back and writing it to the column.
So on the bright side updating 1k rows takes the same amount of time as updating one row. On the other hand every write is a full table write (actually two).
I don't think there is a way to do this efficently with the current API as PostgreSQL is MVCC so it needs to write out each version separately (unless it has some sort of support of partial string sharing, I don't think so). Maybe a better version of this would write every page of the SQLite DB as a separate row so that you only need to update the changed pages.
A query on very ugly tables will count 26 thousands occurrences of the apostrophe, I love it
Are there still reasons to use PostgreSQL?
I like the simplicity of SQLite's "a file is all you need" approach so much, that I started to converge all my projects to SQLite. So far, I have not come across any roadblocks.
Can anyone think of a use case where PostgreSQL is better suited than SQLite?
Sometimes you have applications that should not be able to access an entire database. There are other various scaling reasons, and PG extensions that can be helpful. But I agree that for small to medium sized projects, SQLite is highly underrated.
The biggest one is redundancy. Architecting with Read replicas is much easier with Postgres than Sqlite because of it's server model.
Sqlite on the server is a fantastic starter database. Dead simple to set up, highly performant and scales way higher (vertically) than anyone gives it credit for.
But there certainly is a point you'll have to scale out instead of up, and while there are some great solutions for that (rqlite, litefs, dqlite, marmot) it's not inherent to Sqlite's design.
Should replication really be a concern of the DB layer?
Replication means writing queries which alter the data to multiple machines, right?
Shouldn't that be done by a software one level up? Which takes in the queries via some network protocol and then sends them to all machines.
That would sound more logical to me.
Historically, yes. Databases were software that were concerned with both storage and networking.
It's fine to want to separate those out, but it's not easy to do so and there are reasons they've been coupled for decades.
What makes it hard?
Having a single DB that takes write queries via a proxy which spreads them out to multiple read-only-DBs sounds easy at first.
When do you consider the write/transaction to be completed?
What do you do about out-of-sync read replicas?
ACID gets real hard real fast when introducing replication.
> When do you consider the write/transaction to be completed?
Sending a UPDATE/INSERT/DELETE statement to SQLite is not blocking? I would think it is, because in my code I can read the number of affected rows right after I sent the query.
> What do you do about out-of-sync read replicas?
Delete them and replace them by uploading a checkpoint and replaying a log of the queries since then.
rqlite[1] creator here, happy to answer any questions about it.
When your application scales beyond one machine that needs access to the same database, PostgreSQL becomes an obviously better choice than SQLite. Until that point, SQLite is a fine, and honestly underrated choice.
DuckDB is another option worth considering.
Should the concept of "machines" really be a concern of the DB layer?
SQLite already allows multiple connections, so putting it on a server and adding a program that talks a network protocol and proxies the queries to the DB sounds more logical to me?
High performance software is written acknowledging the reality that it will run on hardware. Databases tend to be a class of software that is hyper-focused on performance.
Writing a networked application that uses SQLite as a database is perfectly reasonable. You're just making the decision to lift the layer of abstraction that is concerned with machines from the DB to your application, which may or may not be a reasonable thing to do.
And after all of that you basically have something that looks like postgres or mysql.
My feeling is that I would have something better.
Because I can use SQLite and its "a file is all you need" approach as long as I don't need multiple machines.
And only bring in the other software (the proxy) when I need it.
Jeebus i keep forgetting how many of you are bootcamp noobs here
Maybe so but this isn't helpful.
MySQL has limited spatial data/function support versus PostGIS extension.
Certainly if you need a network-attached database and aren't creating your own home brew network-attached database (the so-called API server), Postgres is a pretty good choice.
Concurrency.
Yet again SQlite is over represented here in the server context.
Yo dawg I heard you liked databases so we put a database in your database.
Dang wish this wasn't the bottom comment because it is amazing. If it wasn't here, I'd have posted it.
Came for this
(laughs in xzibit)