I will be extra transparent here: we’re in an alpha, and generally don’t share timelines about beta

I will be extra transparent here: we’re in an alpha, and generally don’t share timelines about beta or GA steps, because they almost always lead to disappointment. A beta or GA isn’t just a checklist of features: it’s about ensuring we have observed the system to be reliable in the hands of users. This is especially true for a database product.

Further, when we land the new storage subsystem, we fully expect (we have validated internally, in fact) for these transient query failures to go away. They are an outcome of how we shipped an early alpha — so that users can play with D1 and provide feedback on the rest of the experience — versus holding off until we landed the storage layer we intend to build on proper

Matt SilverlockOP•5/4/23, 11:56 AM

Some upcoming D1 documentation:

Using indexes - https://silverlock-d1-indexes-json-l.cloudflare-docs-7ou.pages.dev/d1/learning/using-indexes/
Querying JSON (if you're used to JSON functions in Postgres, this is a big win): https://silverlock-d1-indexes-json-l.cloudflare-docs-7ou.pages.dev/d1/learning/querying-json/

MMatt Silverlock Some upcoming D1 documentation: - Using indexes - https://silverlock-d1-indexes...

Matt SilverlockOP•5/4/23, 11:57 AM

We'll also be looking at making it clearer on how to use Pages + D1 together as well.

Farhian•5/4/23, 2:27 PM

Hi all, I am trying to define some foreign key relationships and I am probably doing something wrong. I have created a sql script which creates the tables, indexes and the foreign key relationships. The table creation and index creation operations work while the foreign key statements fail. Does anyone have a small example on how to do that? I want to see if maybe my syntax is wrong. Thank you!

Farhian•5/4/23, 2:39 PM

here is a sample sql
CREATE TABLE resellersresellers (

id

id

string PRIMARY KEY,

email

email

string NOT NULL,

name

name

string NOT NULL,
createdcreated datetime
);

CREATE TABLE agenciesagencies (

id

id

string PRIMARY KEY,
reseller_idreseller_id string NOT NULL,

name

name

string NOT NULL,

location

location

string,
createdcreated datetime
FOREIGN KEY(reseller_idreseller_id) REFERENCES resellersresellers(

id

id

),
);

Farhian•5/4/23, 2:40 PM

Farhian•5/4/23, 2:55 PM

omg

Farhian•5/4/23, 2:55 PM

thank you very much! @skye_31

Farhian•5/4/23, 2:55 PM

I need new glasses

Atinux•5/4/23, 4:08 PM

Just made an open source demo of Nuxt + CF Pages + D1: https://github.com/atinux/nuxt-todos-edge
Demo: https://nuxt-todos-edge.pages.dev/

GitHub

GitHub - Atinux/nuxt-todos-edge: A todos application with user auth...

A todos application with user authentication, with SSR & SQLite, working on the edge. - GitHub - Atinux/nuxt-todos-edge: A todos application with user authentication, with SSR & SQL...

Nuxt Todos Edge

A Nuxt demo hosted on CloudFlare Pages with server-side rendering on the edge and using the D1 database

Original message was deleted

Max (@rozenmd)•5/6/23, 6:12 PM

happy to fix it but it isn't clear what's happening in the screenshot

Max (@rozenmd)•5/6/23, 6:18 PM

oh right, will give it a look when I'm back at my laptop

Max (@rozenmd)•5/6/23, 6:18 PM

yeah that's definitely confusing

peachneo•5/7/23, 3:49 AM

hi all, Is there a limit on tables?

peachneo•5/7/23, 3:50 AM

*# of tables

Ppeachneo hi all, Is there a limit on tables?

Cyb3r-Jak3•5/7/23, 3:52 AM

There isn't a documented limit

Ppeachneo hi all, Is there a limit on tables?

Matt SilverlockOP•5/7/23, 9:21 AM

It depends on your total database size limit. At 100MB, it’s about 25,000, but then you wouldn’t have any room left for data nor indexes. In practice, we’re likely to document this as 1,000 (per DB) as a soft limit.

MMatt Silverlock It depends on your total database size limit. At 100MB, it’s about 25,000, but t...

peachneo•5/7/23, 9:21 AM

thanks

Ffate i tried around here and there, but i couldn't find options to set foreign keys i...

Hello, I’m Allie!•5/7/23, 7:07 PM

You should just be able to add foreign keys to your create table statement

TigersWay•5/8/23, 3:41 AM

Hi. Wondering how far/deep we should not be concerned about the (bad) performance of our SQL queries.... I have multiple - relatively simple - queries working with the primary-key-automatic-index which should be done in a few ms, taking now ~1s, while the network latency is not so great but clearly expected.

TigersWay•5/8/23, 3:42 AM

@geelen LOL the usual "soon"

TigersWay•5/8/23, 3:43 AM

Well it's still usable because clearly not in production.... I'll wait. Thanks

lukeed•5/8/23, 5:23 AM

Any pricing updates?

lukeed•5/8/23, 5:24 AM

James•5/8/23, 6:38 AM

Nothing official yet. Functionally, there’s probably not going to be much difference other than your Cloud vendor of choice, which is going to come down to availability and pricing.

D1 is still very early and in alpha right now, but it’ll probably make the most sense to use alongside the rest of the Cloudflare stack.

Murder Chicken•5/8/23, 11:41 AM

Is there a best practice way to accomplish a query of this nature (with many more records) through D1's API?

INSERT INTO example_table (activity_id, user_id, status) 
VALUES 
  (2496, 1, 'AVAILALBE'),
  (2496, 1, 'NOT_AVAILABLE'),
  (2496, 2, 'AVAILABLE') 
ON CONFLICT DO UPDATE SET status = excluded.status;

INSERT INTO example_table (activity_id, user_id, status) 
VALUES 
  (2496, 1, 'AVAILALBE'),
  (2496, 1, 'NOT_AVAILABLE'),
  (2496, 2, 'AVAILABLE') 
ON CONFLICT DO UPDATE SET status = excluded.status;

Murder Chicken•5/8/23, 11:41 AM

Is it leveraging

batch

batch

with an array of objects or something of that nature?

Murder Chicken•5/8/23, 11:42 AM

Though I could not find a way to pass an array of data to batch... just an array of

stmt.bind

stmt.bind

ashutosh•5/8/23, 1:01 PM

Can we access the D1 database outside Cloudflare workers environment?

I will be extra transparent here: we’re in an alpha, and generally don’t share timelines about beta

Similar Threads