Im thinking of a way I could somehow dynamicly allocate maybe by using KV or smth but don't know if

Im thinking of a way I could somehow dynamicly allocate maybe by using KV or smth but don't know if its possible in any way

BBefus Im thinking of a way I could somehow dynamicly allocate maybe by using KV or smt...

Hello, I’m Allie!•2/20/24, 9:04 AM

No, unless you are willing to go via the HTTP API, which is painfully slow

HHello, I’m Allie!No, unless you are willing to go via the HTTP API, which is *painfully* slow

BefusOP•2/20/24, 9:49 AM

Just thinking out loud here but if I were to create the database manually before hand, when in my case a new organisation register check for the first empty database and assign that way. You reckon that would be possible?

BBefus Just thinking out loud here but if I were to create the database manually before...

Hello, I’m Allie!•2/20/24, 9:57 AM

Maybe, but that would only work for the first N databases

bun•2/20/24, 10:13 AM

are there no tools compatible with d1 ? datagrip, tableplus, or anything?

bun•2/20/24, 10:15 AM

does worker use TCP for D1 or HTTP? like does it count as an outgoing request?

Original message was deleted

bun•2/20/24, 10:18 AM

holy f

Júlio Mühlbauer•2/20/24, 10:20 AM

Does D1 have this types of limitations because of SQLite? https://twitter.com/heyandras/status/1758940021640274148?t=1C0_ZdZ4JytNJItZrMJcwg&s=19

Andras Bacsai (@heyandras) on X

The only case when Sqlite is not suitable for your project, is when more than one server/instance of your app needs to access the database.

Your app is probably not in this category.

Twitter

•

2/17/24, 7:42 PM

Bbun does worker use TCP for D1 or HTTP? like does it count as an outgoing request?

Hello, I’m Allie!•2/20/24, 10:21 AM

It uses something internally, which afaik isn't quite TCP or HTTP

bun•2/20/24, 10:25 AM

so queries dont count as outgoign requests?

JJúlio Mühlbauer Does D1 have this types of limitations because of SQLite? https://twitter.com/he...

Hello, I’m Allie!•2/20/24, 10:25 AM

Not quite, since the D1 Layer that is on top of SQLite handles simultaneous connections. It can't actually process multiple requests simultaneously, but it can store requests in a queue. I would note too that while you can use D1 outside of Cloudflare, it isn't meant to be used that way, and thus it is a bit unergonomic

Bbun so queries dont count as outgoign requests?

Hello, I’m Allie!•2/20/24, 10:26 AM

They do. No matter the underlying protocol, an outgoing I/O op is counted as a "request"(though opening a TCP socket is a single connection, even if you do multiple things with it)

Hello, I’m Allie!•2/20/24, 10:26 AM

TCP is the only special snowflake, at least at the moment

HHello, I’m Allie!Not quite, since the D1 Layer that is on top of SQLite handles simultaneous conn...

Júlio Mühlbauer•2/20/24, 10:28 AM

Is there some limitations on scalability compared to a postgres database? It would be cool to see this type of thing on the docs. Thanks!

bun•2/20/24, 10:28 AM

are there any standardized approaches on sharding the database currently? it sucks a lot that databases are 2 gb only....

JJúlio Mühlbauer Is there some limitations on scalability compared to a postgres database? It wou...

Hello, I’m Allie!•2/20/24, 10:28 AM

There is https://developers.cloudflare.com/d1/platform/limits/, if that is what you mean

Bbun are there any standardized approaches on sharding the database currently? it suc...

Hello, I’m Allie!•2/20/24, 10:29 AM

No, though generally the optimal sharding strategy is dependent on the shape of the data you are storing

bun•2/20/24, 10:29 AM

like do i have to do 100 bindings in my wrangler.toml

Bbun like do i have to do 100 bindings in my wrangler.toml

Hello, I’m Allie!•2/20/24, 10:29 AM

That is a way. You can also offload large values to a separate store, if you don't need to query on them directly

HHello, I’m Allie!There is https://developers.cloudflare.com/d1/platform/limits/, if that is what ...

Júlio Mühlbauer•2/20/24, 10:30 AM

As I undertood, this is mostly limits impose by cloudflare that can be increased by a request or paying. But not about the technology the database uses, right?

JJúlio Mühlbauer As I undertood, this is mostly limits impose by cloudflare that can be increased...

Hello, I’m Allie!•2/20/24, 10:32 AM

Not necessarily. While some of the limits are applied by Cloudflare directly, others are in place because the database may not be stable above those limits. You might be able to push it farther, but CF can't guarantee that it will function correctly, so you can't raise the limit that far

HHello, I’m Allie!Not necessarily. While some of the limits are applied by Cloudflare directly, ot...

Júlio Mühlbauer•2/20/24, 10:32 AM

Understand. Thanks!

bun•2/20/24, 10:32 AM

are there no estimates on increasing the storage of the database?

Bbun are there no estimates on increasing the storage of the database?

Hello, I’m Allie!•2/20/24, 10:33 AM

Estimates of Pricing? Timeline? Other?

bun•2/20/24, 10:33 AM

timeline

Bbun timeline

Hello, I’m Allie!•2/20/24, 10:33 AM

:soontm:

Júlio Mühlbauer•2/20/24, 10:33 AM

Would love to see a postgres or something that scales made by cloudflare btw

Original message was deleted

BefusOP•2/20/24, 10:59 AM

I will have a look at this, thanks a lot!

Original message was deleted

BefusOP•2/20/24, 11:03 AM

That will be a solution to the waiting for deployment?

BefusOP•2/20/24, 11:04 AM

Awesome!

Original message was deleted

Max (@rozenmd)•2/20/24, 12:09 PM

here's how we load the entire northwind db (via the REST API/wrangler): https://raw.githubusercontent.com/cloudflare/d1-northwind/main/db/data-big.sql

Vignesh•2/20/24, 12:12 PM

Any ETA on official dynamic binding?

Original message was deleted

BefusOP•2/20/24, 12:25 PM

Gonna be trying this out these next few days first gotta setup next auth with d1

Original message was deleted

itsmatteomanf•2/20/24, 1:07 PM

Oh yeah, if you do a batch with reused statements especially, it’s in the 100s of thousands of inserts per call in a lot cases. Performance wise is really good.

itsmatteomanf•2/20/24, 1:08 PM

There are a couple of issues with reliability (but they are very rare now and they are being fixed), but for the most part it’s really good.

Original message was deleted

itsmatteomanf•2/20/24, 1:10 PM

That even seems slow, but it depends on what you use (Worker or HTTP API?) and the amount of data per row + number of indexes on top

Original message was deleted

itsmatteomanf•2/20/24, 1:36 PM

Ok, yeah, makes more sense

my rows are smaller!

Original message was deleted

Hello, I’m Allie!•2/20/24, 1:42 PM

Yes, but they have to be created when the script receives its first event. You can't create them before the event arrives

Martin Carlsson•2/20/24, 6:16 PM

Are there any way of getting data out of D1?

I mean, like aws dms, airbyte, or fivetran?

Thenerox•2/20/24, 11:44 PM

hello can i connect to D1 using dbeaver or something like that?

Tony Leung•2/21/24, 4:59 AM

Hi guys, I'm querying data with the query above, but it seems like scanning too much data. I've set index for

p.posted

p.posted

and

p.retry

p.retry

but it's still scanning the whole table

. Is there anyway to reduce scanned rows of the query above? I'm running this query every minutes so everyday I'm wasting 10-20k scanned row. This table is going to grow much larger so the waste will be more significant

Screenshot_2024-02-21_at_11.55.02_AM.png

prabha•2/21/24, 10:30 AM

Hi everyone.. How to create REST API for CRUD operations in D1

stefanoros•2/21/24, 10:45 AM

Hello everyone, i'm trying to set up a simple many-to-many join

SELECT *
    FROM categories
    JOIN images_categories ON categories.id = images_categories.category_id
    JOIN images ON images_categories.image_id = images.id;

SELECT *
    FROM categories
    JOIN images_categories ON categories.id = images_categories.category_id
    JOIN images ON images_categories.image_id = images.id;

i get this error

Error: D1_ERROR: A prepared SQL statement must contain only one statement.

Error: D1_ERROR: A prepared SQL statement must contain only one statement.

. Can't seem to figure why.

Sstefanoros Hello everyone, i'm trying to set up a simple many-to-many join ```SELECT * ...

Max (@rozenmd)•2/21/24, 10:56 AM

can you post the exact code, including JS/TS that you're using?

stefanoros•2/21/24, 10:57 AM

sure

    let result = await platform!.env.DB.prepare(
        `SELECT *
    FROM categories
    JOIN images_categories ON categories.id = images_categories.category_id
    JOIN images ON images_categories.image_id = images.id;
    `
    ).all();

    let result = await platform!.env.DB.prepare(
        `SELECT *
    FROM categories
    JOIN images_categories ON categories.id = images_categories.category_id
    JOIN images ON images_categories.image_id = images.id;
    `
    ).all();

if I remove one join statement it works

Max (@rozenmd)•2/21/24, 10:59 AM

that's weird - i've only seen that error with the parser failing to understand comments

Max (@rozenmd)•2/21/24, 11:00 AM

does it work if you combine the SQL onto a single line? like

`SELECT * FROM categories JOIN images_categories ON categories.id = images_categories.category_id JOIN images ON images_categories.image_id = images.id;`

`SELECT * FROM categories JOIN images_categories ON categories.id = images_categories.category_id JOIN images ON images_categories.image_id = images.id;`

MMax (@rozenmd)does it work if you combine the SQL onto a single line? like ``` `SELECT * FROM ...

stefanoros•2/21/24, 11:25 AM

Yes! It works now, thanks

TTony Leung Hi guys, I'm querying data with the query above, but it seems like scanning too ...

Tony Leung•2/21/24, 11:35 AM

hi guys, could anyone help me

TTony Leung Hi guys, I'm querying data with the query above, but it seems like scanning too ...

Max (@rozenmd)•2/21/24, 11:45 AM

is there a single index covering both

posted

posted

and

retry

retry

, or two individual indexes?

Tony Leung•2/21/24, 11:50 AM

yes, I'm having a composite index like this:

CREATE INDEX post_posted_retry_idx ON post (posted, retry);

CREATE INDEX post_posted_retry_idx ON post (posted, retry);

let result = await platform!.env.DB.prepare( `SELECT * FROM categories JOIN images_categories ON categories.id = images_categories.category_id JOIN images ON images_categories.image_id = images.id; ` ).all();

Im thinking of a way I could somehow dynamicly allocate maybe by using KV or smth but don't know if

Similar Threads

Similar Threads

Similar Threads