Cloudflare Developers•2y ago

D1 is not a good fit for large databases (10GB+), we recommend folks look at Hyperdrive instead in t

D1 is not a good fit for large databases (10GB+), we recommend folks look at Hyperdrive instead in that use case.

MMax (@rozenmd)D1 is not a good fit for large databases (10GB+), we recommend folks look at Hyp...

James•3/11/24, 9:57 AM

If I use Turbo that support Unlimited Storage, do I still need a hyperdrive, or is it necessary?

JJames If I use Turbo that support Unlimited Storage, do I still need a hyperdrive, or ...

Max (@rozenmd)OP•3/11/24, 9:57 AM

what's Turbo?

MMax (@rozenmd)what's Turbo?

James•3/11/24, 9:58 AM

Oh sorry this autocorrect I mean turso https://turso.tech/

JJames Oh sorry this autocorrect I mean turso https://turso.tech/

Max (@rozenmd)OP•3/11/24, 9:59 AM

We have a guide for setting that up here, no need for Hyperdrive: https://developers.cloudflare.com/workers/tutorials/connect-to-turso-using-workers/

MMax (@rozenmd)We have a guide for setting that up here, no need for Hyperdrive: https://develo...

James•3/11/24, 10:01 AM

Thx

TTigersWay You should probably try: `createdAt DATETIME NOT NULL DEFAULT (CURRENT_TIMESTAMP...

works thanks

Madyan•3/11/24, 11:02 AM

Hey all! I'm migrating into D1, and I have a seeding script that reads files from a directory (containing mock profile images of users, where each filename represents name of person), then it creates the relevant data and inserts it into the DB.
The challenge is that Node.js seems to be unable to access D1, while a wrangler runtime is unable to access the filesystem.
I'm now thinking of creating a combination of both (a Node script that sends an http request to a wrangler endpoint with the data it should insert into the DB). Any suggestions for an easier solution?

Ccopper @Max (@rozenmd) when will Hyperdrive will be avail for MySQL?

Max (@rozenmd)OP•3/11/24, 11:06 AM

worth asking in the Hyperdrive channel

Madyan•3/11/24, 11:13 AM

Thanks @copper!
The reason I was using Node is to be able to access the filesystem. My understanding is that three frameworks you're referring to would be using the wrangler runtime too, and hence won't be able to access the filesystem either, right?

Madyan•3/11/24, 11:15 AM

makes sense

Madyan•3/11/24, 11:15 AM

this is just a script to provide data for local / staging environments

Ccopper ok and why not put those files in s3 (or cf r2)

Madyan•3/11/24, 11:19 AM

ah makese sense. The script currently does both populating DB and uploading images (with my old provider). Perhaps I should do the image upload first like you suggested. Thank you!

Madyan•3/11/24, 11:24 AM

yeah, ideally I'd like to keep it all automated so that I can copy the script across my differnt project in the future

Madyan•3/11/24, 11:25 AM

but that's very cool! wasn't aware of aws4fetch

Madyan•3/11/24, 11:25 AM

I'll look into r2 first and have that as a plan B

Madyan•3/11/24, 11:37 AM

It creates fake data, adhering to the TS types generated by ORM to ensure everything is in sync. It's using images from the filesystem to ensure the files exist (e.g. I have a directory with food photos with filenames like

pad-thai.jpeg

pad-thai.jpeg

, and I use the script to generate data based on the available files)

Madyan•3/11/24, 11:38 AM

We created a prototype, so we were hoping to keep the mock data realistic so that it resonates better with potential buyers when we demo it to them

Oleksii Kilevoi•3/11/24, 12:09 PM

Is there any way to create d1 database via http with token?

OOleksii Kilevoi Is there any way to create d1 database via http with token?

Hello, I’m Allie!•3/11/24, 12:11 PM

https://developers.cloudflare.com/api/operations/cloudflare-d1-create-database

Cloudflare API Documentation

Interact with Cloudflare's products and services via the Cloudflare API

HHello, I’m Allie!https://developers.cloudflare.com/api/operations/cloudflare-d1-create-database

Oleksii Kilevoi•3/11/24, 12:11 PM

Thanks a lot!

Ccopper @Max (@rozenmd) thanks for hinting to a hyperdrive channel, who could I ask abou...

Max (@rozenmd)OP•3/11/24, 1:17 PM

re: cold-starts D1 Latency

MMax (@rozenmd)re: cold-starts https://discord.com/channels/595317990191398933/1208459278625677...

Hello, I’m Allie!•3/11/24, 3:02 PM

Is that due to runtime updates?

Nicolas•3/11/24, 10:02 PM

Hi all,

So I am trying to see how I can use D1 in a multi-tenant fashion.

Suppose I am developing an onboarding workflow, when a user signs up to my platform, I need to create a database for that user; as the user invites their team members to his group, they will share the same D1 database that was created when they signed up.

As far as I understand, there needs to be bindings between workers and D1 so the workers can access them, so how would this work in this multi-tenant scenario? I suppose I will need to store the database ID's of all the dynamically created D1's in another root level D1 database that keeps track of tenant information, at a minimum, but what about granting access to workers for these dynamically created D1s?

Are there any official guides on this?

Thanks!

matt•3/11/24, 11:27 PM

has anyone noticed how restrictive sqlite is when it comes to migration scripts? there are barely any options to alter an existing column. how are people getting by with all these missing features?

squareclamp•3/12/24, 3:44 AM

Hi, I'm looking for help on how to do multi statement inserts in d1, when the inserts are dependent.

Example, I want to perform a transaction to create a

post

post

and have author inserted in

post_author

post_author

. If this is done separately, a post can be without author. I can't use batch API, because I need

post_id

post_id

for entry in

post_author

post_author

.

Is this an alternative that can be achieved with d1?

Ccopper @Max (@rozenmd) thanks for hinting to a hyperdrive channel, who could I ask abou...

Tony Leung•3/12/24, 3:44 AM

Oh that's why my app are suddenly slow-read on low traffic period.

Tony Leung•3/12/24, 3:45 AM

Bro literally had summoned big boss Kenton to answer your question :MeowHeartCloudflare:

D1 is not a good fit for large databases (10GB+), we recommend folks look at Hyperdrive instead in t

Similar Threads

Similar Threads

Similar Threads