yes running via wrangler dev

Ddinesh_karuppasamy yes running via wrangler dev

C

Cyb3r-Jak3•10/31/23, 1:13 AM

Did the you also apply the migrations locally? Wrangler dev by default uses local mode which doesn’t use any remote data.

CCyb3r-Jak3 Did the you also apply the migrations locally? Wrangler dev by default uses loca...

D

dinesh_karuppasamyOP•10/31/23, 1:16 AM

okay thanks. Let me check again.

P

Paul Shriner•10/31/23, 12:10 PM

D1 question and I think that I know the answer, I would like to sort by similarity of two strings. In a traditional enviornment I think I would use editdist3 which seems like the goto in SQLite, but I don't see it support in D1... Did I miss something?

K

kian•10/31/23, 12:18 PM

The spellfix1 virtual table is not included in the SQLite amalgamation and is not a part of any standard SQLite build. It is a loadable extension.

K

kian•10/31/23, 12:18 PM

It's a non-standard extension so it won't be included in D1 currently.

P

Paul Shriner•10/31/23, 12:24 PM

that is what I figured, but was hoping

P

Paul Shriner•10/31/23, 12:26 PM

Basically what I am trying to do is take a comma delimited list of sorted numbers say [1,2,4,7] and find the record most closely match [1,2,7] .

I think that I can do it by flatteing the column into fields, but It won't be simple to read nor will it be performant.

So I would take ideas

P

Paul Shriner•10/31/23, 12:32 PM

Actually maybe not horrible iwth SQL math

K

kian•10/31/23, 2:38 PM

// TODO(someday): Allow temporary tables. Creating a temporary table actually causes
//   SQLite to open a separate temporary file to place the data in. Currently, our storage
//   engine has no support for this.

// TODO(someday): Allow temporary tables. Creating a temporary table actually causes
//   SQLite to open a separate temporary file to place the data in. Currently, our storage
//   engine has no support for this.

// TODO(someday): Allow temporary tables. Creating a temporary table actually causes
//   SQLite to open a separate temporary file to place the data in. Currently, our storage
//   engine has no support for this.

// TODO(someday): Allow temporary tables. Creating a temporary table actually causes
//   SQLite to open a separate temporary file to place the data in. Currently, our storage
//   engine has no support for this.

K

kian•10/31/23, 2:38 PM

The new storage backend doesn't support temporary tables.

K

kian•10/31/23, 2:58 PM

They fall under the same comment as far as I can see

K

kian•10/31/23, 3:07 PM

sqlite3CreateView

sqlite3CreateView

calls

sqlite3StartTable

sqlite3StartTable

so it would follow the same path of attempting to create a temporary file - that's my understanding

K

kian•10/31/23, 3:14 PM

Creating a temporary view w/

PRAGMA query_only = 1

PRAGMA query_only = 1

results in

attempt to write a readonly database (8)

attempt to write a readonly database (8)

but you might be right, I can't say I've ever really used them before.

K

kian•10/31/23, 3:16 PM

Omitting

TEMP

TEMP

from

CREATE VIEW

CREATE VIEW

just moves it from the

main

main

database to

temp

temp

as far as I can see, so I assume it's still trying to write something at the least - of which, would be on a temporary database file.

K

kian•10/31/23, 3:18 PM

The edge runtime supposedly sets

temp_store

temp_store

to

MEMORY

MEMORY

, which conflicts with the comment - so there might be a quirk in SQLite or D1.

J

JustinNoel•10/31/23, 3:21 PM

I just migrated my project from a Planetscale database to D1 yesterday.

It was super easy and improved my page loads times thanks to lower latency.

Thanks so much to everyone on the CF D1 team!

P

Paul•10/31/23, 3:22 PM

The docs for time travel are a bit confusing.
Does it create a new snapshot when I run this:

wrangler d1 time-travel info YOUR_DATABASE

wrangler d1 time-travel info YOUR_DATABASE

?

Also, how do I get a list of previous snapshots so I can restore to one?

P

Paul•10/31/23, 3:24 PM

Ah I think i get it. Because it's happening every minute for the last 30 days, i can just supply it with any timestamp?

PPaul Ah I think i get it. Because it's happening every minute for the last 30 days, i...

H

Hello, I’m Allie!•10/31/23, 3:25 PM

You need the ID of the backup first:

$ wrangler d1 backup download example-db 123a81a2-ab91-4c2e-8ebc-64d69633faf1

$ wrangler d1 backup download example-db 123a81a2-ab91-4c2e-8ebc-64d69633faf1

$ wrangler d1 backup download example-db 123a81a2-ab91-4c2e-8ebc-64d69633faf1

$ wrangler d1 backup download example-db 123a81a2-ab91-4c2e-8ebc-64d69633faf1

P

Paul•10/31/23, 3:26 PM

I don't think the

backup

backup

works anymore on beta?

H

Hello, I’m Allie!•10/31/23, 3:26 PM

Oh wait

H

Hello, I’m Allie!•10/31/23, 3:26 PM

Ur right

PPaul I don't think the `backup` works anymore on beta?

H

Hello, I’m Allie!•10/31/23, 3:27 PM

Yeah, then what you said should work fine

P

Paul•10/31/23, 3:27 PM

which is honestly a bit concering, i'd like to be able to easily export my entire database as standard sql

P

Paul•10/31/23, 3:27 PM

i feel like time-travel is a form of lock-in

P

Paul•10/31/23, 3:28 PM

it's an excellent feature for sure, but it's still very important for me to be able to down the entirety of my database, so I can load it up in a local client and play with it

PPaul it's an excellent feature for sure, but it's still very important for me to be a...

H

Hello, I’m Allie!•10/31/23, 3:28 PM

I think this falls under the

beta

beta

banner. Afaik they do still intend to have a way to dump the DB at some point

HHello, I’m Allie!I think this falls under the `beta` banner. Afaik they do still intend to have a...

P

Paul•10/31/23, 3:29 PM

Oh excellent.

PPaul Oh excellent.

D

Dahoom•10/31/23, 5:17 PM

https://github.com/nora-soderlund/cloudflare-d1-backups in the meantime you can use this

GitHub

GitHub - nora-soderlund/cloudflare-d1-backups: A package to dump a ...

A package to dump a copy of your experimental D1 database through a worker. - GitHub - nora-soderlund/cloudflare-d1-backups: A package to dump a copy of your experimental D1 database through a worker.

D

Dahoom•10/31/23, 5:19 PM

There should be a discord bot that scans for keywords like "export" and "dump" in this channel and automatically posts this url

Original message was deleted

M

Max (@rozenmd)•10/31/23, 8:52 PM

was on holiday, will get on it

G

GuillaumeLakano•11/1/23, 4:30 PM

Hi! I'm trying to dynamically create D1 database with the CF API, and then to PATCH the script bindings to permit to my worker to reach the new D1 database.
The dynamic creation works great, but the PATCH return me this error:

{
            "result": null,
            "success": false,
            "errors": [
              {
                "code": 10001,
                "message": "workers.api.error.content_type"
              }
            ],
            "messages": []
          }

{
            "result": null,
            "success": false,
            "errors": [
              {
                "code": 10001,
                "message": "workers.api.error.content_type"
              }
            ],
            "messages": []
          }

{
            "result": null,
            "success": false,
            "errors": [
              {
                "code": 10001,
                "message": "workers.api.error.content_type"
              }
            ],
            "messages": []
          }

{
            "result": null,
            "success": false,
            "errors": [
              {
                "code": 10001,
                "message": "workers.api.error.content_type"
              }
            ],
            "messages": []
          }

I'm not sure of how to adapt the documentation for the bindings specificity of D1:
https://developers.cloudflare.com/api/operations/worker-script-get-settings

I've tried multiple things:

{
    bindings: [
        { type: 'd1_databases', database_name: name, database_id: uuid }
    ]
}

{
    bindings: [
        { type: 'd1_databases', database_name: name, database_id: uuid }
    ]
}

{
    bindings: [
        { type: 'd1_databases', database_name: name, database_id: uuid }
    ]
}

{
    bindings: [
        { type: 'd1_databases', database_name: name, database_id: uuid }
    ]
}

and

{
    bindings: [
        { type: 'd1_databases', name, namespace_id: uuid }
    ]
}

{
    bindings: [
        { type: 'd1_databases', name, namespace_id: uuid }
    ]
}

{
    bindings: [
        { type: 'd1_databases', name, namespace_id: uuid }
    ]
}

{
    bindings: [
        { type: 'd1_databases', name, namespace_id: uuid }
    ]
}

If someone have some ideas to help

Cloudflare API Documentation

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

K

kian•11/1/23, 4:32 PM

Are you using FormData?

K

kian•11/1/23, 4:32 PM

K

kian•11/1/23, 4:32 PM

It sounds like you're only sending JSON.

K

kian•11/1/23, 4:33 PM

It should be FormData, with a

settings

settings

key whose value is a stringified object

G

GuillaumeLakano•11/1/23, 4:37 PM

Oh good point, I'll try, thank you @kian !

B

Bram•11/1/23, 7:19 PM

Hi, for my use case I have to store large blobs of JSON (3-10MB files). My idea was to combine D1 (for metadata) and R2 (for blob storage). I was wondering if it is possible to write to both in a transactional way? Is there some kind of transaction API that I could use?

B

Bram•11/1/23, 7:22 PM

Hmm now I think about it, I would probably be better of having the user upload directly to R2 with signed urls?

B

Bram•11/1/23, 7:27 PM

or maybe not:

Typically, you will not share presigned URLs directly with end users or browsers, as presigned URLs are used more for internal applications.

B

Bram•11/1/23, 7:31 PM

Is it easy to subscribe to an event for when R2 is done uploading a new file?

B

Bram•11/1/23, 7:36 PM

ok thanks for the info

Original message was deleted

B

Bram•11/1/23, 7:36 PM

and yea that is true, will take the discussion there

J

jahir9991•11/2/23, 9:27 AM

In batch api how can use data from previous statement to the next statement like we do in db transactions ?

Similar Threads

Similar Threads

Similar Threads