Anyone remember whether D1 Auto-Vacuums, or whether we should run them ourselves periodically?

HHello, I’m Allie!Anyone remember whether D1 Auto-Vacuums, or whether we should run them ourselves...

M

Right now: neither. We need to implement auto-vacuum. You can't run it yourself. But we don't include the free list in the reported/restricted size. More of an "us" problem right now.

MMatt Silverlock Right now: neither. We need to implement auto-vacuum. You can't run it yourself....

H

Hello, I’m Allie!OP•6/7/23, 3:23 PM

Are there any other things you recommend doing to maximize performance? I’ve read about a couple of Pragmas you can set, but I’m not sure what applies/works with D1

H

Hello, I’m Allie!OP•6/7/23, 3:23 PM

Or, more like, is there something about Best Practices?

HHello, I’m Allie!Are there any other things you recommend doing to maximize performance? I’ve rea...

M

Matt Silverlock•6/7/23, 3:26 PM

Indexes! One good index goes a LONG way.

M

Matt Silverlock•6/7/23, 3:27 PM

There will be docs about all of this — just takes time to write.

S

shixzie•6/7/23, 8:44 PM

is there a way to specify the

migrations_dir

migrations_dir

with the cli? I'm working on a pages project and it doesn't have a

wrangler.toml

wrangler.toml

file. So I'm trying a script like

npx wrangler d1 migrations apply <DB> --migrations-dir=<dir>

npx wrangler d1 migrations apply <DB> --migrations-dir=<dir>

but it's not working.

Sshixzie is there a way to specify the `migrations_dir` with the cli? I'm working on a pa...

M

Matt Silverlock•6/7/23, 9:18 PM

Not currently, no. Pages does not support a config file.

MMatt Silverlock Not currently, no. Pages does not support a config file.

S

shixzie•6/7/23, 9:26 PM

Is that a planned feature? Any suggestions on what can I do to manage migrations meanwhile?

Sshixzie Is that a planned feature? Any suggestions on what can I do to manage migrations...

M

Matt Silverlock•6/8/23, 12:31 AM

It is planned. There is not currently a workaround. cc @nevika

M

Murder Chicken•6/8/23, 6:39 PM

Is there an easy way to get a prepared d1 statement as text (with the bindings applied) ?

MMurder Chicken Is there an easy way to get a prepared d1 statement as text (with the bindings a...

M

Matt Silverlock•6/8/23, 8:14 PM

Not at the moment, no: not something we yet expose.

Original message was deleted

M

Matt Silverlock•6/8/23, 8:45 PM

https://github.com/cloudflare/cloudflare-docs/pull/9298 incoming

GitHub

d1: error.cause be gone by elithrar · Pull Request #9298 · cloudfla...

Blocked until cloudflare/workers-sdk#3416 is merged.

As of wrangler 3.1.1 the top-level Error property will now include the detailed error message.
Although Error.cause is standardized & in th...

MMatt Silverlock https://github.com/cloudflare/cloudflare-docs/pull/9298 incoming 🙂

I

itsmatteomanf•6/8/23, 8:46 PM

I just changed my code to show

error.cause

error.cause

Iitsmatteomanf I just changed my code to show `error.cause` <:lul:946531054485930015>

M

Matt Silverlock•6/8/23, 8:54 PM

We’ll intend to keep that as is for a bit, but it’s an overdue change. Doesn’t matter that

.cause

.cause

is standard, unfortunately

MMatt Silverlock We’ll intend to keep that as is for a bit, but it’s an overdue change. Doesn’t m...

I

itsmatteomanf•6/8/23, 8:55 PM

Oh yeah, don't worry... I was one of the people not using it until you told me lul

OOr according to the docs it looks like the maximum arguments per SQL function was b...

M

Matt Silverlock•6/9/23, 12:18 AM

No, the docs just got ahead of the release. Still 8.

L

lokiwind•6/9/23, 12:09 PM

Hello, I would like to establish a connection between workers and mongodb, is there a tutorial on this?

MMatt Silverlock Not at the moment, no: not something we yet expose.

M

Murder Chicken•6/9/23, 12:13 PM

Is there any work around to, locally, somehow inspect what the end query result is with the bindings applied? It's super hard to debug queries if they're working directly in the DB but failing when D1 is trying to apply a bind. Any way to replicate what's under the hood in a custom debug function?

M

Murder Chicken•6/9/23, 12:34 PM

I see what's up now that I'm looking at the facade. The statement and bindings are sent to sqlite so there is no local interpolation. Anyway way to log d1 queries at the database level?

M

Max (@rozenmd)•6/9/23, 1:11 PM

Drizzle has a way of letting you see the actual query it runs, I don't think we have that as part of D1 yet

M

Murder Chicken•6/9/23, 1:20 PM

Does Drizzle even work properly with D1? I was seeing posts around people having to jump through hoops to get it to work.

M

Max (@rozenmd)•6/9/23, 1:26 PM

It works, it just has column name collisions when joining tables with the same column names + not using aliases

M

Murder Chicken•6/9/23, 1:27 PM

Ok, thanks @rozenmd , will take a look. Feels like overkill to debug a query with an ORM, haha.

Original message was deleted

M

Murder Chicken•6/9/23, 2:31 PM

Unfortunately, it's not that simple. Dumping the statement and bindings gives you the component pieces but does not allow you to introspect how they're being married together to form a complete query.

In this instance (just solved it), an integer value was being bound to the query as a string. I had to explicitly cast that particular parameter value with

parseInt

parseInt

to get the query to function. Figuring it out through process of elimination is time-consuming and I ended up having to remove bound parameters one by one until I identified which one was causing the issue so that I could trace back why that value was being bound improperly.

Llokiwind Hello, I would like to establish a connection between workers and mongodb, is th...

I

Isaac McFadyen•6/9/23, 3:33 PM

This isn't really #d1-open-alpha specific, so I'd recommend we move to #workers-discussions.

OOr according to the docs it looks like the maximum arguments per SQL function was b...

M

Matt Silverlock•6/9/23, 4:06 PM

Released: all functions now accept up to 32 arguments.

➜  wrangler d1 execute db-enam --command "SELECT json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7)"

🌀 Mapping SQL input into an array of statements
🌀 Parsing 1 statements
🚣 Executed 1 commands in 0.1332309991121292ms
┌──────────────────────────────────────────────────────────────────────────────────────────┐
│ json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7) │
├──────────────────────────────────────────────────────────────────────────────────────────┤
│ {"hello":1,"world":2,"foo":3,"bar":4,"baz":5,"quux":6,"floop":7}                         │
└──────────────────────────────────────────────────────────────────────────────────────────┘

➜  wrangler d1 execute db-enam --command "SELECT json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7)"

🌀 Mapping SQL input into an array of statements
🌀 Parsing 1 statements
🚣 Executed 1 commands in 0.1332309991121292ms
┌──────────────────────────────────────────────────────────────────────────────────────────┐
│ json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7) │
├──────────────────────────────────────────────────────────────────────────────────────────┤
│ {"hello":1,"world":2,"foo":3,"bar":4,"baz":5,"quux":6,"floop":7}                         │
└──────────────────────────────────────────────────────────────────────────────────────────┘

➜  wrangler d1 execute db-enam --command "SELECT json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7)"

🌀 Mapping SQL input into an array of statements
🌀 Parsing 1 statements
🚣 Executed 1 commands in 0.1332309991121292ms
┌──────────────────────────────────────────────────────────────────────────────────────────┐
│ json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7) │
├──────────────────────────────────────────────────────────────────────────────────────────┤
│ {"hello":1,"world":2,"foo":3,"bar":4,"baz":5,"quux":6,"floop":7}                         │
└──────────────────────────────────────────────────────────────────────────────────────────┘

➜  wrangler d1 execute db-enam --command "SELECT json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7)"

🌀 Mapping SQL input into an array of statements
🌀 Parsing 1 statements
🚣 Executed 1 commands in 0.1332309991121292ms
┌──────────────────────────────────────────────────────────────────────────────────────────┐
│ json_object('hello', 1, 'world', 2, 'foo', 3, 'bar', 4, 'baz', 5, 'quux', 6, 'floop', 7) │
├──────────────────────────────────────────────────────────────────────────────────────────┤
│ {"hello":1,"world":2,"foo":3,"bar":4,"baz":5,"quux":6,"floop":7}                         │
└──────────────────────────────────────────────────────────────────────────────────────────┘

Iitsmatteomanf Done, https://github.com/cloudflare/workerd/issues/729

M

Matt Silverlock•6/9/23, 4:14 PM

We just merged this internally. Had to do a bunch of work on our storage layer to make this work correctly. Goal is to get

ALTER TABLE

ALTER TABLE

statements released within the week or so.

JJacob Wright I'm getting "not authorized to use function" with timestamp functions (e.g. SELE...

M

Matt Silverlock•6/9/23, 4:16 PM

All of

current_timestamp

current_timestamp

,

current_time

current_time

, and

current_date

current_date

should now work outside of

CREATE TABLE

CREATE TABLE

statements now:

➜  wrangler d1 execute db-enam --command "SELECT current_timestamp, current_time, current_date" --json | jq '.[].results'
[
  {
    "current_timestamp": "2023-06-09 16:16:37",
    "current_time": "16:16:37",
    "current_date": "2023-06-09"
  }
]

➜  wrangler d1 execute db-enam --command "SELECT current_timestamp, current_time, current_date" --json | jq '.[].results'
[
  {
    "current_timestamp": "2023-06-09 16:16:37",
    "current_time": "16:16:37",
    "current_date": "2023-06-09"
  }
]

➜  wrangler d1 execute db-enam --command "SELECT current_timestamp, current_time, current_date" --json | jq '.[].results'
[
  {
    "current_timestamp": "2023-06-09 16:16:37",
    "current_time": "16:16:37",
    "current_date": "2023-06-09"
  }
]

➜  wrangler d1 execute db-enam --command "SELECT current_timestamp, current_time, current_date" --json | jq '.[].results'
[
  {
    "current_timestamp": "2023-06-09 16:16:37",
    "current_time": "16:16:37",
    "current_date": "2023-06-09"
  }
]

MMatt Silverlock We just merged this internally. Had to do a bunch of work on our storage layer t...

I

itsmatteomanf•6/9/23, 4:49 PM

Thank you, appreciated

MMatt Silverlock All of `current_timestamp`, `current_time`, and `current_date` should now work o...

I

itsmatteomanf•6/9/23, 4:50 PM

Oh hey, that was a thing I had pinged you about, too ahah

I

itstmillertime•6/9/23, 6:03 PM

How does one run a D1 database migration in CI/CD?
I am running the following:

wrangler d1 migrations apply production

wrangler d1 migrations apply production

But it wants a keyboard press `` to confirm the migration. Any advice is appreciated!

Iitstmillertime How does one run a D1 database migration in CI/CD? I am running the following: `...

M

Matt Silverlock•6/9/23, 6:17 PM

We need to accept the

-y

-y

flag (not supported yet). You should be able to

echo | wrangler d1 migrations apply <db-name>

echo | wrangler d1 migrations apply <db-name>

to non-interactively confirm

I

itstmillertime•6/9/23, 6:42 PM

@skye_31 What's the CI environment variable? CI=true?

MMatt Silverlock We need to accept the `-y` flag (not supported yet). You should be able to `echo...

I

itstmillertime•6/9/23, 6:42 PM

Wrapping it in an echo pipe is the secret?

Iitstmillertime @skye_31 What's the CI environment variable? CI=true?

M

Max (@rozenmd)•6/9/23, 6:47 PM

it should support these: https://github.com/watson/ci-info#supported-ci-tools

M

Max (@rozenmd)•6/9/23, 6:48 PM

otherwise hardcoding CI=1 or CI=true works

K

kian•6/9/23, 7:45 PM

It just means you didn't call

all()

all()

/

raw()

raw()

/

first()

first()

/etc

K

kian•6/9/23, 7:45 PM

Execute returned results - did you mean to call query?

I

itsmatteomanf•6/9/23, 8:36 PM

Hey @rozenmd, I think it's you that handles UI bugs from the dash, right?
If you have a table with

table

table

as the name of a column, it fails to delete rows from the UI. I presume it isn't adding

or

around column names.

I

itsmatteomanf•6/9/23, 8:37 PM

Not sure why it is specifying columns for row deletions, probably to select data?

M

marc•6/9/23, 11:43 PM

I've been trying out D1 recently paired with a Worker. How often are Reads and Writes updated on CF's dashboard? As it seems to be showing 0 for both even though I'm actively reading and writing from them.

Mmarc I've been trying out D1 recently paired with a Worker. How often are Reads and W...

M

Max (@rozenmd)•6/9/23, 11:56 PM

can be a few minutes - also I should add a tooltip that only experimental backend dbs have metrics

MMax (@rozenmd)can be a few minutes - also I should add a tooltip that only experimental backen...

M

marc•6/10/23, 12:01 AM

That makes sense - thank you ^^

MMatt Silverlock We just merged this internally. Had to do a bunch of work on our storage layer t...

I

itsmatteomanf•6/10/23, 10:12 PM