Question, if i bind 100 D1 servers to a worker and use KV to target a specific one. (lets say i run

__Ultima_Question, if i bind 100 D1 servers to a worker and use KV to target a specific o...

L

No, you only pay for rows read and written, and your storage used. https://developers.cloudflare.com/d1/platform/pricing/

Note: empty databases still consume tiny bit of storage for SQLite metadata (few KBs), although for your 100 DBs it's negligible.

P

parisholley•4/17/25, 1:30 AM

can someone clarify what "region" means with replicas? is a single datacenter being chosen within a region (balanced across customers presumably with some internal algorithm) to host the replica, or is the replica always going to be on the nearest datacenter (pop) to the user?

Pparisholley can someone clarify what "region" means with replicas? is a single datacenter be...

L

lambrospetrou•4/17/25, 9:16 AM

The heuristic and algorithm for replica placement is ongoing and will change over time.
The current placement is that it picks a random location within each region and all requests originating from that region will go to that replica. The usual availability applies for replicas too, if the datacenter does down, it will move somewhere else in the same region until the original datacenter comes back online.

This will change down the line, though, and become smarter. We explore several heuristics at the moment how it would look like.

P

parisholley•4/17/25, 2:14 PM

good to know. i have a multi-tenant application for local businesses, so outside of developers, 99% of traffic is always going to originate from the same metro per tenant/d1 instance (eg: texas company, california company, etc). having a replica in another region isn't particularly useful in my case but i do have a lot of async processing (using inngest) and being able to fire off jobs and append a bookmark id could unlock some performance (given that those tend to be read heavy). being able to pin D1 (and its replicas) to a DC (as the primary) would be

S

shaw•4/19/25, 12:37 AM

Having a very strange issue.

D1_ERROR: internal error; reference = ...

Returning empty array of items

But before, it was working for me and not for the people I know in Europe until they got on a VPN to america

S

shaw•4/19/25, 12:38 AM

I disabled caching, enabled dev mode, now I have this error

S

shaw•4/19/25, 12:46 AM

S

shaw•4/19/25, 12:57 AM

any help much appreciated, had to pause our launch since the database unexpectedly doesnt work

Sshaw Having a very strange issue. D1_ERROR: internal error; reference = ... Return...

C

Chaika•4/19/25, 1:02 AM

do you have read replication enabled? d1-database

CChaika do you have read replication enabled? https://discord.com/channels/5953179901913...

S

shaw•4/19/25, 1:45 AM

no

S

shaw•4/19/25, 1:45 AM

i had it enabled but disabled since we were getting these issues

Sshaw i had it enabled but disabled since we were getting these issues

C

Chaika•4/19/25, 2:18 AM

They've said before

FYI, disabling replicas is not instant. If replicas are created and process some of your queries, it can take a few hours for them to stop processing queries and completely disappear.

you said you're getting issues on a prod db? It still impacting, does creating a new db work around it, etc?

L

lbguilherme•4/19/25, 10:24 AM

Hello there! D1 currently isn't available on south america and on middle east, it will currently place the database on other region and route requests. Is there any roadmap or at least an intent of adding those regions in the future? Thanks!

K

Kiwi•4/19/25, 9:43 PM

I've been wavering between PG and D1 because of:

Transaction support
Customer support
Questions about performance
After some experimentation switching in hyperdrive I think I'm resolved that D1 is definitely best for speed, but still have concerns about the other points.
How have others found D1 in practice?

KKiwi I've been wavering between PG and D1 because of: - Transaction support - Custome...

J

James•4/20/25, 1:03 PM

I’d go with PG, personally. I shared my thoughts a while back at d1-database

D1 is always improving, but in my experience you will always have the most flexibility both in terms of technology and support with something like PG.

S

Successful-Arm-3762•4/21/25, 6:36 AM

Workers + D1 API responding very slow

I deployed a CF worker binded with a D1 database.
I know about the cold start delay, but all my APIs are taking approx. 2 second each.

I have an API to fetch 150 product data. That takes 20+ seconds. I optimized it to use batch queries and everything. I did EXPLAIN QUERY and the query itself takes 20ms.

The worker colo is "SIN" and D1 is in "APAC".
Could this be a D1 issue?

Would be grateful if somebody could help me out.

Screenshot_2025-04-21_at_11.48.08_AM.png

S

sayonara•4/21/25, 10:05 AM

Our d1 database is giving error
@Tanmay Kapoor @zegevlier

{
    "errors": [
        {
            "code": 7500,
            "message": "Currently processing a long-running export."
        }
    ],
    "success": false,
    "messages": [],
    "result": []
}

{
    "errors": [
        {
            "code": 7500,
            "message": "Currently processing a long-running export."
        }
    ],
    "success": false,
    "messages": [],
    "result": []
}

{
    "errors": [
        {
            "code": 7500,
            "message": "Currently processing a long-running export."
        }
    ],
    "success": false,
    "messages": [],
    "result": []
}

{
    "errors": [
        {
            "code": 7500,
            "message": "Currently processing a long-running export."
        }
    ],
    "success": false,
    "messages": [],
    "result": []
}

Original message was deleted

I

itsmatteomanf•4/21/25, 10:08 AM

?pings

S

SuperHelpflare•4/21/25, 10:08 AM

Please do not ping community members for non-moderation reasons. Doing so will not solve your issue faster and will make people less likely to want to help you.

S

sayonara•4/21/25, 10:18 AM

 "message": "internal error; reference = e_PVSMDp_85064edabe1b49f79d1f0a92065fcb4d"

"message": "internal error; reference = e_PVSMDp_85064edabe1b49f79d1f0a92065fcb4d"

 "message": "internal error; reference = e_PVSMDp_85064edabe1b49f79d1f0a92065fcb4d"

"message": "internal error; reference = e_PVSMDp_85064edabe1b49f79d1f0a92065fcb4d"

S

sayonara•4/21/25, 10:18 AM

for reference plz fix our db lol

S

Santi•4/21/25, 12:13 PM

D1_ERROR: Currently processing a long-running export.

S

Santi•4/21/25, 12:14 PM

Can someone help us out, our entire aplpication is down

S

sayonara•4/21/25, 12:58 PM

our case numnber-

01479222

01479222

01479222

01479222

S

sayonara•4/21/25, 12:58 PM

help plz

S

sayonara•4/21/25, 1:54 PM

@zegevlier ?

SSanti D1_ERROR: Currently processing a long-running export.

G

gruntlord6•4/21/25, 2:47 PM

Did you do a dump?

G

gruntlord6•4/21/25, 2:47 PM

I'm pretty sure that's how it's documented to function

G

gruntlord6•4/21/25, 2:50 PM

"A running export will block other database requests"
https://developers.cloudflare.com/d1/best-practices/import-export-data/

Cloudflare Docs

Import and export data · Cloudflare D1 docs

D1 allows you to import existing SQLite tables and their data directly, enabling you to migrate existing data into D1 quickly and easily. This can be useful when migrating applications to use Workers and D1, or when you want to prototype a schema locally before importing it to your D1 database(s).

Ggruntlord6 Did you do a dump?

S

sayonara•4/21/25, 2:52 PM

Yes but it hasn’t stopped since hours

Ggruntlord6 Did you do a dump?

S

Santi•4/21/25, 3:00 PM

Yes, but it has been going on for 6 hours now

G

gruntlord6•4/21/25, 3:19 PM

seems like a related issue. That said going forward I wouldn't do dumps on a live db since it blocks reads and can cause your application to become unavailable. I would copy it to another db and then dump that

G

gruntlord6•4/21/25, 3:20 PM

for now a ticket is your only real option since it seems to be an issue you both have

G

gruntlord6•4/21/25, 3:20 PM

the more tickets that are opened the more priority it gets

S

sayonara•4/21/25, 3:24 PM

we are same team

S

shaw•4/21/25, 4:26 PM

D1 doesn't seem usable, I think we should abandon cloudflare

S

shaw•4/21/25, 4:27 PM

Shockingly bad level of support for the amount of money being spent

S

shaw•4/21/25, 4:27 PM

Seems like the only solution is to tell our users we lost our data and never use cf again

Sshaw Seems like the only solution is to tell our users we lost our data and never use...

L

lambrospetrou•4/22/25, 9:23 AM

D1 never deletes your data. As others mentioned, during exports unfortunately the database is blocked from processing other queries until the export completes which depends on the size of your database.
I just checked your DB and it seems responsive to me at the moment, but has zero traffic, I assume you stopped using it.
Also, if you need urgent support in the future, reach out to Cloudflare Support or your account manager. This Discord channel is community driven, and even though we (the CF team) monitor it daily as much as we can, it's not the fastest way to get official support.

J

Jk3ndo•4/22/25, 7:08 PM

Hello guys,
I have created a worker which helps me to execute a cron which read data from a specific KV, and for each of those data make an API call and save the result into a database.

The response saved match the following interface:

export interface IWebsiteVerificationReport {
    domain: string,
    slug: string,
    verification: string,
    status: boolean,
    message: string,
    sanityDocId: string,
    audience: string,
    location?: string,
    runAt?: Date
}

export interface IWebsiteVerificationReport {
    domain: string,
    slug: string,
    verification: string,
    status: boolean,
    message: string,
    sanityDocId: string,
    audience: string,
    location?: string,
    runAt?: Date
}

export interface IWebsiteVerificationReport {
    domain: string,
    slug: string,
    verification: string,
    status: boolean,
    message: string,
    sanityDocId: string,
    audience: string,
    location?: string,
    runAt?: Date
}

export interface IWebsiteVerificationReport {
    domain: string,
    slug: string,
    verification: string,
    status: boolean,
    message: string,
    sanityDocId: string,
    audience: string,
    location?: string,
    runAt?: Date
}

I have created a class to handle the database related operations and it was working very well. see database file.

So as I said it was working very well, but as API call was too much I decided to include a Durable object to execute my API calls in queue. I added then the following line to the entry point file:

export { RequestQueue } from "./storage/requestQueue";

export { RequestQueue } from "./storage/requestQueue";

export { RequestQueue } from "./storage/requestQueue";

export { RequestQueue } from "./storage/requestQueue";

And for my d1 and durable object configuration in my wrangler toml I added those lines:

durable_objects.bindings = [{ name = "REQUEST_QUEUE", class_name = "RequestQueue" }]

d1_databases = [
  { binding = "DB", database_name = "lightmode-store", database_id = "c1766ce5-234c-401b-a854-d614c5c1d300" },
]

[[migrations]]
tag = "v1"
new_classes = ["RequestQueue"]
new_databases = ["lightmode-store"]

durable_objects.bindings = [{ name = "REQUEST_QUEUE", class_name = "RequestQueue" }]

d1_databases = [
  { binding = "DB", database_name = "lightmode-store", database_id = "c1766ce5-234c-401b-a854-d614c5c1d300" },
]

[[migrations]]
tag = "v1"
new_classes = ["RequestQueue"]
new_databases = ["lightmode-store"]

durable_objects.bindings = [{ name = "REQUEST_QUEUE", class_name = "RequestQueue" }]

d1_databases = [
  { binding = "DB", database_name = "lightmode-store", database_id = "c1766ce5-234c-401b-a854-d614c5c1d300" },
]

[[migrations]]
tag = "v1"
new_classes = ["RequestQueue"]
new_databases = ["lightmode-store"]

durable_objects.bindings = [{ name = "REQUEST_QUEUE", class_name = "RequestQueue" }]

d1_databases = [
  { binding = "DB", database_name = "lightmode-store", database_id = "c1766ce5-234c-401b-a854-d614c5c1d300" },
]

[[migrations]]
tag = "v1"
new_classes = ["RequestQueue"]
new_databases = ["lightmode-store"]

However, after having added the durable object to it, at the step of the database insertion data I get the error "Error: TypeError: dba.prepare is not a function".
When I stringify the D1 database to inspect it I get the following object:

{
  "alwaysPrimarySession": {
    "fetcher": {},
    "bookmarkOrConstraint": "first-primary"
  },
  "fetcher": {}
}

{
  "alwaysPrimarySession": {
    "fetcher": {},
    "bookmarkOrConstraint": "first-primary"
  },
  "fetcher": {}
}

{
  "alwaysPrimarySession": {
    "fetcher": {},
    "bookmarkOrConstraint": "first-primary"
  },
  "fetcher": {}
}

{
  "alwaysPrimarySession": {
    "fetcher": {},
    "bookmarkOrConstraint": "first-primary"
  },
  "fetcher": {}
}

Can someone help me resolve the issue please ?
Thanks,

KKent C. Dodds Hey friends, I've never used Cloudflare D1 before, but I'm looking into it. Has ...

G

gnarley_farley.•4/23/25, 2:54 AM

Convert over to drizzle that’s what I’m currently doing. Prisma is not compatible with it.

JJk3ndo Hello guys, I have created a worker which helps me to execute a cron which read ...

L

lambrospetrou•4/23/25, 8:31 AM

What is

new_databases

new_databases

new_databases

new_databases under the

[[migrations]]

[[migrations]]

[[migrations]]

[[migrations]]? The

[[migrations]]

[[migrations]]

[[migrations]]

[[migrations]] block is for the Durable Objects, not D1. See https://developers.cloudflare.com/durable-objects/reference/durable-objects-migrations/

Also, from the error

Error: TypeError: dba.prepare is not a function

Error: TypeError: dba.prepare is not a function

Error: TypeError: dba.prepare is not a function

Error: TypeError: dba.prepare is not a function it seems you are using

dba

dba

dba

dba as the D1 Binding variable, but your actual binding in your wrangler toml seems to be defined as

DB

DB

DB

DB, so without the code you are actually running I am not sure what you are doing wrong.

Z

Zack•4/23/25, 11:25 AM

Just wondering if there is an update on the fixes that were discussed for d1 replicas. I believe the original estimate was end of last week for a rollout, did that happen silently or are those fixes still pending release?

ZZack Just wondering if there is an update on the fixes that were discussed for d1 rep...

L

lambrospetrou•4/23/25, 11:52 AM

A preliminary quick fix was rolled out indeed last week that would improve certain cases, and the main fix is currently gradually rolling out but will take a few days to make sure nothing else breaks.

Llambrospetrou A preliminary quick fix was rolled out indeed last week that would improve certa...

Z

Zack•4/23/25, 11:53 AM

Great to hear, would you be able to post here when that is fully rolled out? I would love to test replicas again after that with a 50/50 rollout in our prod environment

Llambrospetrou What is `new_databases` under the `[[migrations]]`? The `[[migrations]]` block i...

J

Jk3ndo•4/23/25, 12:01 PM

Yes sure, here is the class responsible of database operations
I have removed as well the D1 database from migrations section but it still provide the same issue.
@nora , here attached you can find my function where I use the prepare method (websiteVerificationReport.ts) and the entry point of the worker.

ZZack Great to hear, would you be able to post here when that is fully rolled out? I w...

L

lambrospetrou•4/23/25, 12:02 PM

I will post here, yes. Although, again, I do strongly suggest you try it on a staging/test environment first.

Llambrospetrou I will post here, yes. Although, again, I do strongly suggest you try it on a st...

Z

Zack•4/23/25, 12:02 PM

Yea I will do ofc

Question, if i bind 100 D1 servers to a worker and use KV to target a specific one. (lets say i run

Similar Threads

Similar Threads

Similar Threads