You're right, my initial worry was about cold starts. I guess there are two scenarios I'm wondering

You're right, my initial worry was about cold starts. I guess there are two scenarios I'm wondering about: when a read replica is first created in a new data centre, and when an existing read replica has to be cold started. Which leaves me with three questions: 1. what's the heuristic for creating a read replica in a new data centre? 2. how quickly can a new read replica be created and booted (not looking for precise numbers, just a way to think about the impact of sqlite storage size)? 3. what affects the cold start time for a read replica (sqlite size, amount of data to be updated, something else)? Apologies if I'm asking silly questions, just trying to get my head around how to think about this.

Sscruffy You're right, my initial worry was about cold starts. I guess there are two scen...

lambrospetrou•11/17/25, 1:37 PM

The first a replica (or DO in general) is created in a region it's assigned a specific datacenter location in the region randomly. Then, every time that specific replica comes online it will start there, if that DC is available and healthy, or moved elsewhere in the same region temporarily. This behavior is what happens NOW, and is NOT a guarantee we provide. We want to be flexible to move replicas in the future, since that's the plan, to dynamically create and move replicas based on incoming traffic.
The main delay is fetching the DB from cold storage. That strictly depends on the size. It can be a few milliseconds or many seconds.
DB size is the main contributor.

Llambrospetrou 1. The first a replica (or DO in general) is created in a region it's assigned a...

scruffyOP•11/17/25, 1:43 PM

This is really useful, thanks. Do read replicas only get created when there's a certain amount of traffic in a given region, and if so, what level of traffic is required? Does the delay scale linearly on DB size and do you have any very rough ballpark delays for say 1GB?

Sscruffy This is really useful, thanks. Do read replicas only get created when there's a ...

lambrospetrou•11/17/25, 2:45 PM

The replicas will start booting on the first request to its region, and until they are ready requests are forwarded to the primary. So, if you are only doing 1 request per minute for example, it's unlikely you will get processed by a replica ever, but you will always keep waking it up.
The delay depends on the size yes, I don't have numbers to publicly state, but it's as like downloading N MBs from cold storage.

You can find more details about how it all works under the hood in our blog post: https://blog.cloudflare.com/d1-read-replication-beta/#under-the-hood-how-d1-read-replication-is-implemented

The Cloudflare Blog

Sequential consistency without borders: How D1 implements global re...

D1, Cloudflare’s managed SQL database, announces read replication beta. In this post, we do a technical deep dive of the read replication implementation and how your queries can remain consistent across all regions.

Llambrospetrou The replicas will start booting on the first request to its region, and until th...

scruffyOP•11/18/25, 10:33 AM

thanks for your help, much appreciated

ts•11/19/25, 10:27 PM

Thanks for the great product. I’m hitting an issue with D1 where very simple inserts are taking anywhere from one to eight seconds. I’m only inserting a single user row and returning it. The table has the primary key index and a unique index on email, nothing else. The SQL itself runs in a few milliseconds, but the D1 trace shows long delays before the statement actually completes, which makes it look like platform-level latency rather than anything in the query.

I’m also seeing the same kind of latency spikes on KV puts from time to time. I’m trying to understand whether there are plans to reduce these stalls, or if there’s anything specific I can do on my end to improve D1 insert or KV write performance.
Thanks in advance

fawwaz•11/20/25, 6:52 AM

Hey all I'm excited about the new release of tanstack db (v1 expected to be realsed next month), I'm curious if any one yet used with d1?

fawwaz•11/20/25, 6:53 AM

there is a concern that progressive mode would get expensive with traditional fetch but I'm thinking that cache-tags would solve this, what do you guys think?
https://developers.cloudflare.com/cache/how-to/purge-cache/purge-by-tags/

Cloudflare Docs

Purge cache by cache-tags

Cache-tag purging makes multi-file purging easier because you can instantly bulk purge by adding cache-tags to your assets, such as webpages, image files, and more.

Tts Thanks for the great product. I’m hitting an issue with D1 where very simple ins...

Florent•11/20/25, 9:36 AM

I sent you a DM to get more info, so that I can take a look a your specific D1 DB.

Matthew•11/21/25, 9:22 AM

Is there any indication of when Time Travel might gain the ability to clone and fork databases? Understand you won't have a precise date, but helpful to know if it's closer to "next quarter" or "maybe one day"

https://developers.cloudflare.com/d1/reference/time-travel/#restore-a-database

In the future, D1 will support branching & cloning databases using Time Travel.

At the moment, backups feel like a big missing piece.

Time travel destructively overwrites the entire database - so restoring one deleted row means losing all good data newer than the bookmark.
wrangler exportwrangler export takes the db offline, potentially for multiple minutes depending on size.
I could create a workflow to query tables one by one and output a .sql.sql file, but there could be inconsistent data between tables, foreign key problems, etc.
Maybe I'm missing an option?

MMatthew Is there any indication of when Time Travel might gain the ability to clone and ...

lambrospetrou•11/21/25, 9:49 AM

Asked a few times every month, we do know people want it, we do have plans to do it, but unfortunately cannot promise any delivery estimates.

We are hoping that over the next few quarters we will be switching back to doing all the read replication improvements we want, and that will enable features like this.

Llambrospetrou Asked a few times every month, we do know people want it, we do have plans to do...

Matthew•11/21/25, 10:25 AM

Thanks!

tobystr•11/21/25, 11:28 AM

That and also Change Data Capture to a pipeline / queue would be nice

Earl•11/21/25, 6:59 PM

Is there a possibility to do a D1 restore via HTTP API or bindings API?

I can get the current bookmark at runtime via the binding but I see no way to restore to a bookmark except via wrangler, which is not available at runtime.

EEarl Is there a possibility to do a D1 restore via HTTP API or bindings API? I can ...

Florent•11/21/25, 11:35 PM

There might be a way but it's not documented in our OpenAPI. I'll check with the team if I can get more information but I can't guarantee you a response.

Earl•11/21/25, 11:37 PM

That would be great

currently I'm looking into using a durable object container to run wrangler. But this is pretty hacky.

Florent•11/21/25, 11:38 PM

What would be your use case for restoring a database via an HTTP api call?

FFlorent What would be your use case for restoring a database via an HTTP api call?

Earl•11/22/25, 12:47 AM

I've been talking about it with lambrospetrou in #durable-objects .

Quick summary: I'm processing live data and sometimes that data is faulty and I need to revert to an earlier point and replay the corrected data.

With durable objects that's possible at runtime. But I also have D1s.

I'm trying to avoid writing revert logic for all the incoming data since this is difficult to maintain.

Paul•11/22/25, 7:54 PM

Does D1 have triggers for tables? Like can I raise an event to a Queue when a insert/update/delete happens?

Earl•11/22/25, 11:04 PM

Only internal SQLite triggers, you can't call outside. Maybe durable objects would be a better fit?

PPaul Does D1 have triggers for tables? Like can I raise an event to a Queue when a in...

lambrospetrou•11/23/25, 12:58 AM

This feature is commonly called CDC (change data capture), and it's not supported at the moment. The SQLite triggers feature is supported, but that's only to execute SQL things when certain rows change, so not what you are asking.

PPaul Does D1 have triggers for tables? Like can I raise an event to a Queue when a in...

Hello, I’m Allie!•11/23/25, 1:07 AM

Dumb idea, use SQLite triggers to push to a triggers table, and then periodically poll it?

Aman•11/23/25, 1:08 AM

I am facing an issue where I am unable to access some parts of my site
For example, when I try to remove something from the cart, it throws an error and does not let me remove it, additionally it does not let me logout

When used in development mode, it works fine
I purged cache and all but still not working, any insights?

makima•11/23/25, 5:37 AM

hi all is it possible to use d1 per user, the limit is 50k databases what if i have morethen 50k users , can you increase it and also its 0.75 dollars for a gb after 5gb does this inclulde each database ?

Mmakima hi all is it possible to use d1 per user, the limit is 50k databases what if i h...

David•11/23/25, 7:15 AM

I'm curious, why a separate DB per user?

Mmakima hi all is it possible to use d1 per user, the limit is 50k databases what if i h...

Kevin•11/23/25, 7:59 AM

use durable objects for per user storage. there is no limit on durable objects count. it uses the same storage as D1.

Mmakima hi all is it possible to use d1 per user, the limit is 50k databases what if i h...

lambrospetrou•11/23/25, 9:50 AM

The limit can be increased, yes, but it's going to be a pain to bind them all to the same worker. If you want a single worker to have access to thousands/millions of DBs, as others suggested using SQLite Durable Objects is the recommended approach.

AAman I am facing an issue where I am unable to access some parts of my site For examp...

lambrospetrou•11/23/25, 9:50 AM

Are you using D1, and if yes, are you seeing errors from D1? Otherwise you are probably going to get more help from the right channel for what's failing in your setup.

Cemal•11/23/25, 11:43 AM

Hi, my D1 queries aren't working with Vitest. Is there a solution for this?

Llambrospetrou The limit can be increased, yes, but it's going to be a pain to bind them all to...

makima•11/23/25, 12:07 PM

appreacite the feedback, i needed much more storage for my db so i had a look around and found hyperdrive and planetscale, would this be a good setup , im using nextjs with cloudflare workers everything works but i need a scalable db.

Llambrospetrou The limit can be increased, yes, but it's going to be a pain to bind them all to...

makima•11/23/25, 12:18 PM

planetscale is pricy though , how scalable and effective is this setup that you have reccomended , is per user db even a good approach , i know i will need way more then 10gb of storage but i dont really know which approach to use

Mmakima planetscale is pricy though , how scalable and effective is this setup that you ...

Larry•11/23/25, 12:55 PM

If you need more than 10gb per user, don't go the DO route. Sharding will be complexity you probably don't want to deal with. I recommend using Cloudflare Hyperdrive to connect to a Postgres database hosted somewhere. https://developers.cloudflare.com/workers/tutorials/postgres/

Cloudflare Docs

Connect to a PostgreSQL database with Cloudflare Workers

This tutorial explains how to connect to a Postgres database with Cloudflare Workers. The Workers application you create in this tutorial will interact with a product database inside of Postgres.

LLarry If you need more than 10gb per user, don't go the DO route. Sharding will be com...

makima•11/23/25, 12:57 PM

i was looking at this, although planetscale is pricy i see they do have single nodes now, how effective is hyperdrive though , is the latency good

Mmakima i was looking at this, although planetscale is pricy i see they do have single n...

Larry•11/23/25, 1:04 PM

I've heard people on here say that latency was surprisingly good. One person said it was faster than D1 but I think that's because it does some caching. Cloudflare uses Postgres themselves although I expect that they host it themselves.

LLarry I've heard people on here say that latency was surprisingly good. One person sai...

makima•11/23/25, 1:08 PM

i also heard they use it internally and theres even docs for planetscale, im not sure what to do exactly , it would be nice to stick to cloudflare infra for all things, the thing that im not even sure if user will ever get to 10gb of data, its a second brain type of applicaton, like a advance todo type of appilcation , what i though is that having a db or do for each user would be the best setup but im not really experienced with this so im jsut trying to put together different infras and see how it is.

Mmakima i also heard they use it internally and theres even docs for planetscale, im not...

Larry•11/23/25, 3:34 PM

Definitely use DOs not D1 if it's per user. You will struggle with binding otherwise. The nice thing about going the DO route also is that you can encapsulate everything that your client code needs into methods in the DO. For now, they can use the SQLite 10GB DB inside of the DO. You can monitor and see how fast they are growing. If one of your users reaches 5GB in a year, then you'll have a year to migrate the storage part somewhere else, or figure out how to shard. I think you'll be surprised how far 10GB goes. One bad thing about 5GG+ dbs in DOs though is that cold starts can be 1+ seconds if Cloudflare has to move the db file to the machine where the DO runs.

Panchi•11/24/25, 9:04 AM

yo anyone why dis this happen and how do i fix? pls fm

Panchi•11/24/25, 9:05 AM

Paul•11/25/25, 1:52 AM

Looking for any gotchas or hidden costs or suprises for D1 global table replication. The website docs seem to say it's 100% free (no additional costs), for example: "Pricing
D1 read replication is built into D1, so you don’t pay extra storage or compute costs for read replicas. You incur the exact same D1 usage billing with or without replicas, based on rows_read and rows_written by your queries." - but how could that be true? If the data is replicated to say 6 other global regions, why would the price be exactly the same as the price for a single instance in a single region? What am I missing?

PPaul Looking for any gotchas or hidden costs or suprises for D1 global table replicat...

Chaika•11/25/25, 3:26 AM

There's a lot of cool information here: https://blog.cloudflare.com/d1-read-replication-beta/
The catch I would say is that you have to send requests (one every 1-2 mins or so) to keep the replica active. Otherwise, it sleeps. If a request is sent from a region with a sleeping replica, it just goes to the primary until it wakes up again. Replicas boot from cold storage and replay writes/etc to get up to date, and then say in sync actively until they go to sleep.
It's not like a constant active-active setup though unless all regions are constantly awake, and the data isn't really replicated until it's active.
I would also take that description as they planned for Read Replication since the start so already baked that into the costs.

CChaika There's a lot of cool information here: <https://blog.cloudflare.com/d1-read-rep...

lambrospetrou•11/25/25, 10:31 AM

What Chaika said. D1's pricing included certain number of replicas to be included, hence why there isn't an extra charge. In the future, when there is going to be an arbitrary number of replicas maybe an extra charge will be added for "extras above a certain limit" but this is mostly speculation, we haven't decided anything yet.

Also, you anyway pay for row operations done on replicas, so it's not that you don't give something for using lots of them.

PPaul Looking for any gotchas or hidden costs or suprises for D1 global table replicat...

lambrospetrou•11/25/25, 11:13 AM

Ah, also if by "gotchas" you meant anything other than pricing as well, then I would say that the main limitation right now is that replicas are fixed, one per region. This means that unless your load is from all over the world, you will mostly get a a single replica per region helping with the reads, which is not great if most of your traffic is in a single region.
This is something we are planning on extending, hopefully next quarter(s), and allow plenty more replicas per region.
And what Chaika mentioned that without constant traffic replicas hibernate.

Llambrospetrou What Chaika said. D1's pricing included certain number of replicas to be include...

Paul•11/26/25, 1:28 AM

Just to clarify, when you say "you anyway pay for row operations done on replicas" - does that mean you pay for all the inserts/updates/deletes that are being syncronized to the read-only copies? Like do you pay for CRUD operations happening in each replicated database instance? OR do you still JUST pay the EXACT SAME price you would for rows written to the primary, NO change if you have replication enabled or not.

CChaika There's a lot of cool information here: <https://blog.cloudflare.com/d1-read-rep...

Paul•11/26/25, 1:30 AM

Ok great to understand the cold start behavior, thank you.

PPaul Just to clarify, when you say "you anyway pay for row operations done on replica...

Isaac McFadyen•11/26/25, 3:30 AM

Exact same. 1 read anywhere = 1 read, 1 write on primary = 1 write.

divby0•11/26/25, 8:00 AM

https://github.com/cloudflare/workers-sdk/issues/5438

It says that last month @penalosa added the feature issue type to this. Does that mean its actively being worked on? Because I basically can't use normal relational database design principles with the sqlite limitations of no major alterations without recreating tables combined with the limitations that defering foreign keys does not defer the ON DELETE CASCADE action

IIsaac McFadyen Exact same. 1 read anywhere = 1 read, 1 write on primary = 1 write.

lambrospetrou•11/26/25, 10:16 AM

You don't pay for the replicated writes, you pay for all row READS on replicas and primary, and all row WRITES on the primary.

makima•11/26/25, 10:20 AM

a follow on on the durable objects per user setup , i have somewhat implemented it , how do i even test it though ? im not fully versed in dbs in general. my setup is simply a user signups for an account and this automatically sets up a DO for them. thats what i have so far. it works with a single user(me) but i have zero idea how to test it at scale, not that i will have a so many users of the bat, but its good to test, any suggestions

Matthew•11/26/25, 11:15 PM

Something weird is going with D1 performance for me. Observability shows that every 10 minutes or so, the P99 worker wall time jumps to 30-40 seconds, from a baseline of 100-200ms, and the durationduration of d1_batchd1_batch traces shows the same pattern. Going back the last seven days shows the problem gradually worsening, then fixing itself, then getting worse. My query latency has remained low throughout, at <3ms. My database ID is 585d290a-44f4-4f54-b265-00b35317ae41585d290a-44f4-4f54-b265-00b35317ae41.

Matthew•11/26/25, 11:16 PM

The second and third images graph the P99 and P95 durationMS of d1_batch calls.