Following https://blog.cloudflare.com/d1-read-replication-beta/, CF team is starting to think about

Following https://blog.cloudflare.com/d1-read-replication-beta/, CF team is starting to think about how replication looks for Durable Objects. If you have a use case for DO read replication, reply to this message (feel free to share your use case too) and I'll reach out separately.

Ggruntlord6 the deploy button doesn't work, not sure if thats your template or a limit of du...

lambrospetrou•4/29/25, 3:54 PM

You need to update that to your domain.

Llambrospetrou You need to update that to your domain.

gruntlord6•4/29/25, 3:54 PM

is it possible to update the template to use the worker domain or is that just something that always has to be done manually on the users end

Ggruntlord6 is it possible to update the template to use the worker domain or is that just s...

lambrospetrou•4/29/25, 3:55 PM

You need to remove this line: https://github.com/janwilmake/dorm/blob/a818944b8a55690bed5d209011dae616a832b1e2/wrangler.jsonc#L7
Assuming that's what you try to deploy.

Llambrospetrou You need to remove this line: https://github.com/janwilmake/dorm/blob/a818944b8a...

gruntlord6•4/29/25, 3:56 PM

that would make the deploy button work without user intervention?

Edit: Thank you, as always very helpful

eckoln•4/29/25, 4:08 PM

Hey there. What is the cost of messages in internal websocket connection between DOs? Is it included in incoming messages?
In my usecase, i want to use DO to DO internal websocket connection. I will send one client message to between (3+) DOs. In this case, will 1 message be counted as 3+ messages?

Eeckoln Hey there. What is the cost of messages in internal websocket connection between...

lambrospetrou•4/29/25, 4:52 PM

For compute requests billing-only, a 20:1 ratio is applied to incoming WebSocket messages to factor in smaller messages for real-time communication. For example, 100 WebSocket incoming messages would be charged as 5 requests for billing purposes.

From https://developers.cloudflare.com/durable-objects/platform/pricing/

veeque•4/29/25, 7:55 PM

Is there a way to "export" data from all DOs? Say I were to build my app using DOs, but then for some reason I need to move to some other infra, can I export the SQL data stored in every DO for migration purposes?

Vveeque Is there a way to "export" data from all DOs? Say I were to build my app using D...

boop•4/29/25, 8:44 PM

Not natively, you'll have to implement that logic yourself

Llambrospetrou Yes, that's common. You can avoid the alarms by only having active DOs "heartbea...

Fernando Barroso•4/29/25, 11:29 PM

by heartbeat you mean whenever a significant action is done it calls back to the parent?

Vveeque Is there a way to "export" data from all DOs? Say I were to build my app using D...

gruntlord6•4/29/25, 11:32 PM

I think they shared a script here that does what the d1 export does

jcheese•4/30/25, 12:05 AM

I am getting

for a specific DO ID, other IDs in the same namespace works fine. can anyone help?

Jjcheese I am getting ``` GET http://dummy-example.cloudflare.com/cdn-cgi/partyserver/se...

Larry•4/30/25, 12:40 AM

My first suspicion would be that there is something in the data for the instance that is causing it. I suggest wrapping your fetch, onMessage, onRequest, etc methods with a try/catch and logging the details.

LLarry My first suspicion would be that there is something in the data for the instance...

jcheese•4/30/25, 12:42 AM

yeah, ive done that but i keep getting

Kiwi•4/30/25, 5:57 AM

How do I go about debugging ?

Grzkpl•4/30/25, 6:03 AM

Fix Needed for Bug in CF Email Notification System general-help[Cloudflare]: [ACTION REQUIRED] Daily usage limit exceeded for Cloudflare Durable Objects

VVy Following https://blog.cloudflare.com/d1-read-replication-beta/, CF team is star...

Danny Navarro•4/30/25, 7:36 AM

More than a direct use case, I was hopeful read replication would enable dynamic relocation of DOs. If a worker accesses the DO from another data center it gets read replicated right there. After the DO gets evicted, next time it's accessed, if there was already a read replica in the new data center, that replica gets promoted to write leader.

Also, it'd be nice to have locally replicated DO sqlite files for debugging. Maybe with something like

DDanny Navarro More than a direct use case, I was hopeful read replication would enable dynamic...

lambrospetrou•4/30/25, 9:23 AM

Also, it'd be nice to have locally replicated DO sqlite files for debugging. Maybe with something like wrangler do --namespace <ns> read <id> <db.sqlite>

You mean, doing an export/dump/copy right once? You don't expect things to be replicated locally in real time when changes happen remotely.

After the DO gets evicted, next time it's accessed, if there was already a read replica in the new data center, that replica gets promoted to write leader.

This is something that we are thinking about but not as part of read replication. Moving the primary should take into account more than just a single request, so it wouldn't work like that, but yeah the team at some points wants to be able to move primary DOs.

Llambrospetrou > Also, it'd be nice to have locally replicated DO sqlite files for debugging. M...

Danny Navarro•4/30/25, 9:34 AM

You mean, doing an export/dump/copy right once? You don't expect things to be replicated locally in real time when changes happen remotely.

Right, I don't really care about real-time. I was thinking about explicitly doing a one way sync with a command, maybe

pull

pull

is a better word. A command to export just once would be great, the one way sync would be just an optimization.

VVy Following https://blog.cloudflare.com/d1-read-replication-beta/, CF team is star...

John Spurlock•4/30/25, 1:45 PM

long time wish: export/backup/archive DO with read-only access to another DO's storage but without competing with the main DOs incoming request queue - a minimal replication delay would be fine

veeque•4/30/25, 4:15 PM

why is the 10gb limit a hard limit? it's not like the data has to travel to other locations, given that a DO only ever exists in the datacenter it was created.

Vveeque why is the 10gb limit a hard limit? it's not like the data has to travel to othe...

lambrospetrou•4/30/25, 4:26 PM

Few reasons, cold starts would be very slow and it would be hard to get the fast SQLite performance from local queries if you had to load pages from remote most of the time (even with caching). Also, the fact that the DOs do not move (yet), is just because it's not implemented, yet.

Llambrospetrou Few reasons, cold starts would be very slow and it would be hard to get the fast...

jed•4/30/25, 6:45 PM

is this still in the works? it’s been almost five years since the announcement that cloudflare “can transparently migrate objects between locations.”

Jjed is this still in the works? it’s been almost five years since the announcement ...

lambrospetrou•4/30/25, 7:05 PM

There is transparent migration already, but only when the original location is down, and not due to request traffic.

JJan Wilmake thanks. will remoive it, idk how to fix it either

gruntlord6•4/30/25, 8:27 PM

I think they said this will fix it durable-objects

jcheese•5/1/25, 1:04 AM

will DOs re-initialize whenever theres a new deployment that changes code? meaning, will the

constructor

constructor

get called everytime when theres a new deployment and the DO gets accessed?

Jjcheese will DOs re-initialize whenever theres a new deployment that changes code? meani...

1984 Ford Laser•5/1/25, 1:05 AM

No but any open/hibernated Websockets will be terminated

11984 Ford Laser No but any open/hibernated Websockets will be terminated

jcheese•5/1/25, 1:08 AM

i see, but how can a js class have it's underlying code updated to even receive the new migrations in the file w/o actually re-initializing the class?

1984 Ford Laser•5/1/25, 1:09 AM

It will get called the next time the DO is accessed, but not immediately upon deployment. Sorry if thats what you meant

11984 Ford Laser It will get called the next time the DO is accessed, but not immediately upon de...

jcheese•5/1/25, 1:15 AM

i see. then do you know why for migration libs like this: https://github.com/lambrospetrou/durable-utils?tab=readme-ov-file#sqlite-schema-migrations

require

await this._migrations.runAll();

await this._migrations.runAll();

in all sql queries, instead of a one off

        await this._migrations.runAll();

        await this._migrations.runAll();

in the constructor?

GitHub

GitHub - lambrospetrou/durable-utils: Utilities for Cloudflare Dura...

Utilities for Cloudflare Durable Objects and Workers - lambrospetrou/durable-utils

Jjcheese i see. then do you know why for migration libs like this: https://github.com/lam...

Larry•5/1/25, 1:46 AM

lambrospetrou favors doing it "lazily" in case you have functionality in your DO that doesn't need stored data. If that's not the case for you, doing it in the constructor is fine. That's what I do.

LLarry lambrospetrou favors doing it "lazily" in case you have functionality in your DO...

jcheese•5/1/25, 1:48 AM

thanks!

that doesn't need stored data

what does this mean necessarily?

Jjcheese thanks! > that doesn't need stored data what does this mean necessarily?

Larry•5/1/25, 1:51 AM

Let's say you have a method that accepts two numbers and returns the sum. No stored data is involved.

LLarry Let's say you have a method that accepts two numbers and returns the sum. No sto...

jcheese•5/1/25, 1:51 AM

ah yea i get that, but if you have migrations in your DO, you always have stored data im assuming?

Jjcheese ah yea i get that, but if you have migrations in your DO, you always have stored...

Larry•5/1/25, 1:53 AM

Yes, but part of your DO's functionality might not use stored data. A commonly accepted software engineering principle is to not do work until it's needed.

LLarry Yes, but part of your DO's functionality might not use stored data. A commonly a...

jcheese•5/1/25, 2:12 AM

i see i see. thank you!

LLarry lambrospetrou favors doing it "lazily" in case you have functionality in your DO...

lambrospetrou•5/1/25, 9:43 AM

Exactly what @Larry said. I need to update the README of my library to explain this since many folks ask about it at this point. It seemed a bit obvious to me, but I was wrong assuming that.

Llambrospetrou Exactly what @Larry said. I need to update the README of my library to explain t...

jcheese•5/1/25, 11:23 AM

The internals of the DO and when constructor gets called and stuff isnt so obvious

jcheese•5/1/25, 11:24 AM

Thanks though, amazing library

VVy Following https://blog.cloudflare.com/d1-read-replication-beta/, CF team is star...

DaniFoldi•5/1/25, 11:24 AM

I'm still hoping one day we'll have DOs in all colos, even if storageless - the use case is for websocket handling and chat rooms where persistence is external anyways

DDaniFoldi I'm still hoping one day we'll have DOs in all colos, even if storageless - the ...

Zack•5/1/25, 11:29 AM

Isn't storageless DO just a worker?

ZZack Isn't storageless DO just a worker? 🤔

DaniFoldi•5/1/25, 11:30 AM

you can't use hibernation, connect multiple websockets to the same worker, so not quite

Zack•5/1/25, 11:30 AM

Ah I see

DaniFoldi•5/1/25, 11:30 AM

but I see where you're coming from, for a simple proxy, or maybe a filtering/logging service, yeah regular workers are just fine (even better for some things since no wall time charges)

ZZack Isn't storageless DO just a worker? 🤔

lambrospetrou•5/1/25, 11:34 AM

The big benefit of DOs over workers even without storage is coordination due to the global uniqueness of the DO ID.

Tobias•5/1/25, 6:44 PM

We just had a customer having an outage on an important realtime-flow because a DO was hanging for >1h - no requests timed out, they were hanging in pending state.

I'm a bit out of ideas here - since there are no errors, or no indicators on why this suddenly happened.

A re-deploy fixed this, and all connections were reset.

Tobias•5/1/25, 6:45 PM

(This wall time is from an alarm in a different DO that uses RPC to connect here)

Tobias•5/1/25, 6:46 PM

In the morning we had requests erroring with an overload error - self-recovered just fine after the load was gone. But this time, we're a bit clueless - especially since there was no self recovery :/

Tobias•5/1/25, 6:47 PM

The DB in the DO is still relatively small for DB standards, with just 50k records. Nothing that should hit the limit.