I had to create a new API Key..

O

OndrejD•6/24/25, 6:59 PM

Hmm, not sure where would I need to put the API key... I tried to log in again with wrangler - no change. FYI I'm on the paid plan

O

OndrejD•6/24/25, 7:01 PM

Maybe it's still rolling out and my account isn't on it yet?

N

Niek•6/24/25, 7:06 PM

same for me, i tried logging out and in. for me it loops like this:

Login Succeeded
Image does not exist remotely, pushing: registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer:5f39f398
The push refers to repository [registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer]
e303a827c75d: Retrying in 3 seconds

Login Succeeded
Image does not exist remotely, pushing: registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer:5f39f398
The push refers to repository [registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer]
e303a827c75d: Retrying in 3 seconds

Login Succeeded
Image does not exist remotely, pushing: registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer:5f39f398
The push refers to repository [registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer]
e303a827c75d: Retrying in 3 seconds

Login Succeeded
Image does not exist remotely, pushing: registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer:5f39f398
The push refers to repository [registry.cloudflare.com/3ac73ad157bc18d2890dfd70f000bd07/mycontainer]
e303a827c75d: Retrying in 3 seconds

N

Niek•6/24/25, 7:38 PM

ah:

413 Request Entity Too Large

413 Request Entity Too Large

413 Request Entity Too Large

413 Request Entity Too Large - even though the layer is under 2GB. i guess there is a lower per-layer limit?

K

Khagan•6/24/25, 7:47 PM

Is it possible to trigger cron from another container? For example I got django app in a container. Can I trigger a delayed job from it?

Y

Yoh•6/24/25, 8:06 PM

If I’m already running 'wrangler dev' inside of docker (we have a monorepo where each worker in a monorepo runs in its own docker container), is there any chance this workflow would be supported for workers using cloudflare containers?

C

Constantin•6/24/25, 8:29 PM

hey there! i packed my image as a docker image it's an Elixir Phoenix build with

mix phx.gen.release --docker

mix phx.gen.release --docker

mix phx.gen.release --docker

mix phx.gen.release --docker .. added a wrangler.tml file with a basic config

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "silent-unit-dbda_MyContainer"
class_name = "MyContainer"

[[migrations]]
tag = "v1"
new_sqlite_classes = ["MyContainer"]

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "silent-unit-dbda_MyContainer"
class_name = "MyContainer"

[[migrations]]
tag = "v1"
new_sqlite_classes = ["MyContainer"]

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "silent-unit-dbda_MyContainer"
class_name = "MyContainer"

[[migrations]]
tag = "v1"
new_sqlite_classes = ["MyContainer"]

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "silent-unit-dbda_MyContainer"
class_name = "MyContainer"

[[migrations]]
tag = "v1"
new_sqlite_classes = ["MyContainer"]

and created the durable object on CF side.

however, when i'm trying to run

npx wrangler deploy

npx wrangler deploy

npx wrangler deploy

npx wrangler deploy i'm getting this error:

✘ [ERROR] Your Worker depends on the following Durable Objects, which are not exported in your entrypoint file: MyContainer.

✘ [ERROR] Your Worker depends on the following Durable Objects, which are not exported in your entrypoint file: MyContainer.

✘ [ERROR] Your Worker depends on the following Durable Objects, which are not exported in your entrypoint file: MyContainer.

✘ [ERROR] Your Worker depends on the following Durable Objects, which are not exported in your entrypoint file: MyContainer.

CConstantin hey there! i packed my image as a docker image it's an Elixir Phoenix build wit...

G

gruntlord6•6/24/25, 8:31 PM

did you deploy it already before?

G

gruntlord6•6/24/25, 8:31 PM

usually that means you just need to remove the migration

G

gruntlord6•6/24/25, 8:38 PM

or if its a new migration it means you already use that migration version

C

Constantin•6/24/25, 8:38 PM

looks like creating a DO should be a code driven process (either node/python)

C

Constantin•6/24/25, 8:39 PM

so even if i have

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "MyContainer"
class_name = "MyContainer"

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "MyContainer"
class_name = "MyContainer"

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "MyContainer"
class_name = "MyContainer"

name = "mycontainer"
compatibility_date = "2025-06-24"
[assets]
directory = "./rel"
[[containers]]
class_name = "MyContainer"
image = "./Dockerfile"
max_instances = 10
instance_type = "basic" # Optional, defaults to "dev"
image_vars = { FOO = "BAR" }

[[durable_objects.bindings]]
name = "MyContainer"
class_name = "MyContainer"

the deploy will need the DO

G

gruntlord6•6/24/25, 8:40 PM

haven't checked the docs that closely, if you can use the durable object directly it would make sense why you have to configure it

G

gruntlord6•6/24/25, 8:40 PM

if its something you don't control then it would make sense for them to handle that

G

gruntlord6•6/24/25, 8:43 PM

https://developers.cloudflare.com/durable-objects/reference/durable-objects-migrations/

Cloudflare Docs

Durable Objects migrations

A migration is a mapping process from a class name to a runtime state. This process communicates the changes to the Workers runtime and provides the runtime with instructions on how to deal with those changes.

S

Sameer Ali•6/24/25, 8:49 PM

has someone tried golang/rust app yet? what's the cold start time?

C

Constantin•6/24/25, 8:50 PM

ty @gruntlord6 i'll retry these days. it's quite a new think do me. I'm not sure why wrangler still needs the path for the final build (i assume), via

[assets]
directory = "./_build" <--- the release directory

[assets]
directory = "./_build" <--- the release directory

[assets]
directory = "./_build" <--- the release directory

[assets]
directory = "./_build" <--- the release directory

since a Docker file is present and generates. however, the image size is limited to 25M

CConstantin ty @gruntlord6 i'll retry these days. it's quite a new think do me. I'm not sure...

G

gruntlord6•6/24/25, 8:53 PM

there might be a difference in how your doing it vs what the docs say, it looks like its treating it as an asset
https://developers.cloudflare.com/containers/image-management/

Cloudflare Docs

Image Management

When running wrangler deploy, if you set the image attribute in you Wranlger configuration
file to a path, wrangler will build your container image locally using Docker, then push it to a registry run by Cloudflare.
This registry is integrated with your Cloudflare account and is backed by R2. All authentication is handled automatically by
Cloudf...

D

DanGamble•6/24/25, 8:55 PM

Can I finally run my Ruby on Rails on Cloudflare now?

C

Constantin•6/24/25, 8:55 PM

hmm, getting an error having only

[containers]
image = "./Dockerfile"

[containers]
image = "./Dockerfile"

[containers]
image = "./Dockerfile"

[containers]
image = "./Dockerfile"

1

1984 Ford Laser•6/24/25, 9:12 PM

Have only just started reading the docs, but am I right in thinking these containers are supposed to handle HTTP data/requests only? I can't send an RTMP stream to a container for example?

R

Rapid•6/24/25, 9:15 PM

When deploying a container it deploys it to a single location? Is the pricing relevant to a single location or edge, trying to understand exactly how it is managed, maybe I missed something

RRapid When deploying a container it deploys it to a single location? Is the pricing re...

1

1984 Ford Laser•6/24/25, 9:20 PM

This might help: https://developers.cloudflare.com/containers/architecture/

1

1984 Ford Laser•6/24/25, 9:20 PM

For pricing reasons, you pay for every instance that you have running

11984 Ford Laser Have only just started reading the docs, but am I right in thinking these contai...

1

1984 Ford Laser•6/24/25, 9:23 PM

Last paragraph on Architecture page answered this for me:

Because all Container requests are passed through a Worker, end-users cannot make TCP or UDP requests to a Container instance.

Because all Container requests are passed through a Worker, end-users cannot make TCP or UDP requests to a Container instance.

Because all Container requests are passed through a Worker, end-users cannot make TCP or UDP requests to a Container instance.

Because all Container requests are passed through a Worker, end-users cannot make TCP or UDP requests to a Container instance.

They provide a form to fill to let them know you want this feature as well

T

Thomas Lefebvre•6/24/25, 10:00 PM

To clarify Workers to Container can be TCP as with this example: https://github.com/cloudflare/containers-demos/blob/3ed3765967586cbbca159ef0c6d53ff7e54b0e15/compression/src/index.ts#L26 that uses connect
But client to Workers has to be HTTP for now

GitHub

containers-demos/compression/src/index.ts at 3ed3765967586cbbca159e...

Contribute to cloudflare/containers-demos development by creating an account on GitHub.

CConstantin hmm, getting an error having only ``` [containers] image = "./Dockerfile" ```

B

Beny•6/24/25, 10:16 PM

Should be

[[containers]]

[[containers]]

[[containers]]

[[containers]] I think?

A

andrei•6/24/25, 10:19 PM

is it possible to access D1 bindings inside a container?

A

Aster•6/24/25, 10:30 PM

Hello

I am trying to run an otel collector as a container
It looks like this:

docker pull otel/opentelemetry-collector-contrib:0.128.0
docker tag otel/opentelemetry-collector-contrib:0.128.0 opentelemetry-collector-contrib:0.128.0
npx wrangler containers push opentelemetry-collector-contrib:0.128.0

docker pull otel/opentelemetry-collector-contrib:0.128.0
docker tag otel/opentelemetry-collector-contrib:0.128.0 opentelemetry-collector-contrib:0.128.0
npx wrangler containers push opentelemetry-collector-contrib:0.128.0

docker pull otel/opentelemetry-collector-contrib:0.128.0
docker tag otel/opentelemetry-collector-contrib:0.128.0 opentelemetry-collector-contrib:0.128.0
npx wrangler containers push opentelemetry-collector-contrib:0.128.0

docker pull otel/opentelemetry-collector-contrib:0.128.0
docker tag otel/opentelemetry-collector-contrib:0.128.0 opentelemetry-collector-contrib:0.128.0
npx wrangler containers push opentelemetry-collector-contrib:0.128.0

and then the code is pretty simple:

// ...

export class OpenTelemetryContainer extends Container {
  // Port the container listens on (default: 8080)
  defaultPort = 13133; // http port
  // Time before ~~container~~ sleeps due to inactivity (default: 30s)
  sleepAfter = "2m";
  // Environment variables passed to the container
  envVars = {
    MESSAGE: "I was passed in via the container class!",
  };
  // ...
}

//...

app.get("/health", async (c) => {
  const container = getContainer(c.env.OTEL_CONTAINER);
  return await container.fetch(c.req.raw);
});

// ...

export class OpenTelemetryContainer extends Container {
  // Port the container listens on (default: 8080)
  defaultPort = 13133; // http port
  // Time before ~~container~~ sleeps due to inactivity (default: 30s)
  sleepAfter = "2m";
  // Environment variables passed to the container
  envVars = {
    MESSAGE: "I was passed in via the container class!",
  };
  // ...
}

//...

app.get("/health", async (c) => {
  const container = getContainer(c.env.OTEL_CONTAINER);
  return await container.fetch(c.req.raw);
});

// ...

export class OpenTelemetryContainer extends Container {
  // Port the container listens on (default: 8080)
  defaultPort = 13133; // http port
  // Time before ~~container~~ sleeps due to inactivity (default: 30s)
  sleepAfter = "2m";
  // Environment variables passed to the container
  envVars = {
    MESSAGE: "I was passed in via the container class!",
  };
  // ...
}

//...

app.get("/health", async (c) => {
  const container = getContainer(c.env.OTEL_CONTAINER);
  return await container.fetch(c.req.raw);
});

// ...

export class OpenTelemetryContainer extends Container {
  // Port the container listens on (default: 8080)
  defaultPort = 13133; // http port
  // Time before ~~container~~ sleeps due to inactivity (default: 30s)
  sleepAfter = "2m";
  // Environment variables passed to the container
  envVars = {
    MESSAGE: "I was passed in via the container class!",
  };
  // ...
}

//...

app.get("/health", async (c) => {
  const container = getContainer(c.env.OTEL_CONTAINER);
  return await container.fetch(c.req.raw);
});

But I always get a standard message when hitting /health:

There is no Container instance available at this time.
This is likely because you have reached your max concurrent instance count (set in wrangler config) or are you currently provisioning the Container.
If you are deploying your Container for the first time, check your dashboard to see provisioning status, this may take a few minutes.

There is no Container instance available at this time.
This is likely because you have reached your max concurrent instance count (set in wrangler config) or are you currently provisioning the Container.
If you are deploying your Container for the first time, check your dashboard to see provisioning status, this may take a few minutes.

There is no Container instance available at this time.
This is likely because you have reached your max concurrent instance count (set in wrangler config) or are you currently provisioning the Container.
If you are deploying your Container for the first time, check your dashboard to see provisioning status, this may take a few minutes.

There is no Container instance available at this time.
This is likely because you have reached your max concurrent instance count (set in wrangler config) or are you currently provisioning the Container.
If you are deploying your Container for the first time, check your dashboard to see provisioning status, this may take a few minutes.

Do you know what I could be missing? Despite defaultPort set to 13133, I don't see it being reflected when running

docker ps

docker ps

docker ps

docker ps

78b186a1676b   cloudflare-dev/opentelemetrycontainer:dd5373cf   "/otelcol-contrib --…"   8 seconds ago   Up 7 seconds   0.0.0.0:32813->4317/tcp, [::]:32813->4317/tcp, 0.0.0.0:32814->4318/tcp, [::]:32814->4318/tcp, 0.0.0.0:32815->55678/tcp, [::]:32815->55678/tcp, 0.0.0.0:32816->55679/tcp, [::]:32816->55679/tcp   workerd-opentelemetry-exploration-OpenTelemetryContainer-c8a1bd81d48b5c133d43ff665add35273c6305f82f94a7fe0adae6baa179c408

78b186a1676b   cloudflare-dev/opentelemetrycontainer:dd5373cf   "/otelcol-contrib --…"   8 seconds ago   Up 7 seconds   0.0.0.0:32813->4317/tcp, [::]:32813->4317/tcp, 0.0.0.0:32814->4318/tcp, [::]:32814->4318/tcp, 0.0.0.0:32815->55678/tcp, [::]:32815->55678/tcp, 0.0.0.0:32816->55679/tcp, [::]:32816->55679/tcp   workerd-opentelemetry-exploration-OpenTelemetryContainer-c8a1bd81d48b5c133d43ff665add35273c6305f82f94a7fe0adae6baa179c408

78b186a1676b   cloudflare-dev/opentelemetrycontainer:dd5373cf   "/otelcol-contrib --…"   8 seconds ago   Up 7 seconds   0.0.0.0:32813->4317/tcp, [::]:32813->4317/tcp, 0.0.0.0:32814->4318/tcp, [::]:32814->4318/tcp, 0.0.0.0:32815->55678/tcp, [::]:32815->55678/tcp, 0.0.0.0:32816->55679/tcp, [::]:32816->55679/tcp   workerd-opentelemetry-exploration-OpenTelemetryContainer-c8a1bd81d48b5c133d43ff665add35273c6305f82f94a7fe0adae6baa179c408

78b186a1676b   cloudflare-dev/opentelemetrycontainer:dd5373cf   "/otelcol-contrib --…"   8 seconds ago   Up 7 seconds   0.0.0.0:32813->4317/tcp, [::]:32813->4317/tcp, 0.0.0.0:32814->4318/tcp, [::]:32814->4318/tcp, 0.0.0.0:32815->55678/tcp, [::]:32815->55678/tcp, 0.0.0.0:32816->55679/tcp, [::]:32816->55679/tcp   workerd-opentelemetry-exploration-OpenTelemetryContainer-c8a1bd81d48b5c133d43ff665add35273c6305f82f94a7fe0adae6baa179c408

Aandrei is it possible to access D1 bindings inside a container?

B

Beny•6/24/25, 10:31 PM

You can use the Container's DurableObject for it since it has bindings

A

andrei•6/24/25, 10:46 PM

are there examples of how to access the durable object inside the container?

C

Cyb3r-Jak3•6/24/25, 10:53 PM

I don't believe you can access bindings from inside a container

I

Ivan Chernetsky (CF)•6/24/25, 10:58 PM

@andrei @Mike Nomitch talked about container-to-worker calls in the livestream: https://youtu.be/UiZ8JqSnoa0?t=2252

YouTubeCloudflare Developers

Cloudflare Containers Public Beta — Launch Livestream

Containers are now available on Workers, unlocking a world of possibilities for your applications. Tune in to hear from Cloudflare Product Manager Mike Nomitch, who will be joined by special guests from Coder.com and Val.Town to showcase a glimpse of what Containers make possible.

A

andrei•6/24/25, 11:04 PM

hopefully there's plans to make this more streamlined. otherwise sounds like we need to create a bunch of glue code to talk to the worker from the container to access all these services (e.g. D1, Vectorize, etc)

M

midnight•6/24/25, 11:27 PM

I’m working on deploying an image but keep running into issues building. From what I can tell I’m building on an arm host/VM but deployment requires a amd64 build.

I added

—platform=linux/amd64

—platform=linux/amd64

—platform=linux/amd64

—platform=linux/amd64 to my FROM in docker but I’m still having issues.

1

1984 Ford Laser•6/25/25, 1:02 AM

I'm in Melbourne(Geelong) too and get Singapore as well. I assume this will get better over time, perhaps SIN is the only APAC colo with Containers enabled currently or something

1

1984 Ford Laser•6/25/25, 1:03 AM

I'm getting request times from a hot container and DO to be 4-500ms, which I think is fine. The request goes through the Worker, then DO, then Container, then back to Worker so there's definitely more going on in the pipeline that might slow it down

R

Ru Chern Chong•6/25/25, 1:05 AM

Was there an announcement?

1

1984 Ford Laser•6/25/25, 1:05 AM

Also looks like the accompanying DO will stay active for the entire time the Container is active (which is something I should have expected). Just something for people to note when scaling up this feature, that the DO also needs to be paid for. A single DO instance uses 7.68GB-sec per minute for anyone curious

RRu Chern Chong Was there an announcement?

1

1984 Ford Laser•6/25/25, 1:06 AM

nah

1

1984 Ford Laser•6/25/25, 1:13 AM

Container specific dashboard is very cool! Just need the same for individual DO instances now..

11984 Ford Laser nah

R

Ru Chern Chong•6/25/25, 1:15 AM

Oh okay

1

1984 Ford Laser•6/25/25, 1:16 AM

Is it stated anywhere that deploying the Worker that the container is bound to will immediately take any active instances offline? I know the same has always been true of DOs, but might be good to note for those that are unfamiliar with the deployment pattern

DDefenestrator It looks like the location hint doesn't hold much weight lol. I'm in Melbourne...

1

1984 Ford Laser•6/25/25, 1:18 AM

I've also gotten an instance to spawn in Japan, note I haven't been providing any locationHint, I've been curious to see where instances will spawn on their own

1

1984 Ford Laser•6/25/25, 1:19 AM

ok wtf Florida

Similar Threads

Similar Threads

Similar Threads