My initial log message was from UI, though locally i get the same error. 1) wrangler login again had

My initial log message was from UI, though locally i get the same error.
1) wrangler login again hadn't helped 2) API Token in profile has Edit rights for containers (see image)
3)

rm -rf ~/.config/.wrangler

rm -rf ~/.config/.wrangler

rm -rf ~/.config/.wrangler

rm -rf ~/.config/.wrangler and wrangler login hadn't helped either

Locally i get following stacktrace in logs

--- 2025-07-02T15:27:25.870Z error
✘ [ERROR] Unauthorized


---

--- 2025-07-02T15:27:25.872Z debug
Error: Unauthorized
    at buildAndMaybePush (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:60995:13)
    at process.processTicksAndRejections (node:internal/process/task_queues:105:5)
    at async maybeBuildContainer (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62544:15)
    at async deployContainers (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62593:19)
    at async deploy (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:152723:5)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:81967:66)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:78772:7)
---

--- 2025-07-02T15:27:25.870Z error
✘ [ERROR] Unauthorized


---

--- 2025-07-02T15:27:25.872Z debug
Error: Unauthorized
    at buildAndMaybePush (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:60995:13)
    at process.processTicksAndRejections (node:internal/process/task_queues:105:5)
    at async maybeBuildContainer (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62544:15)
    at async deployContainers (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62593:19)
    at async deploy (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:152723:5)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:81967:66)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:78772:7)
---

--- 2025-07-02T15:27:25.870Z error
✘ [ERROR] Unauthorized


---

--- 2025-07-02T15:27:25.872Z debug
Error: Unauthorized
    at buildAndMaybePush (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:60995:13)
    at process.processTicksAndRejections (node:internal/process/task_queues:105:5)
    at async maybeBuildContainer (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62544:15)
    at async deployContainers (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62593:19)
    at async deploy (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:152723:5)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:81967:66)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:78772:7)
---

--- 2025-07-02T15:27:25.870Z error
✘ [ERROR] Unauthorized


---

--- 2025-07-02T15:27:25.872Z debug
Error: Unauthorized
    at buildAndMaybePush (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:60995:13)
    at process.processTicksAndRejections (node:internal/process/task_queues:105:5)
    at async maybeBuildContainer (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62544:15)
    at async deployContainers (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:62593:19)
    at async deploy (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:152723:5)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:81967:66)
    at async Object.handler (/home/vagrant/Code/hello-containers/node_modules/wrangler/wrangler-dist/cli.js:78772:7)
---

FFedor Vompe My initial log message was from UI, though locally i get the same error. 1) wran...

7eleven•7/2/25, 3:43 PM

I have a similar issue. Images cannot be deleted from the registry if deployed using an Account API token.
https://discordapp.com/channels/595317990191398933/1389543206420742144

77eleven I have a similar issue. Images cannot be deleted from the registry if deployed u...

7eleven•7/2/25, 3:48 PM

I tried logging in with

wrangler login

wrangler login

, but even with OAuth Token privileges I could not delete it.

Yash•7/2/25, 7:16 PM

Is there a way to keep at least one minimum instance running to avoid cold start delays?

goofysystem•7/2/25, 7:39 PM

I'm getting this after manually deleting container and associated DO from workers dashboard

  Cannot create binding for class 'RustTaskContainer' because it is not currently configured to
  implement Durable Objects. Ensure a --new-class migration has been applied to it. [code: 10061]
  
  If you think this is a bug, please open an issue at:
  https://github.com/cloudflare/workers-sdk/issues/new/choose

  Cannot create binding for class 'RustTaskContainer' because it is not currently configured to
  implement Durable Objects. Ensure a --new-class migration has been applied to it. [code: 10061]
  
  If you think this is a bug, please open an issue at:
  https://github.com/cloudflare/workers-sdk/issues/new/choose

I was trying to solve this issue: hey, nice work everyone - very excited

Mike Nomitch•7/2/25, 10:24 PM

Hello again

Here with a general update on some common snags people are seeing/have seen:

sleepAfter not keeping containers alive properly - This is fixed as of (@)cloudflare/containers 0.0.18. I'd upgrade to that and you should be good to go.

Cold Starts being slow - We've got a PR up to bring cold starts lower (2-3 seconds for small web servers). This was a result of CPU restrictions on the smaller instances (IE
```
dev
```
```
dev
```
the default). Our solution is to allow you to use a lot of CPU in the initial boot of your container regardless of instance size, and cold starts will be faster. (note that long-term we're exploring ways to get sub-second). -> PR up now for review. Rollout likely starting at end of this week, and fully live early next week.
OOMs on boot - Some Rails and Python apps were OOMing on boot on
```
dev
```
```
dev
```
instances. We have a PR up to grant more memory to all instance types so your app (excluding the VM) gets the full memory allotment. Like the one above, rollout likely starting at end of this week, but takes a couple business days to finish, so fully live early next week. (I'd use a
```
basic
```
```
basic
```
instant_type as a temporary fix)
there is no container instance that can be provided to this durable objectthere is no container instance that can be provided to this durable object
- This might happen for several reasons. (2 and 3 above for instance). - A likely reason is that
```
max_instances
```
```
max_instances
```
  was not set in wrangler config, and we accidentally had it default to 0 . So if you're seeing this, I'd make sure
```
max_instances
```
```
max_instances
```
  is set. You also might see this if you immediately request a container after first deploy. If so, you should just wait. - In any case, we need to surface better logs here and in the UI. And a new
```
wrangler
```
```
wrangler
```
  will change the default here soon.

Thanks again for all the tire kicking and feedback. Apologies for the snags.

Shawn•7/2/25, 11:49 PM

Hey guys, this may be a noob question, but is CF containers strictly serverless? Can we host long running services like FastAPI / ARQ similar to how it works in k8s? Cost is a not of concern here. Just had enough dealing with AKS

Yash•7/3/25, 2:57 AM

Hello ,
I’ve spent the past 4–5 days exploring Cloudflare Containers. While the idea is impressive—especially for someone like me who runs multiple services on AWS and Azure containers—the platform is still in beta and lacks maturity.

There are too many limitations right now for production use. After trying to make it work, I realized I was spending more time troubleshooting than building. So I’ve decided to put it on hold for now and focus on actual development work.

Hopefully, once Cloudflare improves the service, I’ll revisit it. But for now, I’d suggest avoiding Cloudflare Containers for production until it's more stable.
Really appreciate your support and quick reply, @Mike Nomitch

YYash Hello , I’ve spent the past 4–5 days exploring Cloudflare Containers. While the...

Isaac McFadyen•7/3/25, 3:02 AM

There are too many limitations right now for production use

Just an FYI that it's never recommended to use a service that is in Beta in production, and that includes Containers which just launched and is still very much in beta/is being worked on constantly by the team

slice•7/3/25, 5:04 AM

I would like to easily host a docker file without any config stuff. If that’s not possible then at least not with the extra ts/js code to communicate with the container and act a middleware.

Beny•7/3/25, 1:02 PM

I noticed our account has a 40 vCPU quota instead of the 20 mentioned in docs. Was it increased?

{"error":"SURPASSED_TOTAL_LIMITS","details":{"reason":"This application requires more than the allocated quota of vcpus in this account to run all its requested instances of count 100.","used_vcpu":0,"remaining_vcpu":40,"requested_vcpu":50}}

{"error":"SURPASSED_TOTAL_LIMITS","details":{"reason":"This application requires more than the allocated quota of vcpus in this account to run all its requested instances of count 100.","used_vcpu":0,"remaining_vcpu":40,"requested_vcpu":50}}

Mitya•7/3/25, 2:24 PM

Hi all - can I just clarify: do containers, in the current form, support Docker Compose, or only Dockerfiles?

dustiiin•7/3/25, 3:57 PM

Hey anyone knows whats wrong with my dockerfile?

The apt-get update & install row is taking long
After exactly 120 Seconds the Build throws Temporary failure resolving 'deb.debian.org'Temporary failure resolving 'deb.debian.org'

FROM php:8.4-fpm
WORKDIR /var/www/html

RUN apt-get update && apt-get install -y nginx

COPY app/ /var/www/html/
COPY config/nginx.conf /etc/nginx/nginx.conf
COPY config/supervisord.conf /etc/supervisord.conf

EXPOSE 80

CMD ["nginx", "-g", "daemon off;"]

FROM php:8.4-fpm
WORKDIR /var/www/html

RUN apt-get update && apt-get install -y nginx

COPY app/ /var/www/html/
COPY config/nginx.conf /etc/nginx/nginx.conf
COPY config/supervisord.conf /etc/supervisord.conf

EXPOSE 80

CMD ["nginx", "-g", "daemon off;"]

max•7/3/25, 5:15 PM

@Mike Nomitch any plans to allow filesystem extensions through fuse or otherwise?

Ssaswatdas Is there a way to query a container's logs and metrics to build a custom dashboa...

saswatdas•7/3/25, 7:17 PM

What’s the best way to collect and expose container logs, and resource usage (CPU, memory), so they can be streamed back to a user or agent for debugging, auditing, or billing purposes — especially when running AI-generated or untrusted code?

@Mike Nomitch

11984 Ford Laser One of the devs mentioned somewhere that the ability to mount an R2 bucket in th...

yb•7/4/25, 3:29 AM

Maybe use s3fs in container?

jaden•7/4/25, 8:53 AM

Excuse me, how can I run very different services under one worker? I see that the examples provided show one worker corresponding to one container.

Jjaden Excuse me, how can I run very different services under one worker? I see that th...

1984 Ford Laser•7/4/25, 10:43 AM

Put more services on the one container?

terrxo•7/4/25, 11:03 AM

Are there any examples of containers in a monorepo? Does it even work as of now?

3m0•7/4/25, 11:48 AM

anyone having trouble running the containers only in production? For some reason, when I run my container locally in dev mode, everything works fine, but when the container is deployed, it just doesn't run, and no error logs are generated

11984 Ford Laser Put more services on the one container?

jaden•7/4/25, 12:25 PM

Put more services(container) on the one worker . just this

jaden•7/4/25, 12:45 PM

The code for the worker is together with the code for the container, so I understand that does one worker need to correspond to one container?

jaden•7/4/25, 12:45 PM

@1984 Ford Laser

Jjaden The code for the worker is together with the code for the container, so I unders...

Isaac McFadyen•7/4/25, 1:13 PM

No, if you use one Worker and then add multiple containerscontainers (in your wrangler.jsonc right there) that should work.

Isaac McFadyen•7/4/25, 1:13 PM

The only reason theres 1:1 in the examples is for simplicity/to avoid overwhelming new users.

Jjaden The code for the worker is together with the code for the container, so I unders...

terrxo•7/4/25, 3:03 PM

Did you get it working?

Gabriel Francisco•7/4/25, 4:47 PM

My container is being killed randomly... @Mike Nomitch how can I see the reason?

rufi•7/5/25, 1:05 AM

I can deploy my containers and it works remotely, but I cannot run them locally in dev mode, I get: "A call to blockConcurrencyWhile() in a Durable Object waited for too long. The call was canceled and the Durable Object was reset." (...) "container error: Failed to verify port 8000 is available after 27 attempts, last error: internal error; reference = vs98k.."

Rrufi I can deploy my containers and it works remotely, but I cannot run them locally ...

lupin•7/5/25, 12:46 PM

Same here :NotLikeThis:

Radu•7/5/25, 6:06 PM

  const response = await container.containerFetch("http://container/to-markdown", {
    method: "POST",
    headers: {
      "Content-Type": "multipart/form-data",
      "Content-Disposition": `form-data; name="file"; filename="${file.name}"`,
    },
    body: formData,
  });

  const response = await container.containerFetch("http://container/to-markdown", {
    method: "POST",
    headers: {
      "Content-Type": "multipart/form-data",
      "Content-Disposition": `form-data; name="file"; filename="${file.name}"`,
    },
    body: formData,
  });

✘ [ERROR] DOMException {

code: 25,
name: 'DataCloneError',
message: 'Could not serialize object of type "FormData". This type does not support
serialization.',
stack: 'DataCloneError: Could not serialize object of type "FormData". This type does not
support serialization.\n' +

is this not implemented?

Mr. Technician•7/5/25, 6:13 PM

12:55:42.975    #4 ERROR: failed to register layer: write /usr/src/app/node_modules/playwright-1.51/lib/vite/recorder/assets/codeMirrorModule-B9YMkrwa.js: no space left on device
12:55:53.846    ------
12:55:53.851     > [1/1] FROM ghcr.io/browserless/chromium:latest@sha256:ca1cb28a74a7658175057711ca2b2f68b4eb72a2897b2515d2dac124a4a5d70b:
12:55:53.852    ------
12:55:53.853    Dockerfile:1
12:55:53.853    --------------------
12:55:53.853       1 | >>> FROM ghcr.io/browserless/chromium:latest
12:55:53.853       2 |     
12:55:53.853       3 |     ENV \
12:55:53.853    --------------------
12:55:53.853    ERROR: failed to build: failed to solve: failed to register layer: write /usr/src/app/node_modules/playwright-1.51/lib/vite/recorder/assets/codeMirrorModule-B9YMkrwa.js: no space left on device
12:55:53.869    
12:55:53.983    ✘ [ERROR] Build exited with code: 1

12:55:42.975    #4 ERROR: failed to register layer: write /usr/src/app/node_modules/playwright-1.51/lib/vite/recorder/assets/codeMirrorModule-B9YMkrwa.js: no space left on device
12:55:53.846    ------
12:55:53.851     > [1/1] FROM ghcr.io/browserless/chromium:latest@sha256:ca1cb28a74a7658175057711ca2b2f68b4eb72a2897b2515d2dac124a4a5d70b:
12:55:53.852    ------
12:55:53.853    Dockerfile:1
12:55:53.853    --------------------
12:55:53.853       1 | >>> FROM ghcr.io/browserless/chromium:latest
12:55:53.853       2 |     
12:55:53.853       3 |     ENV \
12:55:53.853    --------------------
12:55:53.853    ERROR: failed to build: failed to solve: failed to register layer: write /usr/src/app/node_modules/playwright-1.51/lib/vite/recorder/assets/codeMirrorModule-B9YMkrwa.js: no space left on device
12:55:53.869    
12:55:53.983    ✘ [ERROR] Build exited with code: 1

Mr. Technician•7/5/25, 6:13 PM

can I get a file space increase pretty please?

Mr. Technician•7/5/25, 7:35 PM

    ✘ [ERROR] Image too large: needs 2467 MB, but your app is limited to images with size 2000 MB. Your account needs more disk size per instance to run this container. The default disk size is 2GB.

    ✘ [ERROR] Image too large: needs 2467 MB, but your app is limited to images with size 2000 MB. Your account needs more disk size per instance to run this container. The default disk size is 2GB.

Mr. Technician•7/5/25, 7:35 PM

MMr. Technician ``` ✘ [ERROR] Image too large: needs 2467 MB, but your app is limited to imag...

Alex•7/5/25, 8:23 PM

Update ur docker ignore file. 2GB mostly enough even for a large projects.

AAlex Update ur docker ignore file. 2GB mostly enough even for a large projects.

Mr. Technician•7/5/25, 8:24 PM

can i do this when all im doing is pulling an existing image?

Mr. Technician•7/5/25, 8:24 PM

i’m not surprised it’s big since both images i tested include Chromium

Alex•7/5/25, 8:26 PM

You can get original dockerfile and drop the stuff you don't need. I would start from that point

AAlex You can get original dockerfile and drop the stuff you don't need. I would start...

Mr. Technician•7/5/25, 10:39 PM

ah ok, i doubt there much i can drop unfortunately

619napster•7/5/25, 11:10 PM

The code below was working fine a day ago
but now I am getting this error
anything I have missed ?
I was not calling start before as well

619napster•7/5/25, 11:59 PM

Never mind i just missed a pacakge in requirements.txt

Muhammad•7/6/25, 6:37 AM

I'm getting: Container suddenly disconnected, try again

it's working perfectly on localhost but once I deploy I get the error

IIsaac McFadyen No, if you use one Worker and then add multiple `containers` (in your wrangler.j...

jaden•7/7/25, 2:06 AM

In our current architecture, I want it to be similar to "worker for platform." A worker can mount multiple containers (services) and use subdomains to distribute traffic to different containers. Currently, when I look at the deployment process, containers are represented as an array. In this case, it seems that deploying multiple containers simultaneously is not feasible, as this is a distributed scenario where several deployments may occur at the same time. How can this be handled?

Tterrxo Did you get it working?

jaden•7/7/25, 2:08 AM

yes.

jaden•7/7/25, 2:23 AM

The wrangler tail prints the logs of the current worker. Is there any way to query the real-time logs of the container through a command? This is because the stack output of the container is also of concern.

My initial log message was from UI, though locally i get the same error. 1) wrangler login again had

Similar Threads