If we connect a container to R2, does that trigger paid egress on the container, or like workers is

AAri If we connect a container to R2, does that trigger paid egress on the container,...

1

1984 Ford Laser•10/7/25, 11:34 PM

I feel like there needs to be a sticky for this question or something. Currently, a Container sending to R2 does count as egress

A

AriOP•10/7/25, 11:54 PM

Ahh okok, ty! Yeah

C

CodingCoop | Co-Founder Nullshot•10/8/25, 8:44 AM

Hey all - Anyone submit the use case to being able to override the instance type of a container as a start param / argument?

Right now I would need to setup a N classes that extend my CustomContainerClass which is configured as separate durable objects / containers in wrangler.

Ideally I am able to have a default instance_type in wrangler, and have the ability to increase resources dynamically.

If this is not on the roadmap / happening anytime soon, happy to implement the above solution - Just need guidance if I missed something and wondering if anyone else ran into this case.

J

jshm•10/8/25, 8:13 PM

Hi folks. Am I correct in saying Containers aren't supported in WfP at the moment? Is support on the roadmap at all?

P

pandalambda2•10/9/25, 10:41 AM

I don't think they've figured out the billing for standalone yet, be surprised if I can start spinning up containers on WfP's account like shopify and smashing egress

P

pandalambda2•10/9/25, 10:42 AM

getting clarification around even egress is like extracting blood from a stone, you're far back in the queue 555

M

Maulik Tyagi•10/9/25, 12:14 PM

Does anyone know what time length container keeps itself alive, basically my worker which is cron worker, calls containers, I'm uing cloudflare/containers npm package, and I can see logs that my container has started, in container I have my python code, which has print statements, in container logs I can see some of these print statements being logged, but then nothing happens, container code doesnt get full executed, my container then shuts down by itself, with possibly incomplete code completion, I have used sleepAfter 30m, i use wrangler tail to track logs and i get some Alarm logs every 3 minutes, then after 10 minutes, container shuts down, what can i do that code gets completed, it can take 15-20 mins to complete, my container code is a python function which do some tasks and then exit, can you guide me through this

MMaulik Tyagi Does anyone know what time length container keeps itself alive, basically my wor...

A

arnar•10/9/25, 12:24 PM

containers-beta sounds like the same issues as here?

M

Maulik Tyagi•10/9/25, 12:39 PM

Hmm, let me try and see this one, thanks for highlighting it

K

kian•10/9/25, 6:56 PM

Does it work in

npx wrangler dev

npx wrangler dev

npx wrangler dev

npx wrangler dev? Surfacing our errors (since only your own logs go into the logs tab right now) in that tab is on our roadmap.

K

kian•10/9/25, 6:57 PM

If it is due to your Container exiting then there'll also be an exit code in that Recent Errors view in the next dash release.

N

nicoritschel•10/9/25, 8:32 PM

Should stdout/stderr from the container be captured in logs? I don't see either using

bunx wrangler dev

bunx wrangler dev

bunx wrangler dev

bunx wrangler dev or on my live deployed container with logging enabled. That pane is completely empty.

K

kian•10/9/25, 8:55 PM

During local dev? Not yet
Deployed with Workers Observability enabled? Yes

A

Arden•10/10/25, 8:13 AM

hey guys any plans to setup a starter template for java , c# and any other languages? I have a hard time to connect to my container when its deployed :PES_SadShrug:

K

kian•10/10/25, 10:54 AM

There shouldn’t really be more to it than “start a web server listening on all interfaces on your EXPOSE’d port”

Z

zalsyen•10/10/25, 8:11 PM

bro where is the containers dashboard lol, just disappeared

Z

zalsyen•10/10/25, 8:11 PM

did u nuke containers?

Zzalsyen did u nuke containers? 🙁

I

Isaac McFadyen•10/10/25, 8:20 PM

Nope, there's a dashboard issue. Team is aware and looking at it.

Z

zalsyen•10/10/25, 8:20 PM

awesome ty

JJon M.Did anyone else fully lose access to containers from their CF Dashboard?

I

Isaac McFadyen•10/10/25, 8:38 PM

Ongoing dashboard issue, yup.

I

Isaac McFadyen•10/10/25, 8:51 PM

https://www.cloudflarestatus.com/incidents/7jv9tr6793pd

Some Dashboard Features are Unavailable

1

1984 Ford Laser•10/11/25, 6:48 AM

Has anyone done any ffmpeg video rendering in a Container and can provide some performance numbers?

M

Maulik Tyagi•10/13/25, 6:58 AM

Hey, I keep getting error in 'wrangler tail' when i manually call my worker which is supposed to call the container, i found that it corresponds to that worker cannot find the container? but in dashboard I can see both worker and container attached to same durable object, for reference in worker.js I'm using container class and getContainer object to get my object, and my wrangler.toml follow same pattern as official documentation

MMaulik Tyagi Hey, I keep getting error in 'wrangler tail' when i manually call my worker whic...

K

kian•10/13/25, 1:35 PM

That sounds like your Container isn't responding on the port within 5 seconds

K

kian•10/13/25, 1:36 PM

https://github.com/cloudflare/containers/blob/e4879c8d973a97a67cdb8162a860fc00a9e51257/src/lib/container.ts#L938-L955

Kkian That sounds like your Container isn't responding on the port within 5 seconds

M

Maulik Tyagi•10/13/25, 1:46 PM

Is that so? although container, when I go to dashboard goes from ready to active state, and in logs I can see logs from my container code, although it doesn't get completed, but I still see some logs. So I believe container is running when I manually call my worker.

K

kian•10/13/25, 1:50 PM

The "The operation was aborted" error will be coming from the 5 second timeout we have on the "is this port ready for us to send requests?" health check.

K

kian•10/13/25, 1:51 PM

Your Container will start but we won't send the

fetch

fetch

fetch

fetch request to it until that check passes

M

Maulik Tyagi•10/13/25, 2:10 PM

Sorry to ask, but then how do i make it pass health check?

K

kian•10/13/25, 2:29 PM

Is your Container responding to requests for any hostname / interface / path (even if a 404)?

K

kian•10/13/25, 2:32 PM

We're basically not getting a response to our health check request.

L

lusons•10/14/25, 2:12 AM

Can I use the restful api or cloudflare-typescript sdk to upload a container? I haven't found any documentation on this yet.

Kkian Is your Container responding to requests for any hostname / interface / path (ev...

M

Maulik Tyagi•10/14/25, 6:18 AM

No, it is just a function which is supposed to run on its own when container starts, it is a cron container, so it does not respond back anything,

import { Container, getContainer } from "@cloudflare/containers";

export class dummy extends Container {
  constructor(state, env) {
    super(state, env);
    // Set envVars in constructor
    this.envVars = {
      NAME: env.name,
    };
  }
  sleepAfter = "30m";
  async onStart() {
    console.log("Container started");
  }
  enableInternet = true;
  onStop() {
    console.log("Container shut down");
  }
  onError(error) {
    console.log("Error:", error);
  }

}
export default {
  async fetch(request, env, ctx) {
    // Trigger Durable Object
    console.log("Manual trigger → starting container");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          // You can pass dynamic values here
          MANUAL_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
    return new Response("Container manually triggered and executed", {
      status: 200,
    });
  },
  async scheduled(controller, env, ctx) {
    console.log("Cron triggered");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          CRON_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
  },
};

import { Container, getContainer } from "@cloudflare/containers";

export class dummy extends Container {
  constructor(state, env) {
    super(state, env);
    // Set envVars in constructor
    this.envVars = {
      NAME: env.name,
    };
  }
  sleepAfter = "30m";
  async onStart() {
    console.log("Container started");
  }
  enableInternet = true;
  onStop() {
    console.log("Container shut down");
  }
  onError(error) {
    console.log("Error:", error);
  }

}
export default {
  async fetch(request, env, ctx) {
    // Trigger Durable Object
    console.log("Manual trigger → starting container");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          // You can pass dynamic values here
          MANUAL_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
    return new Response("Container manually triggered and executed", {
      status: 200,
    });
  },
  async scheduled(controller, env, ctx) {
    console.log("Cron triggered");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          CRON_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
  },
};

import { Container, getContainer } from "@cloudflare/containers";

export class dummy extends Container {
  constructor(state, env) {
    super(state, env);
    // Set envVars in constructor
    this.envVars = {
      NAME: env.name,
    };
  }
  sleepAfter = "30m";
  async onStart() {
    console.log("Container started");
  }
  enableInternet = true;
  onStop() {
    console.log("Container shut down");
  }
  onError(error) {
    console.log("Error:", error);
  }

}
export default {
  async fetch(request, env, ctx) {
    // Trigger Durable Object
    console.log("Manual trigger → starting container");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          // You can pass dynamic values here
          MANUAL_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
    return new Response("Container manually triggered and executed", {
      status: 200,
    });
  },
  async scheduled(controller, env, ctx) {
    console.log("Cron triggered");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          CRON_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
  },
};

import { Container, getContainer } from "@cloudflare/containers";

export class dummy extends Container {
  constructor(state, env) {
    super(state, env);
    // Set envVars in constructor
    this.envVars = {
      NAME: env.name,
    };
  }
  sleepAfter = "30m";
  async onStart() {
    console.log("Container started");
  }
  enableInternet = true;
  onStop() {
    console.log("Container shut down");
  }
  onError(error) {
    console.log("Error:", error);
  }

}
export default {
  async fetch(request, env, ctx) {
    // Trigger Durable Object
    console.log("Manual trigger → starting container");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          // You can pass dynamic values here
          MANUAL_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
    return new Response("Container manually triggered and executed", {
      status: 200,
    });
  },
  async scheduled(controller, env, ctx) {
    console.log("Cron triggered");
    const container = getContainer(env.DUMMY);
    await container.start({
      startOptions: {
        envVars: {
          CRON_TRIGGERED_AT: new Date().toISOString(),
        },
      },
    });
  },
};

For reference this is my worker code, and my container is a python function.

MMaulik Tyagi No, it is just a function which is supposed to run on its own when container sta...

K

kian•10/14/25, 3:00 PM

I don't think the Containers library will let you start a Container without it having a web server / port for us to check on.

K

kian•10/14/25, 3:01 PM

The direct API available with Durable Objects (which we're wrapping in the library) might be a better fit: https://developers.cloudflare.com/durable-objects/api/container/

C

christianklotz•10/14/25, 3:04 PM

I've been using the Sandbox SDK, and while it works locally, once deployed to Cloudflare, as the DO is trying to write a file with

"[HTTP Client] Making POST request to http://localhost:3000/api/write"

"[HTTP Client] Making POST request to http://localhost:3000/api/write"

"[HTTP Client] Making POST request to http://localhost:3000/api/write"

"[HTTP Client] Making POST request to http://localhost:3000/api/write", I get the following error message:

"Sandbox error:",
"Error: Container connectivity was lost"

"Sandbox error:",
"Error: Container connectivity was lost"

"Sandbox error:",
"Error: Container connectivity was lost"

"Sandbox error:",
"Error: Container connectivity was lost"

DO ID:

dcb4030e8dbbd797c6f34192fba09c75e77c42d32244b52680d3431302e0a7ee

dcb4030e8dbbd797c6f34192fba09c75e77c42d32244b52680d3431302e0a7ee

dcb4030e8dbbd797c6f34192fba09c75e77c42d32244b52680d3431302e0a7ee

dcb4030e8dbbd797c6f34192fba09c75e77c42d32244b52680d3431302e0a7ee

After some debugging it looks like, the sandbox preflight succeeds, the first call to

writeFile

writeFile

writeFile

writeFile causes the container to lose connectivity, and then it kinda 'hangs'.

Anyone seen similar things before?

K

kian•10/14/25, 3:42 PM

What's the application ID for the Containers app? It's the UUID in the URL when you go to the Containers overview (Workers & Pages in sidebar -> Containers -> click on the app)

Kkian What's the application ID for the Containers app? It's the UUID in the URL when ...

C

christianklotz•10/14/25, 3:48 PM

a03e7cbd-ed24-485c-92c4-eeb9bac1ae5d

a03e7cbd-ed24-485c-92c4-eeb9bac1ae5d

a03e7cbd-ed24-485c-92c4-eeb9bac1ae5d

a03e7cbd-ed24-485c-92c4-eeb9bac1ae5d

K

kian•10/14/25, 3:51 PM

I think that’s your account ID - the Containers application is a UUID that comes later

K

kian•10/14/25, 3:52 PM

/accountId/containers/applicationId/ - something like that

Kkian I think that’s your account ID - the Containers application is a UUID that comes...

C

christianklotz•10/14/25, 3:53 PM

corrected above

K

kian•10/14/25, 7:00 PM

I'm not seeing much in the way of errors on our side - I can see the "Container" (in this case the Sandbox SDK entrypoint) exited successfully (exit code 0) at 15:15 UTC.

K

kian•10/14/25, 7:00 PM

Is it always reproducible for you when using that

writeFile

writeFile

writeFile

writeFile API?

Kkian I'm not seeing much in the way of errors on our side - I can see the "Container"...

C

christianklotz•10/14/25, 8:28 PM

Thank you for looking into it! Yeah, it's always

writeFile

writeFile

writeFile

writeFile. Interestingly I get this on

execute

execute

execute

execute:

"[HTTP Client] Error executing in session:",
"Error: Session 'sandbox-default' not found"

"[HTTP Client] Error executing in session:",
"Error: Session 'sandbox-default' not found"

"[HTTP Client] Error executing in session:",
"Error: Session 'sandbox-default' not found"

"[HTTP Client] Error executing in session:",
"Error: Session 'sandbox-default' not found"

Not sure if related.

I am whipping up a more isolated implementation right now to see it still breaks

If we connect a container to R2, does that trigger paid egress on the container, or like workers is

Similar Threads

Similar Threads

Similar Threads