I always have to deploy twice lol

MMaurice When working inside containers there is no access to worker bindings, so when us...

G

gruntlord6OP•8/22/25, 8:28 PM

Can't you just have the durable object make those requests?

Ggruntlord6 I always have to deploy twice lol

G

goofysystem•8/22/25, 8:31 PM

So is everyone deploying locally bc ci builds the image afresh everytime so it will always fail

. but also it used to work before so

G

gruntlord6OP•8/22/25, 8:32 PM

no it seems to work fine either though CI/CID when I push with github. If using wrangler it works the second time I push

G

gruntlord6OP•8/22/25, 8:33 PM

I have noticed a higher error rate from wrangler the last few days, both on my machine and the CI/CID pipeline. Most of them internal errors or "malformed responses"

Ggruntlord6 no it seems to work fine either though CI/CID when I push with github. If using...

G

goofysystem•8/22/25, 8:36 PM

I use wrangler in ci too thats probably why. I really want to stay using wrangler to deploy from ci so hopefully fix comes soon

Ggoofysystem Running into this on deploy: ``` --- 2025-08-22T00:33:52.056Z debug Checking fo...

G

goofysystem•8/23/25, 3:07 AM

Found out that wrangler deploy does not replace the image with the new image ever. The image shown on cloudflare containers dashboard was much older version regardless of how many new image I deploy. Obviously, the first failure uploads to the repository but fails to update the image to use the newst tag. And the second run just skips pushing the new image and treats it as already existing. Only way to get new image up is deleting the container from the dashboard and redeploying, which is not ideal.

B

bentlegen•8/23/25, 3:51 AM

Came in here to basically repeat the same as above.

"✘ [ERROR] Error rolling out application <appname> due to an internal error (request id: undefined): Internal Server Error"

We basically can't auto-deploy new containers anymore from CI. Pretty painful.

B

bentlegen•8/23/25, 4:09 AM

My latest attempt was to just try deploying 2x ... no luck.

T

TobiLG•8/23/25, 9:41 AM

I see a lot of RuntimeStartFailed messages when attempting to deploy as well... Deployments take >15min now for me

G

gruntlord6OP•8/23/25, 3:40 PM

I see a lot of errors about not being able to allocate containers, then they allocate anyway apprently

TTobiLG I see a lot of RuntimeStartFailed messages when attempting to deploy as well... ...

G

gruntlord6OP•8/23/25, 3:40 PM

as in the rolling deployments?

G

gruntlord6OP•8/23/25, 3:41 PM

or the actual push to cloudflare

Ggruntlord6 as in the rolling deployments?

T

TobiLG•8/23/25, 4:27 PM

Pre-existing application with updated images…

TTobiLG Pre-existing application with updated images…

G

gruntlord6OP•8/23/25, 4:47 PM

They take a while for me as well, and they are not exactly large images

I

iddogino•8/24/25, 1:21 AM

Experiencing a new issue and wondering if anyone has experienced it too. I have worker A conencting to worker B using a service binding (with

fetch

fetch

), and worker B uses a container. It used to work (last tested about a week ago), but as of yesterday it doesn't seem to work at all. Worker A makes the fetch request and it just gets stuck.
1. Running both locally with wrangler - everything works as expected (I can see A fetching from B locally).
2. Making a regular request to B works (when I switch to

fetch("https://worker-b.account.workers.dev/process/video"

fetch("https://worker-b.account.workers.dev/process/video"

fetch("https://worker-b.account.workers.dev/process/video"

fetch("https://worker-b.account.workers.dev/process/video" it works fine).

Anyone seeing something similar?

Iiddogino Experiencing a new issue and wondering if anyone has experienced it too. I have ...

D

darkpool•8/24/25, 2:46 AM

does

wrangler tail

wrangler tail

wrangler tail

wrangler tail show any logs?

Ddarkpool does `wrangler tail` show any logs?

I

iddogino•8/24/25, 4:09 PM

I can see the request going out from A but no request makes it to B...

T

TobiLG•8/25/25, 2:15 PM

Has anyone else problems when redeploying container projects? E.g. I see the following errors:

Checking for local image flexstore:6601d7c5 failed with error: Cannot read properties of undefined (reading 'split')

Checking for local image flexstore:6601d7c5 failed with error: Cannot read properties of undefined (reading 'split')

Checking for local image flexstore:6601d7c5 failed with error: Cannot read properties of undefined (reading 'split')

Checking for local image flexstore:6601d7c5 failed with error: Cannot read properties of undefined (reading 'split'). If I delete the containers app in the Dashboard, I can deploy again

T

Tory•8/25/25, 4:26 PM

What would cause container instances to just stop. These instances have been running steadily, without any code changes for several weeks. Then suddenly they all completely died.

D

deved9036•8/25/25, 7:47 PM

are there issues with deploying containers? just stopped working all of a sudden today for me

D

deved9036•8/25/25, 7:48 PM

seems to be mainly containers, also the observability is throwing a bunch of errors all of a sudden

T

TobiLG•8/25/25, 9:34 PM

yeah, same for me. Have you tried updating to the latest wrangler?

D

deved9036•8/25/25, 9:41 PM

i'll give it a god now

D

deved9036•8/25/25, 9:51 PM

okaayy nerverming! seems that fixed it lmao, works on 4.32.0

D

deved9036•8/25/25, 9:52 PM

@TobiLG you still having problem on the latest?

Ddeved9036 @TobiLG you still having problem on the latest?

T

TobiLG•8/25/25, 9:54 PM

not currently, but it was pretty unstable the last few days

D

deved9036•8/25/25, 9:55 PM

yeah, for sure!

J

julian7797•8/25/25, 11:33 PM

When using a container from another durable object, I appear to be running into issues, but from inside regular worker, all is well? Anyone else seen this

D

danamajid•8/26/25, 10:03 AM

hi @Mike Nomitch | Workers PM i'm getting:
"The image registry does not exist"

according to https://www.answeroverflow.com/m/1389977616509047009
you shared it's because the account was part of the private beta (which i my case it was).

can you pls help?

Anyone seen this before when pushing - Cloudflare Developers

Anyone seen this before when pushing their image?

npx wrangler@latest containers images list

 ⛅️ wrangler 4.22.0
───────────────────

Error listing images: ApiError: The image registry does not exist

npx wrangler@latest containers images list

 ⛅️ wrangler 4.22.0
───────────────────

Error listing images: ApiError: The image registry does not exist

npx wrangler@latest containers images list

 ⛅️ wrangler 4.22.0
───────────────────

Error listing images: ApiError: The image registry does not exist

npx wrangler@latest containers images list

 ⛅️ wrangler 4.22.0
───────────────────

Error listing images: ApiError: The image registry does not exist

supposed to be using the built-in registry, maybe I've missed something obvious

N

Nicolasmp•8/26/25, 11:57 AM

Hi there! Is there any way to have different

sleepAfter

sleepAfter

sleepAfter

sleepAfter variables according to env variables? For example, I want my container to stop after an hour in dev but never shutdown in prod.

NNicolasmp Hi there! Is there any way to have different `sleepAfter` variables according to...

K

kian•8/26/25, 5:58 PM

I think you can just set it to a

this.env.???

this.env.???

this.env.???

this.env.??? value

K

kian•8/26/25, 5:58 PM

export class MyContainer extends Container {
  defaultPort = 8080;
  sleepAfter = this.env.SLEEP_AFTER;
}

export class MyContainer extends Container {
  defaultPort = 8080;
  sleepAfter = this.env.SLEEP_AFTER;
}

export class MyContainer extends Container {
  defaultPort = 8080;
  sleepAfter = this.env.SLEEP_AFTER;
}

export class MyContainer extends Container {
  defaultPort = 8080;
  sleepAfter = this.env.SLEEP_AFTER;
}

Kkian ```js export class MyContainer extends Container { defaultPort = 8080; sleep...

N

Nicolasmp•8/26/25, 5:59 PM

Alright, and if that variable is undefined it will default to never sleeping?

K

kian•8/26/25, 6:00 PM

The default value is 10 minutes so that

S

Sean•8/26/25, 7:09 PM

Wondering if anyone has some guidance they can share - we have an AI process that runs in a container and takes about 15-20 minutes per invocation. This is triggered by an R2 event subscription where the consumer worker calls my Cloudflare Container app. Right now the issue is the proxy worker that forwards the request to the container is timing out at 180,000 ms (3 minutes) and causing the process to be retried. The actual process works within the deployed container, but we don't want it to be continuously retried by the platform timeout error. This seems to be the worker CPU time limit. Unfortunately pushing it up to 5 minutes won't help.

SSean Wondering if anyone has some guidance they can share - we have an AI process tha...

C

Chaika•8/26/25, 7:21 PM

Cpu time is not consumed by waiting on i/o like a fetch. If the client stays connected, there's no limit on duration. Do they all timeout at exactly 3 minutes? That doesn't really match any platform limit that comes to mind, hmm. I assume you're using a queue and calling the container worker via rpc from the queue consumer?

CChaika Cpu time is not consumed by waiting on i/o like a fetch. If the client stays con...

S

Sean•8/26/25, 7:32 PM

Yes, correct. I have a message worker (queue consumer) that makes a request to my CF Container app. The worker associated with the container app consistently errors out at ~180k ms

SSean Yes, correct. I have a message worker (queue consumer) that makes a request to m...

C

Chaika•8/26/25, 7:35 PM

Queue Consumers should have a max of 15 minutes duration time. Proxy Worker should stay alive as long as the request is pending. The container will sleep if no requests in 10 minutes by default, unless you've configured sleepAfter to 3 minutes.
If you click on the event, is there any more info on the error?

CChaika Queue Consumers should have a max of 15 minutes duration time. Proxy Worker shou...

S

Sean•8/26/25, 7:37 PM

Here's the full error

SSean Here's the full error

C

Chaika•8/26/25, 7:38 PM

That's from the container worker, not the queue consumer

CChaika That's from the container worker, not the queue consumer

S

Sean•8/26/25, 7:39 PM

Correct, that's where I'm getting the errors (the container worker). Apologies if I misstated.

SSean Correct, that's where I'm getting the errors (the container worker). Apologies i...

C

Chaika•8/26/25, 7:45 PM

Your queue consumer should be awaiting the fetch/rpc and then returning success, otherwise it'd just be building up the queue over and over. Are you doing something weird like waitUntil in it and ignoring errors?

S

Sean•8/26/25, 7:53 PM

@Chaika This is my queue consumer code:

Error details attempting to hit the container app start-async endpoint:

Msg 2000d8094472de33684f1967e7365521: Error processing message (attempt 1) for d1cab0d6-be82-4211-a6a9-c53c00dbff40: Error: HTTP error! status: 500 - {"error":"Internal Server Error"}

Msg 2000d8094472de33684f1967e7365521: Error processing message (attempt 1) for d1cab0d6-be82-4211-a6a9-c53c00dbff40: Error: HTTP error! status: 500 - {"error":"Internal Server Error"}

Msg 2000d8094472de33684f1967e7365521: Error processing message (attempt 1) for d1cab0d6-be82-4211-a6a9-c53c00dbff40: Error: HTTP error! status: 500 - {"error":"Internal Server Error"}

Msg 2000d8094472de33684f1967e7365521: Error processing message (attempt 1) for d1cab0d6-be82-4211-a6a9-c53c00dbff40: Error: HTTP error! status: 500 - {"error":"Internal Server Error"}

SSean @Chaika This is my queue consumer code: Error details attempting to hit the co...

C

Chaika•8/26/25, 8:18 PM

hmm, yea I don't think this is a container thing. Your worker log says cancelled, meaning the calling worker gave up on it.
Using a layer inbetween/Mastra Client is going to complicate getting the underlying error out, a quick google search shows there's a configurable timeout for Mastra which a default, worth looking into

CChaika hmm, yea I don't think this is a container thing. Your worker log says cancelled...

S

Sean•8/26/25, 8:19 PM

Ok thanks for your help

D

Dipan•8/27/25, 1:31 AM

Hey, I am deploying a Server in cloudflare container, deployed it successfully and get this response -

Total Upload: 53.47 KiB / gzip: 13.58 KiB
Worker Startup Time: 14 ms
Your worker has access to the following bindings:
- Durable Objects:
  - MORPH_CONTAINER: MorphContainer
- Vars:
  - NODE_ENV: "production"
Uploaded morph-server (9.65 sec)
Deployed morph-server triggers (1.76 sec)

Total Upload: 53.47 KiB / gzip: 13.58 KiB
Worker Startup Time: 14 ms
Your worker has access to the following bindings:
- Durable Objects:
  - MORPH_CONTAINER: MorphContainer
- Vars:
  - NODE_ENV: "production"
Uploaded morph-server (9.65 sec)
Deployed morph-server triggers (1.76 sec)

Total Upload: 53.47 KiB / gzip: 13.58 KiB
Worker Startup Time: 14 ms
Your worker has access to the following bindings:
- Durable Objects:
  - MORPH_CONTAINER: MorphContainer
- Vars:
  - NODE_ENV: "production"
Uploaded morph-server (9.65 sec)
Deployed morph-server triggers (1.76 sec)

Total Upload: 53.47 KiB / gzip: 13.58 KiB
Worker Startup Time: 14 ms
Your worker has access to the following bindings:
- Durable Objects:
  - MORPH_CONTAINER: MorphContainer
- Vars:
  - NODE_ENV: "production"
Uploaded morph-server (9.65 sec)
Deployed morph-server triggers (1.76 sec)

But when I am seng a request to the server getting this error (attached in screenshot)