[BUG] KV Population fails when warming the cache

[roybarber]

Describe the bug

Hey guys, is there a way i can debug whats going on here: Locally on my macbook i can deploy, on my mac pro i cant (get the same issue) and every standard deployment build fails in the cloudflare dashboard with the following:

13:50:15.573	Populating KV incremental cache...
13:50:15.584	Inserting 200 assets to KV in chunks of 25
13:50:15.972	ERROR Wrangler command failed
13:50:16.032	12%|█████ | 1/8 [00:00:<00:00:, 0.00it/s] ELIFECYCLE  Command failed with exit code 1.
13:50:16.061	Failed: error occurred while running deploy command

Thats all, nothing is written to KV. Can we skip this?

I've downgraded openNext to a older version as we have another replicated project thats running fine.

Steps to reproduce

Deploy a ISR project using the recommended setup

My wrangler file:

{
    "$schema": "node_modules/wrangler/config-schema.json",
    "main": ".open-next/worker.js",
    "name": "uber-uat",
    "compatibility_date": "2025-11-17",
    "compatibility_flags": ["nodejs_compat", "global_fetch_strictly_public"],
    "account_id": "XXX",
    "minify": true,
    "assets": {
        "directory": ".open-next/assets",
        "binding": "ASSETS"
    },
    "observability": {
        "enabled": true
    },
    "placement": {
        "mode": "smart"
    },
    "preview_urls": false,
    "keep_vars": true,
    "durable_objects": {
        "bindings": [
            {
                "name": "NEXT_CACHE_DO_QUEUE",
                "class_name": "DOQueueHandler"
            },
            {
                "name": "NEXT_TAG_CACHE_DO_SHARDED",
                "class_name": "DOShardedTagCache"
            },
            {
                "name": "NEXT_CACHE_DO_PURGE",
                "class_name": "BucketCachePurge"
            }
        ]
    },
    "migrations": [
        {
            "tag": "v1",
            "new_sqlite_classes": ["DOQueueHandler", "DOShardedTagCache", "BucketCachePurge"]
        }
    ],
    "kv_namespaces": [
        {
            "binding": "NEXT_INC_CACHE_KV",
            "id": "XXX"
        }
    ],
    "services": [
        {
            "binding": "WORKER_SELF_REFERENCE",
            "service": "uber-uat"
        }
    ]
}

Opennext config:

import { defineCloudflareConfig } from '@opennextjs/cloudflare'
import { purgeCache } from '@opennextjs/cloudflare/overrides/cache-purge/index'
import kvIncrementalCache from '@opennextjs/cloudflare/overrides/incremental-cache/kv-incremental-cache'
import doQueue from '@opennextjs/cloudflare/overrides/queue/do-queue'
import doShardedTagCache from '@opennextjs/cloudflare/overrides/tag-cache/do-sharded-tag-cache'

export default defineCloudflareConfig({
    incrementalCache: kvIncrementalCache,
    queue: doQueue,
    tagCache: doShardedTagCache({
        baseShardSize: 12,
        regionalCache: true, // Enable regional cache to reduce the load on the DOs
        regionalCacheTtlSec: 5, // The TTL for the regional cache
        shardReplication: {
            numberOfSoftReplicas: 4,
            numberOfHardReplicas: 2,
            regionalReplication: {
                defaultRegion: 'enam'
            }
        }
    }),
    enableCacheInterception: true,
    // you can also use the `durableObject` option to use a durable object as a cache purge
    cachePurge: purgeCache({ type: 'durableObject' })
})

Expected behavior

Cache populates

@opennextjs/cloudflare version

1.14.3

Wrangler version

4.53.0

next info output

Operating System:
  Platform: darwin
  Arch: arm64
  Version: Darwin Kernel Version 24.6.0: Mon Jul 14 11:30:29 PDT 2025; root:xnu-11417.140.69~1/RELEASE_ARM64_T6000
  Available memory (MB): 32768
  Available CPU cores: 10
Binaries:
  Node: 22.13.0
  npm: 11.6.0
  Yarn: N/A
  pnpm: 10.22.0
Relevant Packages:
  next: 15.5.5
  eslint-config-next: N/A
  react: 19.2.0
  react-dom: 19.2.0
  typescript: 5.9.3
Next.js Config:
  output: N/A

Additional context

No response

[vicb]

Locally on my macbook i can deploy, on my mac pro i cant

This make me think that you might not be logged in on your Mac Pro

Can you try pnpm wrangler whoami to get the info

If you see

-> % pnpm wrangler whoami

 ⛅️ wrangler 4.49.1 (update available 4.53.0)
─────────────────────────────────────────────
Getting User settings...
You are not authenticated. Please run `wrangler login`.

Then please run wrangler login and try to deploy again

[roybarber]

Can confirm logged in on both macpro and my laptop. But its doing the same in the worker dashboard which made me think its nothing to do with auth

[vicb]

Can you try to update @opennextjs/cloudflare to the version from #1034 (using "https://pkg.pr.new/@opennextjs/cloudflare@1034" in your package.json) and report what the error is?

[roybarber]

@vicb

thank you, deployed with the pkg.pr version and heres the end of the logs

13:51:59.035 | Populating KV incremental cache...
-- | --
13:51:59.048 | Inserting 245 assets to KV in chunks of 25
13:51:59.530 | 10%\|████                                         \| 1/10 [00:00:<00:00:, 0.00it/s]undefined
13:51:59.530 | ERR_PNPM_RECURSIVE_EXEC_FIRST_FAIL  Command "wrangler" not found
13:51:59.530 |  
13:51:59.530 | Did you mean "pnpm exec changeset"?
13:51:59.531 | ERROR Wrangler command failed
13:51:59.617 | ELIFECYCLE  Command failed with exit code 1.
13:51:59.652 | Failed: error occurred while running deploy command

[vicb]

Command "wrangler" not found

Does you package have a dependency on wrangler ?

[roybarber]

@vicb this is copied from the cloudflare dashboard build logs. We have wrangler installed as a dependancy in the project but dont use it in the pnpm build/deply commands.

we run:pnpm run deploy:cloudflare:uat that generatyes the above errors

"scripts": {
        "build": "next build",
        "build:cloudflare": "opennextjs-cloudflare build",
        "deploy:cloudflare": "opennextjs-cloudflare build && opennextjs-cloudflare deploy",
        "deploy:cloudflare:uat": "ENVIRONMENT=uat pnpm deploy:cloudflare"
},

[vicb]

opennextjs-cloudflare deploy do call wrangler and it sounds like this is where the error originates

[roybarber]

@vicb exactly, the build all works fine, its the deploy which fails on KV warming

[roybarber]

Although ive noticed a double build happens as the commands duplicate on workers. It builds, then runs the deploy, which builds again and then deploys

[vicb]

The deploy script build because that's how you have implemented it:

        "deploy:cloudflare": "opennextjs-cloudflare build && opennextjs-cloudflare deploy",
        "deploy:cloudflare:uat": "ENVIRONMENT=uat pnpm deploy:cloudflare"

One of the ways to solve this would be to:

• set ENVIRONMENT=uat in the build env vars
• use opennextjs-cloudflare deploy as the deploy command

See https://opennext.js.org/cloudflare/howtos/dev-deploy#deploy-your-application-to-cloudflare-workers

[roybarber]

using a seperate build and deploy command means the deploy cant find the build se below:

14:20:43.124 | Success: Build command completed
-- | --
14:20:43.252 | Executing user deploy command: pnpm run deploy:uat
14:20:43.553 |  
14:20:43.553 | > frontend@0.1.0 deploy:uat /opt/buildhome/repo/apps/frontend
14:20:43.553 | > ENVIRONMENT=uat opennextjs-cloudflare deploy
14:20:43.553 |  
14:20:44.754 |  
14:20:44.755 | ┌──────────────────────────────┐
14:20:44.755 | │ OpenNext — Cloudflare deploy │
14:20:44.755 | └──────────────────────────────┘
14:20:44.755 |  
14:20:44.755 | ERROR Could not find compiled Open Next config, did you run the build command?
14:20:44.781 | ELIFECYCLE  Command failed with exit code 1.
14:20:44.807 | Failed: error occurred while running deploy command

So instead ive emptied the build command and left the deploy command to build first. however this also fails because it cant find a build

14:20:43.124 | Success: Build command completed
-- | --
14:20:43.252 | Executing user deploy command: pnpm run deploy:uat
14:20:43.553 |  
14:20:43.553 | > frontend@0.1.0 deploy:uat /opt/buildhome/repo/apps/frontend
14:20:43.553 | > ENVIRONMENT=uat opennextjs-cloudflare deploy
14:20:43.553 |  
14:20:44.754 |  
14:20:44.755 | ┌──────────────────────────────┐
14:20:44.755 | │ OpenNext — Cloudflare deploy │
14:20:44.755 | └──────────────────────────────┘
14:20:44.755 |  
14:20:44.755 | ERROR Could not find compiled Open Next config, did you run the build command?
14:20:44.781 | ELIFECYCLE  Command failed with exit code 1.
14:20:44.807 | Failed: error occurred while running deploy command

so instead we have to add in the build and deploy commands as my screenshot above.

This gets us to this:

14:27:55.046 |  
-- | --
14:27:55.587 | workerd/server/server.c++:1953: warning: A DurableObjectNamespace in the config referenced the class "BucketCachePurge", but no such Durable Object class is exported from the worker. Please make sure the class name matches, it is exported, and the class extends 'DurableObject'. Attempts to call to this Durable Object class will fail at runtime, but historically this was not a startup-time error. Future versions of workerd may make this a startup-time error.
14:27:55.587 | workerd/server/server.c++:1953: warning: A DurableObjectNamespace in the config referenced the class "DOShardedTagCache", but no such Durable Object class is exported from the worker. Please make sure the class name matches, it is exported, and the class extends 'DurableObject'. Attempts to call to this Durable Object class will fail at runtime, but historically this was not a startup-time error. Future versions of workerd may make this a startup-time error.
14:27:55.587 | workerd/server/server.c++:1953: warning: A DurableObjectNamespace in the config referenced the class "DOQueueHandler", but no such Durable Object class is exported from the worker. Please make sure the class name matches, it is exported, and the class extends 'DurableObject'. Attempts to call to this Durable Object class will fail at runtime, but historically this was not a startup-time error. Future versions of workerd may make this a startup-time error.
14:27:55.886 |  
14:27:55.886 | Populating KV incremental cache...
14:27:55.902 | Inserting 245 assets to KV in chunks of 25
14:27:56.470 | 10%\|████                                         \| 1/10 [00:00:<00:00:, 0.00it/s]undefined
14:27:56.471 | ERR_PNPM_RECURSIVE_EXEC_FIRST_FAIL  Command "wrangler" not found
14:27:56.471 |  
14:27:56.471 | Did you mean "pnpm exec changeset"?
14:27:56.472 | ERROR Wrangler command failed
14:27:56.596 | ELIFECYCLE  Command failed with exit code 1.
14:27:56.634 | Failed: error occurred while running deploy command

setting ENVIRONMENT in vars still returns the same error, we validated that one early.

13:51:59.530 | 10%\|████                                         \| 1/10 [00:00:<00:00:, 0.00it/s]undefined

Is it this undefined thats the clue?

[vicb]

Not sure what's going on here.

I can take a look if you upload a minimal repro (strip most of you app routes ans upload the code to a repo)