After executing "deploy --sync" im getting this "Could not find Build action processor" after migrating a project to 0.13, is there anything how to resolve this, maybe something missing in the kubernetes cluster?

Hi there, I have a varfile as a dot env files: garden.env ---- url=https://app.${var.userId}.${environment.name}.example.com -- with a garden.yaml -- apiVersion: garden.io/v1 kind: Project name: myproject defaultEnvironment: "local" environments: - name: local varfile: garden.env __ and I use it in a pod manifest -- spec: manifests: - apiVersion: v1 kind: ConfigMap metadata: name: config data: $merge: ${var} LOG_LEVEL: info -- I would expect that the variable substituted, but it still shows up as url=https://app.${var.userId}.${environment.name}.example.com This has worked before, but now doesn't- garden version: 0.13.53 Could some one please help me debug this issue?

I’ve recently started leveraging spot instances to optimize workload management on my Kubernetes cluster, which has significantly improved bin packing efficiency. However, I’ve been encountering the following issue:

Failed to pull image "gardendev/k8s-util:0.6.2@sha256:f51e7ce040e2e23bc0eaa7216e4d976f13786d96773ef7b8c8f349e7a63d74e9"

The problem seems to be related to rate limits when pulling the

gardendev/k8s-util image.

This is now happening more since spot instance is trying to place the pods in appropriate nodes more often. I’m looking for a way to specify a custom image hosted in my ECR repository to avoid hitting these rate limits. I’m using Kaniko as the builder, and it already uses its own custom image, which is also hosted on ECR. Does anyone have any suggestions on how I can configure this to resolve the issue and prevent hitting these rate limits? My current provider configuration looks something like this:

providers:
  - name: kubernetes
    context: arn:aws:eks:us-east-1:123456789:cluster/garden
    buildMode: kaniko
    kaniko:
      image: 123456789.dkr.ecr.us-east-1.amazonaws.com/local-mirror:gcr.io-kaniko-project-executor-v1.11.0-debug
      extraFlags:
        - --snapshot-mode=redo
        - --use-new-run
      nodeSelector:
        node/namespace: "build"
      tolerations:
        - key: node
          value: build
      namespace: dara-${local.username}
    deploymentRegistry:
      hostname: 123456789.dkr.ecr.us-west-2.amazonaws.com
      namespace: garden

Also on a side note, I tried following the documentation here for utilImageRegistry setup https://docs.garden.io/reference/providers/kubernetes#complete-yaml-schema but it wasn't helpful as I kept getting error

Error validating provider configuration: key "utilImageRegistryDomain" is not allowed at path [utilImageRegistryDomain]. Available keys: dependencies, environments, buildMode, etc

I am using

garden version: 0.13.13

just fyi

when i try to build my project i get the following warning

WARNING: No output specified with docker-container driver. Build result will only remain in the build cache. To push result image into registry use --push or to load image into docker use --load

I can beliefe that this is not garden related, but maybe someone can help or have in the future the same issue. when i add to the

/Users/…/.garden/tools/docker/af81096fe0110658/docker/docker buildx build

command a

--load

the image is available for a deployment 🤔 System: macOS Garden: 0.13.54

How can one perform a "garden login" and "garden dev" on a remote server without UI and web browser? Thank you

i'm not able to figure out: - how to get

garden tools kubectl

to select the local-kubernetes provider (vs

garden tools local-kubernetes.kubectl')
- not able to get a garden exec action to use the right context without passing

--context=docker-desktop` to kubectl (if that's what's needed, how to reference it from the provider config?)(also, doesn't necessarily work for a kubectl plugin)

As part of my garden file for a microservice, I have a script which is being executed in a

run

block that's part of the deployment process and is listed as a dependency for the service. It does not need to be run when cleaning up. When the

type

for that block is

kubernetes

and I provide the manifest, this script is not run during

garden cleanup deploy

. When the type is

container

and I provide an image for the container, all scripts that are executed in

run

blocks are executed during the cleanup process. I'm not changing the dependencies, only the

type

for the service. AFAICT there shouldn't be any reason for these

run

blocks to be executing during cleanup ever, but I'm not sure why they specifically do when the

type

is container rather than

kubernetes

. This feels like a bug to me but I can't say for sure.

━━━ Latest events from Deployment backend-api ━━━
Deployment backend-api: ScalingReplicaSet - Scaled up replica set backend-api-d5f7f57fb to 1
Pod backend-api-d5f7f57fb-clqlf: Scheduled - Successfully assigned custom-runner/backend-api-d5f7f57fb-clqlf to runner
Pod backend-api-d5f7f57fb-clqlf: Pulling - Pulling image "backend-api:v-bdfcbe7b43"
Pod backend-api-d5f7f57fb-clqlf: Failed - Failed to pull image "backend-api:v-bdfcbe7b43": Error response from daemon: pull access denied for backend-api, repository does not exist or may require 'docker login': denied: requested access to the resource is denied
Pod backend-api-d5f7f57fb-clqlf: Failed - Error: ErrImagePull
Pod backend-api-d5f7f57fb-clqlf: BackOff - Back-off pulling image "backend-api:v-bdfcbe7b43"
Pod backend-api-d5f7f57fb-clqlf: Failed - Error: ImagePullBackOff

It seems it's trying to pull a built image from docker-hub and not locally.

Hi. I am working through the upgrade guide for upgrading to 0.14, and I've encountered an error that I don't quite understand. Is this something on my end, and if so could I get some advice on how to correct it? If posting any of my

yml

configurations will help please let me know. Thanks!

Command "/home/runner/.garden/tools/mutagen/722d0c1941ba875e/mutagen sync flush k8s--build-sync--monitor--grid-detect-monitor--nats--tl7rnnkr" failed with code 1:

Error: unable to flush session: session is not currently able to synchronize

Here's the full output:


Forcing synchronization cycle for session sync_2ALXRvkanBcP8pHpXTlz50iHljNn2VWHz
Error: unable to flush session: session is not currently able to synchronize

On Arch (CachyOS) and only recently (probably since 0.14) I can't self-update without specifying the version:

bash
❯ garden self-update
Update Garden

No installation directory specified via --install-dir option. Garden will be re-installed to the current installation directory: /home/rapha/.garden/bin
Checking for target and latest versions...
Current Garden version: 0.14.1
Target Garden version to be installed: undefined
Latest release version: 0.14.1

Downloading version undefined from https://download.garden.io/core/undefined/garden-undefined-linux-amd64.tar.gz...

Could not find version undefined for linux-amd64.
Here are the latest available versions: edge-acorn, edge-bonsai, edge-cedar, 0.14.1, 0.13.57, 0.14.0, 0.13.56, 0.13.55, 0.13.54, 0.13.53, 0.13.52, 0.13.51, 0.13.50

(This also happened before I manually updated:

garden self-update 0.14.1

)

Intermittently I get the following failure when running an action such as:

yml
---
kind: Build
name: dcd-linux-armv8-rpi-pi4
description: Build the deepcore-daemon for rpi4
type: container
spec:
  dockerfile: "docker/deepcore-daemon/linux-armv8.Dockerfile"
  platforms: [linux/arm64]
  # extraFlags: ["--push"]
  buildArgs:
    CARGO_C_MESON_IMAGE_ID: ${actions.build.cargo-c-meson.outputs.deploymentImageId}
  targetStage: runner-rpi4
  publishId: registry.gitlab.com/aivero/open-source/see/dcd-linux-armv8-rpi-pi4
timeout: 6000
dependencies: [build.cargo-c-meson, run.publish-acms]
# variables:
#   gstreamer-builder-container: ${actions.build.gstreamer-builder.outputs.deploymentImageId}
include:
  - docker/deepcore-daemon/linux-armv8.Dockerfile
...

bash
SYNCHRONOUS TERMINATION NOTICE: When explicitly exiting the process via process.exit or via a parent process, asynchronous tasks in your exitHooks will not run. Either remove these tasks, use gracefulExit() instead of process.exit(), or ensure your parent process sends a SIGINT to the process running this code.
node:events:497
      throw er; // Unhandled 'error' event
      ^

Error: write EPIPE
    at WriteWrap.onWriteComplete [as oncomplete] (node:internal/stream_base_commons:87:19)
    at WriteWrap.callbackTrampoline (node:internal/async_hooks:130:17)
Emitted 'error' event on Socket instance at:
    at emitErrorNT (node:internal/streams/destroy:170:8)
    at emitErrorCloseNT (node:internal/streams/destroy:129:3)
    at process.processTicksAndRejections (node:internal/process/task_queues:90:21) {
  errno: -32,
  code: 'EPIPE',
  syscall: 'write'
}

Node.js v22.9.0

Hi Team, I would like to know if Garden supports service dependency management. For example, I want to perform integration testing in an isolated Garden environment, where: Service A depends on Service B Service B depends on Service C Each service has its own garden.yaml file defined in its respective repository If a developer working on Service A wants to spin up Services B and C within the same ephemeral namespace, does Garden provide a way to achieve this? Specifically, can the developer specify the repository links for Services B and C in Service A’s garden.yaml file, so that Garden automatically deploys all required services within the same ephemeral namespace? Thanks!

Hi, after the recent update to Garden Cloud it seems like there have been some changes (correct me if I'm wrong): - The previous dashboard that showed the dependency graph is no longer accessible - The new dashboard, without the graph, is only available to 3 people on the free tier - There is no way to go back to the old dashboard Are there any plans to provide a locally running UI for the dependency graph, like the previous Garden Cloud UI? Will this be added to the new dashboard soon? The reason I'm asking is because the UI was one of the main reasons we adopted Garden for our usecase, which is exlusively for local development at this time. The development tasks for some members are also sporadic, it's not something most developers are using every day (more like every other month). If there are no plans to reintroduce the graph to the UI and provide access to a UI on the free tier, we lose our main usecase for using Garden. Even if it's reintroduced on the paid tier, there's not much point in us paying $60/month per developer when most developers only use it sporadically (even though it helps them out a lot when they do use it). Even though I'd not prefer to throw away all the work we've put into it, and the possibility of expanding the use of Garden, this is the direction things seem to be pointing towards as I'm looking at the state of things.

I'm trying to get the output of a run action, which is just a bash script that writes a string to stdout with 'echo`. I'm trying to get this output in my manifest template file in my deploy action and set it as an environment variable but I'm not able to.

garden.io/service: incident-...): Could not find key stdout under actions.run.initialize-mongodb.outputs. Available keys: (none).
────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────────

Invalid template string (apiVersion: apps/v1\nkind: Deployment\nmetadata:\n  name: service\n  labels:\n    garden.io/service: service\n  annotations:\n    garden.io/mode: default\n    garden.io/service: incident-...): Could not find key stdout under actions.run.initialize-mongodb.outputs. Available keys: (none).

I've started getting this message when I deploy a project (it appears to have started this morning with no changes on my end). Just wanted to report it since the message asked me to 🙂 thanks!

Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: Unable to release the lock on the global config store; lock was already released. If you run into this error repeatedly, please report on our GitHub or Discord
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: Unable to release the lock on the global config store; lock was already released. If you run into this error repeatedly, please report on our GitHub or Discord
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: The lock on the global config store was compromised. Updating the config anyway.
Warning: Unable to release the lock on the global config store; lock was already released. If you run into this error repeatedly, please report on our GitHub or Discord
Warning: Unable to release the lock on the global config store; lock was already released. If you run into this error repeatedly, please report on our GitHub or Discord

Hello! Thank you for the amazing software you have developed. I have mirrored the garden utility images to a private google artifact registry. I have the following config in my project

- name: local-kubernetes
  imagePullSecrets:
  - name: image-pull-credentials
    namespace: ${environment.namespace}
  utilImageRegistryDomain: <...docker.pkg.dev/...>

This works great for my containers, but it fails when fetching garden's "default-backend". It appears the

imagePullSecrets

is not set on the deployment. I believe this is the PR which introduced this functionality https://github.com/garden-io/garden/pull/6552 Im using latests garden v0.14.3

Hey! We have been using Garden 0.12 and planning to move on for version 0.13 then 0.14. The problem we are facing atm, is that we have a base-chart, that is declared as a Module. We use it to specify common variables across our charts. The way we used it in 0.12:

yaml
kind: Module
description: Base Chart
name: base-chart
type: helm
skipDeploy: true
allowPublish: false
spec:
  values: ...
---
kind: Module
name: chalco
type: helm
base: base-chart
values: ...

(We didn't find a migration path to apply the

skipDeploy: true

field also.) Could you please give us some ideas how we can do this in version 0.13 / 0.14?

I've been getting this error inconsistently but very often. I think that it's because garden is trying to deploy an image while it's still being built, or is trying to deploy more things at once than is allowed. I'm really not sure though. My environment is k3s with docker as the containerization runtime (not containerd). I'm using garden 0.13.57 for now while I work on migrating to 0.14 I upgraded to 0.14.3 and am seeing the same thing. Error:

SYNCHRONOUS TERMINATION NOTICE: When explicitly exiting the process via process.exit or via a parent process, asynchronous tasks in your exitHooks will not run. Either remove these tasks, use gracefulExit() instead of process.exit(), or ensure your parent process sends a SIGINT to the process running this code.
node:events:497
      throw er; // Unhandled 'error' event
      ^

Error: write EPIPE
    at afterWriteDispatched (node:internal/stream_base_commons:159:15)
    at writeGeneric (node:internal/stream_base_commons:150:3)
    at Socket._writeGeneric (node:net:958:11)
    at Socket._write (node:net:970:8)
    at writeOrBuffer (node:internal/streams/writable:572:12)
    at _write (node:internal/streams/writable:501:10)
    at Writable.write (node:internal/streams/writable:510:10)
    at Transform.<anonymous> (file:///home/ryan/.local/share/garden/1746479452-NIEIYT4.r/rollup/container-CIK3SmVD.mjs:565:37)
    at Transform.emit (node:events:519:28)
    at Transform.emit (node:domain:488:12)
Emitted 'error' event on Socket instance at:
    at emitErrorNT (node:internal/streams/destroy:170:8)
    at emitErrorCloseNT (node:internal/streams/destroy:129:3)
    at process.processTicksAndRejections (node:internal/process/task_queues:90:21) {
  errno: -32,
  code: 'EPIPE',
  syscall: 'write'
}

Node.js v22.9.0

This example from the documentation does not validate:

kind: Deploy
type: terraform
name: tf
autoApply: true

The exception thrown is:

Error validating terraform Deploy pubsub-tf 'pubsub-tf' (pubsub-utility):

/Users/jackhall/projects/kraken/pubsub-utility/garden.yaml:5
...
4  | name: pubsub-tf
5  | autoApply: true
----------------^
key "autoApply" is not allowed at path [autoApply]. Available keys: apiVersion, type, name, description, internal, source, dependencies, disabled, environments, include, exclude, variables, varfiles, spec, build, kind, timeout)

It seems that provider outputs are available in the action template context, but not the other way around. Can I map a project output to ${actions.<my-action.outputs.deploymentImageId} ?

We are looking to manage part of our infrastructure using ACK controllers (https://aws.amazon.com/blogs/containers/aws-controllers-for-kubernetes-ack/). Does anybody know/have experience with Garden+ACK?

Hi, im getting this error: Failed getting status for service 'reports-worker' (from module 'reports-worker'). Here is the output: ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Cannot read properties of null (reading 'key') ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ This causes that changing anything in the reports-worker dont cause a hot reload or if i modify the garden.yml the deployment does not start. How can i debug where this is comming from because there is no obvious log even with the most verbose settings.

I have services which use garden + terraform. For some reason they randomly have their secrets wiped out which are stored in secrets manager. I intialy create a secret in secret manager using terraform with an intial value later I manully add more values to it and thus the service can use it but randomly not everytime they get wiped out and get reset to the intial value I have the below block to prevent this but its still happening any suggestions ? resource "aws_secretsmanager_secret_version" "app_secret_version" { secret_id = aws_secretsmanager_secret.app_secret.id secret_string = var.secret_string lifecycle { ignore_changes = all } }

I am trying to display project outputs after running a Build actions. The reference in `project.garden.yaml `is

outputs:
  - name: app_image
    value: "${actions.build.app.outputs.deploymentImageName}"

Running

garden get outputs

the Build actions are triggered but I receive the following error:

...(removed for char limits)
ℹ build.app                 → Getting status for Build app (type container) at version v-e5b0c8e672...
ℹ build.app2                → Getting status for Build app2 (type container) at version v-b0365ed458...
ℹ build.app                 → Building app (type container) at version v-e5b0c8e672...
ℹ build.app2                → Building app2 (type container) at version v-b0365ed458...
ℹ build.app                 → Building app:v-ef287852fe...
ℹ build.app2                → Building app2:v-c41698070a...
✔ build.app2                → Done (took 1.1 sec)
✔ build.app                 → Done (took 1.2 sec)
project.garden.yaml:58
...
57  |   - name: app_image
58  |     value: "${actions.build.app.outputs.deploymentImageName}"
-----------------^
Could not find key actions. Available keys: local, command, datetime, project, git, secrets, variables, var, environment, providers, modules, runtime.

See .garden/error.log for detailed error message

My garden project api version is set to

apiVersion: garden.io/v2

and my garden version is

garden version: 0.14.7

. Why aren't actions outputs available? Very strange that

modules

is available since it is a not supported in

0.14

afaik. Anyway, any help would be appreciated! Thanks!

Hey! I saw this forum uses Linen to sync Discord threads to the web. I’m building [Paladin](https://trypaladin.com/), a smoother, cleaner alternative that makes Discord forums easier to browse on the web. It’s free to use and can run alongside Linen, so you don’t have to switch anything to try it out. Would you be interested in trying it out? 😀 (example preview) https://cdn.discordapp.com/attachments/1423846090104373368/1423846090385526834/checkoutly_image_1.jpg?ex=68e1cb81&is=68e07a01&hm=c433b66107ebc56d0890528e77780ac58aed0654f15d3b70000908e22090de7a&

Hi, I'd am trying to configure Garden builds to use registry caching so that builds are faster in CI in situations where the image does not change. I'd like to be able to do this without using the Remote Container Builder since we currently do not share images etc. Just something simple in CI to speed up builds for each team. I know this can easily be done outside of Garden using the

setup-buildx-action@v3

+

build-push-action@v6

(with cache-from, cache-to set). BUT I want to do it using Garden 🙂 Is there a recommended method to get this working (or an example I can look at)? I have tried the following but it seems to work sporadically (sometimes cache hits or misses on builds even if nothing changes). First, I'd like to confirm this method is valid before I troubleshoot any further, but also am open to suggestions/improvements to get this design working. (more info in thread due to message lenght issues

hi everyone, after reading through the doc, i find it only mentions Garden Cloud where you do

garden login

or checking garden dashboard on SaaS. i am curious what are missing in OSS and if i can easily self-host it on my own K8s cluster (e.g. do you plan to provide a Helm Chart)

We are moving Garden community conversations/questions from Discord and to GitHub discussions. You can access Garden Discussions on GitHub here: https://github.com/garden-io/garden/discussions Why? We're happy Discord users but we simply haven't had the bandwidth recently to keep a close eye on the Discord community and answer questions as needed. We've also gotten requests from users that don't have a Discord account that we enable GitHub Discussions. So keeping everything on GitHub will simplify things and allows us and others to respond faster to questions. Timeline We've already enabled GitHub Discussions and are actively monitoring it. Our Discord community server will officially be sunset on January 26th.

#1044642930075373588 Say I needed to investigate how Oauth flow is working at the moment as there seems to be issue with certain time outs, how can garden help? #1044642930075373588

Greeting what the fastest way to deploy a bunch of action all at once? I want to deploy 50 see Ives but it look like it only deploys a couple at a time