failed to create cluster

[David-eb97]

Environment

Device and OS: vmware Ubuntu 22.04.5
App version: latest
Kubernetes distro being used: latest
Other: this is a brand new install

Steps to reproduce

1. Step 2: Deploy the UDS Bundle

uds deploy k3d-core-demo:latest y

Expected result

install the cluster on the OS

Actual Result

fails to install cluster see code below

Visual Proof (screenshots, videos, text, etc)

see code below

Severity/Priority

high?

CODE:

`d@uds:~$ uds deploy k3d-core-demo:latest

NOTE Saving log file to /tmp/uds-2025-05-19-16-47-39-3453232787.log

🎁 BUNDLE DEFINITION

Metadata: information about this bundle

name: k3d-core-demo
description: A UDS bundle for deploying the standard UDS Core package on a development cluster
version: 0.42.0
architecture: amd64

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Build: info about the machine, UDS version, and the user that created this bundle

terminal: pkrvmjydaqcps29
user: runner
architecture: amd64
timestamp: Mon, 12 May 2025 20:47:15 +0000
version: v0.27.2

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Packages: definition of packages this bundle deploys, including variable overrides

name: uds-k3d-dev
ref: 0.12.3@sha256:66fbb0d0f7a5c238053f9a50552eea5a353ff8a34aff623db3ff7047246930fc
repo: ghcr.io/defenseunicorns/packages/uds-k3d

overrides: []

name: init
ref: v0.54.0@sha256:fe32d285a49564045adf61990e394696e44272c5856345304b7769bb2495c753
repo: ghcr.io/zarf-dev/packages/init

overrides: []

name: core
path: build/zarf-package-core-amd64-0.42.0.tar.zst
ref: 0.42.0@sha256:7d1d07d1bb62f45826fca4d24b17b6bd2406dca27dab00abd9a66f7eb7fc2ef9

overrides:

• loki:
  variables:
  LOKI_BACKEND_REPLICAS: 1
  LOKI_READ_REPLICAS: 1
  LOKI_WRITE_REPLICAS: 1
• module:
  variables:
  PEPR_ADMISSION_CPU_REQUEST: 200m
  PEPR_ADMISSION_MEMORY_REQUEST: 256Mi
  PEPR_WATCHER_CPU_REQUEST: 200m
  PEPR_WATCHER_MEMORY_REQUEST: 256Mi

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
? Deploy this bundle? Yes
✔ Successfully pulled: uds-k3d-dev (83.52 KBs)

kind: ZarfPackageConfig

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
metadata: information about this package

name: uds-k3d-dev
description: "UDS K3d Cluster Setup. WARNING: This will destroy the cluster if it already exists."
version: 0.12.3
url: https://github.com/defenseunicorns/uds-k3d
architecture: amd64
yolo: true
aggregateChecksum: bfe305fbe767c28726ed3a9961d25d06b99a5bcfc3d49be5a3d1349fc85117a5

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
build: info about the machine, zarf version, and user that created this package

terminal: fv-az1922-614
user: runner
architecture: amd64
timestamp: Wed, 16 Apr 2025 18:05:04 +0000
version: v0.51.0
migrations:

• scripts-to-actions
• pluralize-set-variable
  lastNonBreakingVersion: v0.27.0

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
components: components selected for this operation

• name: destroy-cluster
  description: Optionally destroy the cluster before creating it
  required: true
  actions:
  onDeploy:
  before:
  - cmd: k3d cluster delete ${ZARF_VAR_CLUSTER_NAME}
  description: Destroy the cluster

• name: create-cluster
  description: Create the k3d cluster
  required: true
  actions:
  onDeploy:
  before:
  - cmd: "k3d_version=$(k3d version | grep -E -o "([0-9]+\.?){3}$")\nrequired_version="5.7.1"\nif ! echo "$required_version\n$k3d_version" | sort -V -C; then\n echo "This package requires a minimum k3d version of $required_version"\n echo "Please upgrade k3d (https://k3d.io/stable/#install-current-latest-release) and try again"\n exit 1\nfi\n"
  description: Check k3d version compatibility
  - cmd: "k3d cluster create \\n-p "80:80@server:" \\n-p "443:443@server:" \\n--api-port 6550 \\n--k3s-arg "--disable=traefik@server:" \\n--k3s-arg "--disable=metrics-server@server:" \\n--k3s-arg "--disable=servicelb@server:" \\n--k3s-arg "--disable=local-storage@server:" \\n--image ${ZARF_VAR_K3D_IMAGE} ${ZARF_VAR_K3D_EXTRA_ARGS} \\n${ZARF_VAR_CLUSTER_NAME}\n"
  description: Create the cluster
  onSuccess:
  - description: Wait for CoreDNS to be ready
  wait:
  cluster:
  kind: Pod
  name: k8s-app=kube-dns
  namespace: kube-system
  condition: Ready
  - cmd: |
  echo "You can access this cluster over SSH (note http redirect will redirect to port 80 instead of 8080):"
  echo "ssh -N -L 8080:localhost:80 -L 8443:localhost:443 -L 6550:localhost:6550"
  echo
  echo "To get the kubeconfig:"
  echo "k3d kubeconfig get ${ZARF_VAR_CLUSTER_NAME}"
  echo
  echo "This cluster can be destroyed with:"
  echo "k3d cluster delete ${ZARF_VAR_CLUSTER_NAME}"
  description: Print out information about how to access the cluster remotely

• name: uds-dev-stack
  description: Install MetalLB, NGINX, Minio, local-path-rwx and Ensure MachineID to meet UDS developer needs without later config changes
  required: true
  charts:

  • name: metallb
    version: 0.14.9
    url: https://metallb.github.io/metallb
    namespace: uds-dev-stack
  • name: uds-dev-stack
    version: 0.12.3
    localPath: chart
    namespace: uds-dev-stack
    valuesFiles:
    • values/dev-stack-values.yaml

━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
variables: deployment-specific values that are set on each package deployment
- name: COREDNS_OVERRIDES
description: CoreDNS overrides
path: coreDnsOverrides

• name: minio
  version: 5.4.0
  url: https://charts.min.io/
  namespace: uds-dev-stack
  valuesFiles:

  • values/minio-values.yaml
    actions:
    onDeploy:
    before:
    • cmd: ./zarf tools kubectl get nodes -o=jsonpath='{.items[0].status.addresses[?(@.type=="InternalIP")].address}' | cut -d'.' -f1-3
      setVariables:
      • name: BASE_IP
        description: Load network ip base for MetalLB
        after:
    • cmd: ./zarf tools kubectl rollout restart deployment coredns -n kube-system
      description: Restart CoreDNS to pick up internal DNS override for uds.dev
      variables:

• name: CLUSTER_NAME currently set to 'uds' (default)
  description: Name of the cluster
  default: uds

• name: K3D_IMAGE currently set to 'rancher/k3s:v1.31...' (default)
  description: K3d image to use
  default: rancher/k3s:v1.31.7-k3s1

• name: K3D_EXTRA_ARGS currently set to '' (default)
  description: Optionally pass k3d arguments to the default

• name: NGINX_EXTRA_PORTS currently set to '[]' (default)
  description: Optionally allow more ports through Nginx (combine with K3D_EXTRA_ARGS '-p :@server:*')
  default: "[]"

• name: DOMAIN currently set to 'uds.dev' (default)
  description: Cluster domain
  default: uds.dev

• name: ADMIN_DOMAIN currently set to '' (default)
  description: Domain for admin services, defaults to admin.DOMAIN
  2025-05-19 16:47:45 INF deploying component name=destroy-cluster
  2025-05-19 16:47:45 INF running command cmd=Destroy the cluster
  2025-05-19 16:47:45 INF waiting for action (no timeout) cmd=Destroy the cluster
  2025-05-19 16:47:45 INF action succeeded cmd=Destroy the cluster stdout=INFO[0000] No nodes found for cluster 'uds', nothing to delete.
  INFO[0000] No clusters found
  stderr=ERRO[0000] Failed to get nodes for cluster 'uds': docker failed to get containers with labels 'map[k3d.cluster:uds]': failed to list containers: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?

2025-05-19 16:47:45 INF deploying component name=create-cluster
2025-05-19 16:47:45 INF running command cmd=Check k3d version compatibility
2025-05-19 16:47:45 INF waiting for action (no timeout) cmd=Check k3d version compatibility
2025-05-19 16:47:45 INF action succeeded cmd=Check k3d version compatibility stdout= stderr=
2025-05-19 16:47:45 INF running command cmd=Create the cluster
2025-05-19 16:47:45 INF waiting for action (no timeout) cmd=Create the cluster
2025-05-19 16:47:45 WRN action failed cmd=Create the cluster stdout=INFO[0000] Prep: Network
INFO[0000] Deleting cluster 'uds'
stderr=ERRO[0000] Failed to get nodes for cluster 'uds': docker failed to get containers with labels 'map[k3d.cluster:uds]': failed to list containers: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?
ERRO[0000] Failed Cluster Preparation: Failed Network Preparation: failed to create cluster network: failed to check for duplicate docker networks: docker failed to list networks: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?
ERRO[0000] Failed to create cluster >>> Rolling Back
ERRO[0000] Failed to get nodes for cluster 'uds': docker failed to get containers with labels 'map[k3d.cluster:uds]': failed to list containers: Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?
ERRO[0000] failed to get cluster: No nodes found for given cluster
FATA[0000] Cluster creation FAILED, also FAILED to rollback changes!

 ERROR:  failed to deploy bundle: unable to deploy component "create-cluster": unable to run component before action: command "Create the cluster" failed after 0 retries

`

[mjnagel]

Hey @David-eb97 based on the error here - Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running? it looks like you might not have a running docker daemon. Do you have docker installed/running on this machine (i.e. can you docker run -it busybox echo hello successfully)? If not you could walk through this - https://docs.docker.com/engine/install/ubuntu/ - since this is Ubuntu.

[David-eb97]

thank you for the update... my question is wouldn't that be a part of the initial install of the software?

[mjnagel]

UDS-CLI / UDS Core will not handle that for you. This is one of the prerequisites for deploying k3d - https://uds.defenseunicorns.com/getting-started/basic-requirements/#setup-guides

[David-eb97]

understood about docker, so i installed docker from source and reinstalled k3d and i got the error messages in the log file

uds-2025-06-06-20-39-49-2128022821.log

[mjnagel]

Not super clear from that error what the issue might be. I would double check a few things:

• Can you run a basic docker image/container (like the example provided before)
• Can you run a basic k3d cluster (k3d cluster create)

If you can do both of those then you could retry the full demo deployment and see again if it gets stuck.