132 lines
3.6 KiB
Makefile
132 lines
3.6 KiB
Makefile
set fallback := true
|
|
|
|
export OLLAMA_NAMESPACE := env("OLLAMA_NAMESPACE", "ollama")
|
|
export OLLAMA_CHART_VERSION := env("OLLAMA_CHART_VERSION", "1.35.0")
|
|
export OLLAMA_HOST := env("OLLAMA_HOST", "")
|
|
export OLLAMA_GPU_ENABLED := env("OLLAMA_GPU_ENABLED", "")
|
|
export OLLAMA_GPU_TYPE := env("OLLAMA_GPU_TYPE", "nvidia")
|
|
export OLLAMA_GPU_COUNT := env("OLLAMA_GPU_COUNT", "1")
|
|
export OLLAMA_MODELS := env("OLLAMA_MODELS", "")
|
|
export OLLAMA_STORAGE_SIZE := env("OLLAMA_STORAGE_SIZE", "30Gi")
|
|
|
|
[private]
|
|
default:
|
|
@just --list --unsorted --list-submodules
|
|
|
|
# Add Helm repository
|
|
add-helm-repo:
|
|
helm repo add ollama https://otwld.github.io/ollama-helm
|
|
helm repo update ollama
|
|
|
|
# Remove Helm repository
|
|
remove-helm-repo:
|
|
helm repo remove ollama
|
|
|
|
# Create Ollama namespace
|
|
create-namespace:
|
|
#!/bin/bash
|
|
set -euo pipefail
|
|
if ! kubectl get namespace ${OLLAMA_NAMESPACE} &>/dev/null; then
|
|
kubectl create namespace ${OLLAMA_NAMESPACE}
|
|
fi
|
|
kubectl label namespace ${OLLAMA_NAMESPACE} \
|
|
pod-security.kubernetes.io/enforce=restricted \
|
|
pod-security.kubernetes.io/enforce-version=latest \
|
|
pod-security.kubernetes.io/warn=restricted \
|
|
pod-security.kubernetes.io/warn-version=latest \
|
|
--overwrite
|
|
|
|
# Delete Ollama namespace
|
|
delete-namespace:
|
|
kubectl delete namespace ${OLLAMA_NAMESPACE} --ignore-not-found
|
|
|
|
# Install Ollama
|
|
install:
|
|
#!/bin/bash
|
|
set -euo pipefail
|
|
|
|
just create-namespace
|
|
just add-helm-repo
|
|
|
|
if [ -z "${OLLAMA_GPU_ENABLED}" ]; then
|
|
if gum confirm "Enable GPU support?"; then
|
|
OLLAMA_GPU_ENABLED="true"
|
|
else
|
|
OLLAMA_GPU_ENABLED="false"
|
|
fi
|
|
fi
|
|
|
|
if [ -z "${OLLAMA_MODELS}" ]; then
|
|
OLLAMA_MODELS=$(
|
|
gum input --prompt="Models to pull (comma-separated): " --width=100 \
|
|
--placeholder="e.g., llama3.2:1b,deepseek-r1:7b" \
|
|
--value="llama3.2:1b"
|
|
)
|
|
fi
|
|
|
|
gomplate -f values.gomplate.yaml -o values.yaml
|
|
helm upgrade --install ollama ollama/ollama \
|
|
--version ${OLLAMA_CHART_VERSION} -n ${OLLAMA_NAMESPACE} --wait \
|
|
-f values.yaml
|
|
|
|
echo ""
|
|
echo "Ollama installed successfully"
|
|
echo "GPU enabled: ${OLLAMA_GPU_ENABLED}"
|
|
echo "Models: ${OLLAMA_MODELS}"
|
|
|
|
# Upgrade Ollama
|
|
upgrade:
|
|
#!/bin/bash
|
|
set -euo pipefail
|
|
|
|
if [ -z "${OLLAMA_GPU_ENABLED}" ]; then
|
|
if gum confirm "Enable GPU support?"; then
|
|
OLLAMA_GPU_ENABLED="true"
|
|
else
|
|
OLLAMA_GPU_ENABLED="false"
|
|
fi
|
|
fi
|
|
|
|
if [ -z "${OLLAMA_MODELS}" ]; then
|
|
OLLAMA_MODELS=$(
|
|
gum input --prompt="Models to pull (comma-separated): " --width=100 \
|
|
--placeholder="e.g., llama3.2:1b,deepseek-r1:7b" \
|
|
--value="llama3.2:1b"
|
|
)
|
|
fi
|
|
|
|
gomplate -f values.gomplate.yaml -o values.yaml
|
|
helm upgrade ollama ollama/ollama \
|
|
--version ${OLLAMA_CHART_VERSION} -n ${OLLAMA_NAMESPACE} --wait \
|
|
-f values.yaml
|
|
|
|
echo "Ollama upgraded successfully"
|
|
|
|
# Uninstall Ollama
|
|
uninstall:
|
|
#!/bin/bash
|
|
set -euo pipefail
|
|
helm uninstall ollama -n ${OLLAMA_NAMESPACE} --wait --ignore-not-found
|
|
just delete-namespace
|
|
echo "Ollama uninstalled"
|
|
|
|
# Pull a model
|
|
pull model:
|
|
kubectl exec -it -n ${OLLAMA_NAMESPACE} deploy/ollama -- ollama pull {{ model }}
|
|
|
|
# Run a model
|
|
run model:
|
|
kubectl exec -it -n ${OLLAMA_NAMESPACE} deploy/ollama -- ollama run {{ model }}
|
|
|
|
# List models
|
|
list:
|
|
kubectl exec -it -n ${OLLAMA_NAMESPACE} deploy/ollama -- ollama list
|
|
|
|
# Show Ollama logs
|
|
logs:
|
|
kubectl logs -n ${OLLAMA_NAMESPACE} deploy/ollama -f
|
|
|
|
# Get pod status
|
|
status:
|
|
kubectl get pods -n ${OLLAMA_NAMESPACE}
|