This project is a helm wrapper to fix some bugs:
Also implement upgrade locker. When using multiple deploy with the same non-concurrent resource additional locks needed.
When you run helm-assistant upgrade application run new sub process helm upgrade <some arguments>
, pipe logs
and also run kubectl process for grab additional data about deployment.
Also, if you run deployment with flag --wait-for-jobs application check deployed job, and if status is Failed
application send helm signal to revert realise
wget https://github.com/SomeBlackMagic/helm-assistant/releases/latest/download/helm-assistant-linux-amd64
chmod +x helm-assistant-linux-amd64
mv helm-assistant-linux-amd64 /usr/local/bin/helm-assistant
export HELM_CMD_ARGS="--kubeconfig /root/.kube/some-cluster.yaml"
export KUBECTL_CMD_ARGS="--kubeconfig /root/.kube/some-cluster.yaml"
export HELM_ASSISTANT_UPGRADE_PIPE_LOGS=true
export HELM_ASSISTANT_UPGRADE_JOB_STRICT=true
export HELM_ASSISTANT_RELEASE_LOCK_ENABLED=true
helm-assistant upgrade \
--install helm-assistant-demo-deployment \
--namespace ${KUBE_NAMESPACE} \
--atomic \
--debug \
--wait \
./test/charts/worker
Warning: install argument must be first in arguments list because application grabs release name from cil args (first argument after upgrade and without "--")
Todolist:
- implement logs filter(helm/helm#7275)