Quickstart Guide
This guide covers how you can quickly get started using Helm.
Prerequisites
The following prerequisites are required for a successful and properly secured use of Helm.
- A Kubernetes cluster
- Deciding what security configurations to apply to your installation, if any
- Installing and configuring Helm and Tiller, the cluster-side service.
Install Kubernetes or have access to a cluster
- You must have Kubernetes installed. For the latest release of Helm, we recommend the latest stable release of Kubernetes, which in most cases is the second-latest minor release.
- You should also have a local configured copy of
kubectl.
NOTE: Kubernetes versions prior to 1.6 have limited or no support for role-based access controls (RBAC).
Helm will figure out where to install Tiller by reading your Kubernetes
configuration file (usually $HOME/.kube/config). This is the same file
that kubectl uses.
To find out which cluster Tiller would install to, you can run
kubectl config current-context or kubectl cluster-info.
$ kubectl config current-context
my-cluster
Understand your Security Context
As with all powerful tools, ensure you are installing it correctly for your scenario.
If you’re using Helm on a cluster that you completely control, like minikube or a cluster on a private network in which sharing is not a concern, the default installation – which applies no security configuration – is fine, and it’s definitely the easiest. To install Helm without additional security steps, install Helm and then initialize Helm.
However, if your cluster is exposed to a larger network or if you share your cluster with others – production clusters fall into this category – you must take extra steps to secure your installation to prevent careless or malicious actors from damaging the cluster or its data. To apply configurations that secure Helm for use in production environments and other multi-tenant scenarios, see Securing a Helm installation
If your cluster has Role-Based Access Control (RBAC) enabled, you may want to configure a service account and rules before proceeding.
Install Helm
Download a binary release of the Helm client. You can use tools like
homebrew, or look at the official releases page.
For more details, or for other options, see the installation guide.
Initialize Helm and Install Tiller
Once you have Helm ready, you can initialize the local CLI and also install Tiller into your Kubernetes cluster in one step:
$ helm init --history-max 200
TIP: Setting --history-max on helm init is recommended as configmaps and other objects in helm history can grow large in number if not purged by max limit. Without a max history set the history is kept indefinitely, leaving a large number of records for helm and tiller to maintain.
This will install Tiller into the Kubernetes cluster you saw with
kubectl config current-context.
TIP: Want to install into a different cluster? Use the
--kube-context flag.
TIP: When you want to upgrade Tiller, just run helm init --upgrade.
By default, when Tiller is installed, it does not have authentication enabled. To learn more about configuring strong TLS authentication for Tiller, consult the Tiller TLS guide.
Install an Example Chart
To install a chart, you can run the helm install command. Helm has
several ways to find and install a chart, but the easiest is to use one
of the official stable charts.
$ helm repo update # Make sure we get the latest list of charts
$ helm install stable/mysql
NAME: wintering-rodent
LAST DEPLOYED: Thu Oct 18 14:21:18 2018
NAMESPACE: default
STATUS: DEPLOYED
RESOURCES:
==> v1/Secret
NAME AGE
wintering-rodent-mysql 0s
==> v1/ConfigMap
wintering-rodent-mysql-test 0s
==> v1/PersistentVolumeClaim
wintering-rodent-mysql 0s
==> v1/Service
wintering-rodent-mysql 0s
==> v1beta1/Deployment
wintering-rodent-mysql 0s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
wintering-rodent-mysql-6986fd6fb-988x7 0/1 Pending 0 0s
NOTES:
MySQL can be accessed via port 3306 on the following DNS name from within your cluster:
wintering-rodent-mysql.default.svc.cluster.local
To get your root password run:
MYSQL_ROOT_PASSWORD=$(kubectl get secret --namespace default wintering-rodent-mysql -o jsonpath="{.data.mysql-root-password}" | base64 --decode; echo)
To connect to your database:
1. Run an Ubuntu pod that you can use as a client:
kubectl run -i --tty ubuntu --image=ubuntu:16.04 --restart=Never -- bash -il
2. Install the mysql client:
$ apt-get update && apt-get install mysql-client -y
3. Connect using the mysql cli, then provide your password:
$ mysql -h wintering-rodent-mysql -p
To connect to your database directly from outside the K8s cluster:
MYSQL_HOST=127.0.0.1
MYSQL_PORT=3306
# Execute the following command to route the connection:
kubectl port-forward svc/wintering-rodent-mysql 3306
mysql -h ${MYSQL_HOST} -P${MYSQL_PORT} -u root -p${MYSQL_ROOT_PASSWORD}
In the example above, the stable/mysql chart was released, and the name of
our new release is wintering-rodent. You get a simple idea of the
features of this MySQL chart by running helm inspect stable/mysql.
Whenever you install a chart, a new release is created. So one chart can be installed multiple times into the same cluster. And each can be independently managed and upgraded.
The helm install command is a very powerful command with many
capabilities. To learn more about it, check out the Using Helm
Guide
Learn About Releases
It’s easy to see what has been released using Helm:
$ helm ls
NAME REVISION UPDATED STATUS CHART APP VERSION NAMESPACE
wintering-rodent 1 Thu Oct 18 15:06:58 2018 DEPLOYED mysql-0.10.1 5.7.14 default
The helm list function will show you a list of all deployed releases.
Uninstall a Release
To uninstall a release, use the helm delete command:
$ helm delete wintering-rodent
release "wintering-rodent" deleted
This will uninstall wintering-rodent from Kubernetes, but you will
still be able to request information about that release:
$ helm status wintering-rodent
LAST DEPLOYED: Thu Oct 18 14:21:18 2018
NAMESPACE: default
STATUS: DELETED
NOTES:
MySQL can be accessed via port 3306 on the following DNS name from within your cluster:
wintering-rodent-mysql.default.svc.cluster.local
To get your root password run:
MYSQL_ROOT_PASSWORD=$(kubectl get secret --namespace default wintering-rodent-mysql -o jsonpath="{.data.mysql-root-password}" | base64 --decode; echo)
To connect to your database:
1. Run an Ubuntu pod that you can use as a client:
kubectl run -i --tty ubuntu --image=ubuntu:16.04 --restart=Never -- bash -il
2. Install the mysql client:
$ apt-get update && apt-get install mysql-client -y
3. Connect using the mysql cli, then provide your password:
$ mysql -h wintering-rodent-mysql -p
To connect to your database directly from outside the K8s cluster:
MYSQL_HOST=127.0.0.1
MYSQL_PORT=3306
# Execute the following command to route the connection:
kubectl port-forward svc/wintering-rodent-mysql 3306
mysql -h ${MYSQL_HOST} -P${MYSQL_PORT} -u root -p${MYSQL_ROOT_PASSWORD}
Because Helm tracks your releases even after you’ve deleted them, you
can audit a cluster’s history, and even undelete a release (with helm
rollback).
Reading the Help Text
To learn more about the available Helm commands, use helm help or type
a command followed by the -h flag:
$ helm get -h
Installing Helm
There are two parts to Helm: The Helm client (helm) and the Helm
server (Tiller). This guide shows how to install the client, and then
proceeds to show two ways to install the server.
IMPORTANT: If you are responsible for ensuring your cluster is a controlled environment, especially when resources are shared, it is strongly recommended installing Tiller using a secured configuration. For guidance, see Securing your Helm Installation.
Installing the Helm Client
The Helm client can be installed either from source, or from pre-built binary releases.
From the Binary Releases
Every release of Helm provides binary releases for a variety of OSes. These binary versions can be manually downloaded and installed.
- Download your desired version
- Unpack it (
tar -zxvf helm-v2.0.0-linux-amd64.tgz) - Find the
helmbinary in the unpacked directory, and move it to its desired destination (mv linux-amd64/helm /usr/local/bin/helm)
From there, you should be able to run the client: helm help.
From Snap (Linux)
The Snap package for Helm is maintained by Snapcrafters.
$ sudo snap install helm --classic
From Homebrew (macOS)
Members of the Kubernetes community have contributed a Helm formula build to Homebrew. This formula is generally up to date.
brew install kubernetes-helm
(Note: There is also a formula for emacs-helm, which is a different project.)
From Chocolatey or scoop (Windows)
Members of the Kubernetes community have contributed a Helm package build to Chocolatey. This package is generally up to date.
choco install kubernetes-helm
The binary can also be installed via scoop command-line installer.
scoop install helm
From Script
Helm now has an installer script that will automatically grab the latest version of the Helm client and install it locally.
You can fetch that script, and then execute it locally. It’s well documented so that you can read through it and understand what it is doing before you run it.
$ curl -LO https://git.io/get_helm.sh
$ chmod 700 get_helm.sh
$ ./get_helm.sh
Yes, you can curl -L https://git.io/get_helm.sh | bash that if you want to live on the edge.
From Canary Builds
“Canary” builds are versions of the Helm software that are built from the latest master branch. They are not official releases, and may not be stable. However, they offer the opportunity to test the cutting edge features.
Canary Helm binaries are stored in the Kubernetes Helm GCS bucket. Here are links to the common builds:
From Source (Linux, macOS)
Building Helm from source is slightly more work, but is the best way to go if you want to test the latest (pre-release) Helm version.
You must have a working Go environment with glide installed.
$ cd $GOPATH
$ mkdir -p src/k8s.io
$ cd src/k8s.io
$ git clone https://github.com/helm/helm.git
$ cd helm
$ make bootstrap build
The bootstrap target will attempt to install dependencies, rebuild the
vendor/ tree, and validate configuration.
The build target will compile helm and place it in bin/helm.
Tiller is also compiled, and is placed in bin/tiller.
Installing Tiller
Tiller, the server portion of Helm, typically runs inside of your Kubernetes cluster. But for development, it can also be run locally, and configured to talk to a remote Kubernetes cluster.
Special Note for RBAC Users
Most cloud providers enable a feature called Role-Based Access Control - RBAC for short. If your cloud provider enables this feature, you will need to create a service account for Tiller with the right roles and permissions to access resources.
Check the Kubernetes Distribution Guide to see if there’s any further points of interest on using Helm with your cloud provider. Also check out the guide on Tiller and Role-Based Access Control for more information on how to run Tiller in an RBAC-enabled Kubernetes cluster.
Easy In-Cluster Installation
The easiest way to install tiller into the cluster is simply to run
helm init. This will validate that helm’s local environment is set
up correctly (and set it up if necessary). Then it will connect to
whatever cluster kubectl connects to by default (kubectl config
view). Once it connects, it will install tiller into the
kube-system namespace.
After helm init, you should be able to run kubectl get pods --namespace
kube-system and see Tiller running.
You can explicitly tell helm init to…
- Install the canary build with the
--canary-imageflag - Install a particular image (version) with
--tiller-image - Install to a particular cluster with
--kube-context - Install into a particular namespace with
--tiller-namespace - Install Tiller with a Service Account with
--service-account(for RBAC enabled clusters) - Install Tiller without mounting a service account with
--automount-service-account false
Once Tiller is installed, running helm version should show you both
the client and server version. (If it shows only the client version,
helm cannot yet connect to the server. Use kubectl to see if any
tiller pods are running.)
Helm will look for Tiller in the kube-system namespace unless
--tiller-namespace or TILLER_NAMESPACE is set.
Installing Tiller Canary Builds
Canary images are built from the master branch. They may not be
stable, but they offer you the chance to test out the latest features.
The easiest way to install a canary image is to use helm init with the
--canary-image flag:
$ helm init --canary-image
This will use the most recently built container image. You can always
uninstall Tiller by deleting the Tiller deployment from the
kube-system namespace using kubectl.
Running Tiller Locally
For development, it is sometimes easier to work on Tiller locally, and configure it to connect to a remote Kubernetes cluster.
The process of building Tiller is explained above.
Once tiller has been built, simply start it:
$ bin/tiller
Tiller running on :44134
When Tiller is running locally, it will attempt to connect to the
Kubernetes cluster that is configured by kubectl. (Run kubectl config
view to see which cluster that is.)
You must tell helm to connect to this new local Tiller host instead of
connecting to the one in-cluster. There are two ways to do this. The
first is to specify the --host option on the command line. The second
is to set the $HELM_HOST environment variable.
$ export HELM_HOST=localhost:44134
$ helm version # Should connect to localhost.
Client: &version.Version{SemVer:"v2.0.0-alpha.4", GitCommit:"db...", GitTreeState:"dirty"}
Server: &version.Version{SemVer:"v2.0.0-alpha.4", GitCommit:"a5...", GitTreeState:"dirty"}
Importantly, even when running locally, Tiller will store release configuration in ConfigMaps inside of Kubernetes.
Upgrading Tiller
As of Helm 2.2.0, Tiller can be upgraded using helm init --upgrade.
For older versions of Helm, or for manual upgrades, you can use kubectl to modify
the Tiller image:
$ export TILLER_TAG=v2.0.0-beta.1 # Or whatever version you want
$ kubectl --namespace=kube-system set image deployments/tiller-deploy tiller=gcr.io/kubernetes-helm/tiller:$TILLER_TAG
deployment "tiller-deploy" image updated
Setting TILLER_TAG=canary will get the latest snapshot of master.
Deleting or Reinstalling Tiller
Because Tiller stores its data in Kubernetes ConfigMaps, you can safely
delete and re-install Tiller without worrying about losing any data. The
recommended way of deleting Tiller is with kubectl delete deployment
tiller-deploy --namespace kube-system, or more concisely helm reset.
Tiller can then be re-installed from the client with:
$ helm init
Advanced Usage
helm init provides additional flags for modifying Tiller’s deployment
manifest before it is installed.
Using --node-selectors
The --node-selectors flag allows us to specify the node labels required
for scheduling the Tiller pod.
The example below will create the specified label under the nodeSelector property.
helm init --node-selectors "beta.kubernetes.io/os"="linux"
The installed deployment manifest will contain our node selector label.
...
spec:
template:
spec:
nodeSelector:
beta.kubernetes.io/os: linux
...
Using --override
--override allows you to specify properties of Tiller’s
deployment manifest. Unlike the --set command used elsewhere in Helm,
helm init --override manipulates the specified properties of the final
manifest (there is no “values” file). Therefore you may specify any valid
value for any valid property in the deployment manifest.
Override annotation
In the example below we use --override to add the revision property and set
its value to 1.
helm init --override metadata.annotations."deployment\.kubernetes\.io/revision"="1"
Output:
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
annotations:
deployment.kubernetes.io/revision: "1"
...
Override affinity
In the example below we set properties for node affinity. Multiple
--override commands may be combined to modify different properties of the
same list item.
helm init --override "spec.template.spec.affinity.nodeAffinity.preferredDuringSchedulingIgnoredDuringExecution[0].weight"="1" --override "spec.template.spec.affinity.nodeAffinity.preferredDuringSchedulingIgnoredDuringExecution[0].preference.matchExpressions[0].key"="e2e-az-name"
The specified properties are combined into the “preferredDuringSchedulingIgnoredDuringExecution” property’s first list item.
...
spec:
strategy: {}
template:
...
spec:
affinity:
nodeAffinity:
preferredDuringSchedulingIgnoredDuringExecution:
- preference:
matchExpressions:
- key: e2e-az-name
operator: ""
weight: 1
...
Using --output
The --output flag allows us skip the installation of Tiller’s deployment
manifest and simply output the deployment manifest to stdout in either
JSON or YAML format. The output may then be modified with tools like jq
and installed manually with kubectl.
In the example below we execute helm init with the --output json flag.
helm init --output json
The Tiller installation is skipped and the manifest is output to stdout in JSON format.
"apiVersion": "extensions/v1beta1",
"kind": "Deployment",
"metadata": {
"creationTimestamp": null,
"labels": {
"app": "helm",
"name": "tiller"
},
"name": "tiller-deploy",
"namespace": "kube-system"
},
...
Storage backends
By default, tiller stores release information in ConfigMaps in the namespace
where it is running.
Secret storage backend
As of Helm 2.7.0, there is now a beta storage backend that
uses Secrets for storing release information. This was added for additional
security in protecting charts in conjunction with the release of Secret
encryption in Kubernetes.
To enable the secrets backend, you’ll need to init Tiller with the following options:
helm init --override 'spec.template.spec.containers[0].command'='{/tiller,--storage=secret}'
Currently, if you want to switch from the default backend to the secrets backend, you’ll have to do the migration for this on your own. When this backend graduates from beta, there will be a more official path of migration
SQL storage backend
As of Helm 2.14.0 there is now a beta SQL storage backend that stores release information in an SQL database (only postgres has been tested so far).
Using such a storage backend is particularly useful if your release information weighs more than 1MB (in which case, it can’t be stored in ConfigMaps/Secrets because of internal limits in Kubernetes’ underlying etcd key-value store).
To enable the SQL backend, you’ll need to deploy a SQL database and init Tiller with the following options:
helm init \
--override \
'spec.template.spec.containers[0].args'='{--storage=sql,--sql-dialect=postgres,--sql-connection-string=postgresql://tiller-postgres:5432/helm?user=helm&password=changeme}'
PRODUCTION NOTES: it’s recommended to change the username and password of the SQL database in production deployments. Enabling SSL is also a good idea. Last, but not least, perform regular backups/snapshots of your SQL database.
Currently, if you want to switch from the default backend to the SQL backend, you’ll have to do the migration for this on your own. When this backend graduates from beta, there will be a more official migration path.
Conclusion
In most cases, installation is as simple as getting a pre-built helm binary
and running helm init. This document covers additional cases for those
who want to do more sophisticated things with Helm.
Once you have the Helm Client and Tiller successfully installed, you can move on to using Helm to manage charts.
Kubernetes Distribution Guide
This document captures information about using Helm in specific Kubernetes environments.
We are trying to add more details to this document. Please contribute via Pull Requests if you can.
MiniKube
Helm is tested and known to work with minikube. It requires no additional configuration.
scripts/local-cluster and Hyperkube
Hyperkube configured via scripts/local-cluster.sh is known to work. For raw
Hyperkube you may need to do some manual configuration.
GKE
Google’s GKE hosted Kubernetes platform enables RBAC by default. You will need to create a service account for tiller, and use the –service-account flag when initializing the helm server.
See Tiller and role-based access control for more information.
Ubuntu with ‘kubeadm’
Kubernetes bootstrapped with kubeadm is known to work on the following Linux
distributions:
- Ubuntu 16.04
- Fedora release 25
Some versions of Helm (v2.0.0-beta2) require you to export KUBECONFIG=/etc/kubernetes/admin.conf
or create a ~/.kube/config.
Container Linux by CoreOS
Helm requires that kubelet have access to a copy of the socat program to proxy connections to the Tiller API. On Container Linux the Kubelet runs inside of a hyperkube container image that has socat. So, even though Container Linux doesn’t ship socat the container filesystem running kubelet does have socat. To learn more read the Kubelet Wrapper docs.
Openshift
Helm works straightforward on OpenShift Online, OpenShift Dedicated, OpenShift Container Platform (version >= 3.6) or OpenShift Origin (version >= 3.6). To learn more read this blog post.
Platform9
Helm Client and Helm Server (Tiller) are pre-installed with Platform9 Managed Kubernetes. Platform9 provides access to all official Helm charts through the App Catalog UI and native Kubernetes CLI. Additional repositories can be manually added. Further details are available in this Platform9 App Catalog article.
DC/OS
Helm (both client and server) has been tested and is working on Mesospheres DC/OS 1.11 Kubernetes platform, and requires no additional configuration.
Installation: Frequently Asked Questions
This section tracks some of the more frequently encountered issues with installing or getting started with Helm.
We’d love your help making this document better. To add, correct, or remove information, file an issue or send us a pull request.
Downloading
I want to know more about my downloading options.
Q: I can’t get to GitHub releases of the newest Helm. Where are they?
A: We no longer use GitHub releases. Binaries are now stored in a GCS public bucket.
Q: Why aren’t there Debian/Fedora/… native packages of Helm?
We’d love to provide these or point you toward a trusted provider. If you’re interested in helping, we’d love it. This is how the Homebrew formula was started.
Q: Why do you provide a curl ...|bash script?
A: There is a script in our repository (scripts/get) that can be executed as
a curl ..|bash script. The transfers are all protected by HTTPS, and the script
does some auditing of the packages it fetches. However, the script has all the
usual dangers of any shell script.
We provide it because it is useful, but we suggest that users carefully read the script first. What we’d really like, though, are better packaged releases of Helm.
Installing
I’m trying to install Helm/Tiller, but something is not right.
Q: How do I put the Helm client files somewhere other than ~/.helm?
Set the $HELM_HOME environment variable, and then run helm init:
export HELM_HOME=/some/path
helm init --client-only
Note that if you have existing repositories, you will need to re-add them
with helm repo add....
Q: How do I configure Helm, but not install Tiller?
A: By default, helm init will ensure that the local $HELM_HOME is configured,
and then install Tiller on your cluster. To locally configure, but not install
Tiller, use helm init --client-only.
Q: How do I manually install Tiller on the cluster?
A: Tiller is installed as a Kubernetes deployment. You can get the manifest
by running helm init --dry-run --debug, and then manually install it with
kubectl. It is suggested that you do not remove or change the labels on that
deployment, as they are sometimes used by supporting scripts and tools.
Q: Why do I get Error response from daemon: target is unknown during Tiller install?
A: Users have reported being unable to install Tiller on Kubernetes instances that are using Docker 1.13.0. The root cause of this was a bug in Docker that made that one version incompatible with images pushed to the Docker registry by earlier versions of Docker.
This issue was fixed shortly after the release, and is available in Docker 1.13.1-RC1 and later.
Getting Started
I successfully installed Helm/Tiller but I can’t use it.
Q: Trying to use Helm, I get the error “client transport was broken”
E1014 02:26:32.885226 16143 portforward.go:329] an error occurred forwarding 37008 -> 44134: error forwarding port 44134 to pod tiller-deploy-2117266891-e4lev_kube-system, uid : unable to do port forwarding: socat not found.
2016/10/14 02:26:32 transport: http2Client.notifyError got notified that the client transport was broken EOF.
Error: transport is closing
A: This is usually a good indication that Kubernetes is not set up to allow port forwarding.
Typically, the missing piece is socat. If you are running CoreOS, we have been
told that it may have been misconfigured on installation. The CoreOS team
recommends reading this:
Here are a few resolved issues that may help you get started:
Q: Trying to use Helm, I get the error “lookup XXXXX on 8.8.8.8:53: no such host”
Error: Error forwarding ports: error upgrading connection: dial tcp: lookup kube-4gb-lon1-02 on 8.8.8.8:53: no such host
A: We have seen this issue with Ubuntu and Kubeadm in multi-node clusters. The issue is that the nodes expect certain DNS records to be obtainable via global DNS. Until this is resolved upstream, you can work around the issue as follows. On each of the control plane nodes:
1) Add entries to /etc/hosts, mapping your hostnames to their public IPs
2) Install dnsmasq (e.g. apt install -y dnsmasq)
3) Remove the k8s api server container (kubelet will recreate it)
4) Then systemctl restart docker (or reboot the node) for it to pick up the /etc/resolv.conf changes
See this issue for more information: https://github.com/helm/helm/issues/1455
Q: On GKE (Google Container Engine) I get “No SSH tunnels currently open”
Error: Error forwarding ports: error upgrading connection: No SSH tunnels currently open. Were the targets able to accept an ssh-key for user "gke-[redacted]"?
Another variation of the error message is:
Unable to connect to the server: x509: certificate signed by unknown authority
A: The issue is that your local Kubernetes config file must have the correct credentials.
When you create a cluster on GKE, it will give you credentials, including SSL
certificates and certificate authorities. These need to be stored in a Kubernetes
config file (Default: ~/.kube/config so that kubectl and helm can access
them.
Q: When I run a Helm command, I get an error about the tunnel or proxy
A: Helm uses the Kubernetes proxy service to connect to the Tiller server.
If the command kubectl proxy does not work for you, neither will Helm.
Typically, the error is related to a missing socat service.
Q: Tiller crashes with a panic
When I run a command on Helm, Tiller crashes with an error like this:
Tiller is listening on :44134
Probes server is listening on :44135
Storage driver is ConfigMap
Cannot initialize Kubernetes connection: the server has asked for the client to provide credentials 2016-12-20 15:18:40.545739 I | storage.go:37: Getting release "bailing-chinchilla" (v1) from storage
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x8053d5]
goroutine 77 [running]:
panic(0x1abbfc0, 0xc42000a040)
/usr/local/go/src/runtime/panic.go:500 +0x1a1
k8s.io/helm/vendor/k8s.io/kubernetes/pkg/client/unversioned.(*ConfigMaps).Get(0xc4200c6200, 0xc420536100, 0x15, 0x1ca7431, 0x6, 0xc42016b6a0)
/home/ubuntu/.go_workspace/src/k8s.io/helm/vendor/k8s.io/kubernetes/pkg/client/unversioned/configmap.go:58 +0x75
k8s.io/helm/pkg/storage/driver.(*ConfigMaps).Get(0xc4201d6190, 0xc420536100, 0x15, 0xc420536100, 0x15, 0xc4205360c0)
/home/ubuntu/.go_workspace/src/k8s.io/helm/pkg/storage/driver/cfgmaps.go:69 +0x62
k8s.io/helm/pkg/storage.(*Storage).Get(0xc4201d61a0, 0xc4205360c0, 0x12, 0xc400000001, 0x12, 0x0, 0xc420200070)
/home/ubuntu/.go_workspace/src/k8s.io/helm/pkg/storage/storage.go:38 +0x160
k8s.io/helm/pkg/tiller.(*ReleaseServer).uniqName(0xc42002a000, 0x0, 0x0, 0xc42016b800, 0xd66a13, 0xc42055a040, 0xc420558050, 0xc420122001)
/home/ubuntu/.go_workspace/src/k8s.io/helm/pkg/tiller/release_server.go:577 +0xd7
k8s.io/helm/pkg/tiller.(*ReleaseServer).prepareRelease(0xc42002a000, 0xc42027c1e0, 0xc42002a001, 0xc42016bad0, 0xc42016ba08)
/home/ubuntu/.go_workspace/src/k8s.io/helm/pkg/tiller/release_server.go:630 +0x71
k8s.io/helm/pkg/tiller.(*ReleaseServer).InstallRelease(0xc42002a000, 0x7f284c434068, 0xc420250c00, 0xc42027c1e0, 0x0, 0x31a9, 0x31a9)
/home/ubuntu/.go_workspace/src/k8s.io/helm/pkg/tiller/release_server.go:604 +0x78
k8s.io/helm/pkg/proto/hapi/services._ReleaseService_InstallRelease_Handler(0x1c51f80, 0xc42002a000, 0x7f284c434068, 0xc420250c00, 0xc42027c190, 0x0, 0x0, 0x0, 0x0, 0x0)
/home/ubuntu/.go_workspace/src/k8s.io/helm/pkg/proto/hapi/services/tiller.pb.go:747 +0x27d
k8s.io/helm/vendor/google.golang.org/grpc.(*Server).processUnaryRPC(0xc4202f3ea0, 0x28610a0, 0xc420078000, 0xc420264690, 0xc420166150, 0x288cbe8, 0xc420250bd0, 0x0, 0x0)
/home/ubuntu/.go_workspace/src/k8s.io/helm/vendor/google.golang.org/grpc/server.go:608 +0xc50
k8s.io/helm/vendor/google.golang.org/grpc.(*Server).handleStream(0xc4202f3ea0, 0x28610a0, 0xc420078000, 0xc420264690, 0xc420250bd0)
/home/ubuntu/.go_workspace/src/k8s.io/helm/vendor/google.golang.org/grpc/server.go:766 +0x6b0
k8s.io/helm/vendor/google.golang.org/grpc.(*Server).serveStreams.func1.1(0xc420124710, 0xc4202f3ea0, 0x28610a0, 0xc420078000, 0xc420264690)
/home/ubuntu/.go_workspace/src/k8s.io/helm/vendor/google.golang.org/grpc/server.go:419 +0xab
created by k8s.io/helm/vendor/google.golang.org/grpc.(*Server).serveStreams.func1
/home/ubuntu/.go_workspace/src/k8s.io/helm/vendor/google.golang.org/grpc/server.go:420 +0xa3
A: Check your security settings for Kubernetes.
A panic in Tiller is almost always the result of a failure to negotiate with the Kubernetes API server (at which point Tiller can no longer do anything useful, so it panics and exits).
Often, this is a result of authentication failing because the Pod in which Tiller is running does not have the right token.
To fix this, you will need to change your Kubernetes configuration. Make sure
that --service-account-private-key-file from controller-manager and
--service-account-key-file from apiserver point to the same x509 RSA key.
Upgrading
My Helm used to work, then I upgrade. Now it is broken.
Q: After upgrade, I get the error “Client version is incompatible”. What’s wrong?
Tiller and Helm have to negotiate a common version to make sure that they can safely communicate without breaking API assumptions. That error means that the version difference is too great to safely continue. Typically, you need to upgrade Tiller manually for this.
The Installation Guide has definitive information about safely upgrading Helm and Tiller.
The rules for version numbers are as follows:
- Pre-release versions are incompatible with everything else.
Alpha.1is incompatible withAlpha.2. - Patch revisions are compatible: 1.2.3 is compatible with 1.2.4
- Minor revisions are not compatible: 1.2.0 is not compatible with 1.3.0, though we may relax this constraint in the future.
- Major revisions are not compatible: 1.0.0 is not compatible with 2.0.0.
Uninstalling
I am trying to remove stuff.
Q: When I delete the Tiller deployment, how come all the releases are still there?
Releases are stored in ConfigMaps inside of the kube-system namespace. You will
have to manually delete them to get rid of the record, or use helm delete --purge.
Q: I want to delete my local Helm. Where are all its files?
Along with the helm binary, Helm stores some files in $HELM_HOME, which is
located by default in ~/.helm.
Using Helm
This guide explains the basics of using Helm (and Tiller) to manage
packages on your Kubernetes cluster. It assumes that you have already
installed the Helm client and the Tiller server (typically by helm
init).
If you are simply interested in running a few quick commands, you may wish to begin with the Quickstart Guide. This chapter covers the particulars of Helm commands, and explains how to use Helm.
Three Big Concepts
A Chart is a Helm package. It contains all of the resource definitions necessary to run an application, tool, or service inside of a Kubernetes cluster. Think of it like the Kubernetes equivalent of a Homebrew formula, an Apt dpkg, or a Yum RPM file.
A Repository is the place where charts can be collected and shared. It’s like Perl’s CPAN archive or the Fedora Package Database, but for Kubernetes packages.
A Release is an instance of a chart running in a Kubernetes cluster. One chart can often be installed many times into the same cluster. And each time it is installed, a new release is created. Consider a MySQL chart. If you want two databases running in your cluster, you can install that chart twice. Each one will have its own release, which will in turn have its own release name.
With these concepts in mind, we can now explain Helm like this:
Helm installs charts into Kubernetes, creating a new release for each installation. And to find new charts, you can search Helm chart repositories.
‘helm search’: Finding Charts
When you first install Helm, it is preconfigured to talk to the official
Kubernetes charts repository. This repository contains a number of
carefully curated and maintained charts. This chart repository is named
stable by default.
You can see which charts are available by running helm search:
$ helm search
NAME VERSION DESCRIPTION
stable/drupal 0.3.2 One of the most versatile open source content m...
stable/jenkins 0.1.0 A Jenkins Helm chart for Kubernetes.
stable/mariadb 0.5.1 Chart for MariaDB
stable/mysql 0.1.0 Chart for MySQL
...
With no filter, helm search shows you all of the available charts. You
can narrow down your results by searching with a filter:
$ helm search mysql
NAME VERSION DESCRIPTION
stable/mysql 0.1.0 Chart for MySQL
stable/mariadb 0.5.1 Chart for MariaDB
Now you will only see the results that match your filter.
Why is
mariadb in the list? Because its package description relates it to
MySQL. We can use helm inspect chart to see this:
$ helm inspect stable/mariadb
Fetched stable/mariadb to mariadb-0.5.1.tgz
description: Chart for MariaDB
engine: gotpl
home: https://mariadb.org
keywords:
- mariadb
- mysql
- database
- sql
...
Search is a good way to find available packages. Once you have found a
package you want to install, you can use helm install to install it.
‘helm install’: Installing a Package
To install a new package, use the helm install command. At its
simplest, it takes only one argument: The name of the chart.
$ helm install stable/mariadb
Fetched stable/mariadb-0.3.0 to /Users/mattbutcher/Code/Go/src/k8s.io/helm/mariadb-0.3.0.tgz
NAME: happy-panda
LAST DEPLOYED: Wed Sep 28 12:32:28 2016
NAMESPACE: default
STATUS: DEPLOYED
Resources:
==> extensions/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
happy-panda-mariadb 1 0 0 0 1s
==> v1/Secret
NAME TYPE DATA AGE
happy-panda-mariadb Opaque 2 1s
==> v1/Service
NAME CLUSTER-IP EXTERNAL-IP PORT(S) AGE
happy-panda-mariadb 10.0.0.70 <none> 3306/TCP 1s
Notes:
MariaDB can be accessed via port 3306 on the following DNS name from within your cluster:
happy-panda-mariadb.default.svc.cluster.local
To connect to your database run the following command:
kubectl run happy-panda-mariadb-client --rm --tty -i --image bitnami/mariadb --command -- mysql -h happy-panda-mariadb
Now the mariadb chart is installed. Note that installing a chart
creates a new release object. The release above is named
happy-panda. (If you want to use your own release name, simply use the
--name flag on helm install.)
During installation, the helm client will print useful information
about which resources were created, what the state of the release is,
and also whether there are additional configuration steps you can or
should take.
Helm does not wait until all of the resources are running before it exits. Many charts require Docker images that are over 600M in size, and may take a long time to install into the cluster.
To keep track of a release’s state, or to re-read configuration
information, you can use helm status:
$ helm status happy-panda
Last Deployed: Wed Sep 28 12:32:28 2016
Namespace: default
Status: DEPLOYED
Resources:
==> v1/Service
NAME CLUSTER-IP EXTERNAL-IP PORT(S) AGE
happy-panda-mariadb 10.0.0.70 <none> 3306/TCP 4m
==> extensions/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
happy-panda-mariadb 1 1 1 1 4m
==> v1/Secret
NAME TYPE DATA AGE
happy-panda-mariadb Opaque 2 4m
Notes:
MariaDB can be accessed via port 3306 on the following DNS name from within your cluster:
happy-panda-mariadb.default.svc.cluster.local
To connect to your database run the following command:
kubectl run happy-panda-mariadb-client --rm --tty -i --image bitnami/mariadb --command -- mysql -h happy-panda-mariadb
The above shows the current state of your release.
Customizing the Chart Before Installing
Installing the way we have here will only use the default configuration options for this chart. Many times, you will want to customize the chart to use your preferred configuration.
To see what options are configurable on a chart, use helm inspect
values:
helm inspect values stable/mariadb
Fetched stable/mariadb-0.3.0.tgz to /Users/mattbutcher/Code/Go/src/k8s.io/helm/mariadb-0.3.0.tgz
## Bitnami MariaDB image version
## ref: https://hub.docker.com/r/bitnami/mariadb/tags/
##
## Default: none
imageTag: 10.1.14-r3
## Specify a imagePullPolicy
## Default to 'Always' if imageTag is 'latest', else set to 'IfNotPresent'
## ref: https://kubernetes.io/docs/user-guide/images/#pre-pulling-images
##
# imagePullPolicy:
## Specify password for root user
## ref: https://github.com/bitnami/bitnami-docker-mariadb/blob/master/README.md#setting-the-root-password-on-first-run
##
# mariadbRootPassword:
## Create a database user
## ref: https://github.com/bitnami/bitnami-docker-mariadb/blob/master/README.md#creating-a-database-user-on-first-run
##
# mariadbUser:
# mariadbPassword:
## Create a database
## ref: https://github.com/bitnami/bitnami-docker-mariadb/blob/master/README.md#creating-a-database-on-first-run
##
# mariadbDatabase:
You can then override any of these settings in a YAML formatted file, and then pass that file during installation.
$ cat << EOF > config.yaml
mariadbUser: user0
mariadbDatabase: user0db
EOF
$ helm install -f config.yaml stable/mariadb
The above will create a default MariaDB user with the name user0, and
grant this user access to a newly created user0db database, but will
accept all the rest of the defaults for that chart.
There are two ways to pass configuration data during install:
--values(or-f): Specify a YAML file with overrides. This can be specified multiple times and the rightmost file will take precedence--set(and its variants--set-stringand--set-file): Specify overrides on the command line.
If both are used, --set values are merged into --values with higher precedence.
Overrides specified with --set are persisted in a configmap. Values that have been
--set can be viewed for a given release with helm get values <release-name>.
Values that have been --set can be cleared by running helm upgrade with --reset-values
specified.
The Format and Limitations of --set
The --set option takes zero or more name/value pairs. At its simplest, it is
used like this: --set name=value. The YAML equivalent of that is:
name: value
Multiple values are separated by , characters. So --set a=b,c=d becomes:
a: b
c: d
More complex expressions are supported. For example, --set outer.inner=value is
translated into this:
outer:
inner: value
Lists can be expressed by enclosing values in { and }. For example,
--set name={a, b, c} translates to:
name:
- a
- b
- c
As of Helm 2.5.0, it is possible to access list items using an array index syntax.
For example, --set servers[0].port=80 becomes:
servers:
- port: 80
Multiple values can be set this way. The line --set servers[0].port=80,servers[0].host=example becomes:
servers:
- port: 80
host: example
Sometimes you need to use special characters in your --set lines. You can use
a backslash to escape the characters; --set name="value1\,value2" will become:
name: "value1,value2"
Similarly, you can escape dot sequences as well, which may come in handy when charts use the
toYaml function to parse annotations, labels and node selectors. The syntax for
--set nodeSelector."kubernetes\.io/role"=master becomes:
nodeSelector:
kubernetes.io/role: master
Deeply nested data structures can be difficult to express using --set. Chart
designers are encouraged to consider the --set usage when designing the format
of a values.yaml file.
Helm will cast certain values specified with --set to integers.
For example, --set foo=true results Helm to cast true into an int64 value.
In case you want a string, use a --set’s variant named --set-string. --set-string foo=true results in a string value of "true".
--set-file key=filepath is another variant of --set.
It reads the file and use its content as a value.
An example use case of it is to inject a multi-line text into values without dealing with indentation in YAML.
Say you want to create a brigade project with certain value containing 5 lines JavaScript code, you might write a values.yaml like:
defaultScript: |
const { events, Job } = require("brigadier")
function run(e, project) {
console.log("hello default script")
}
events.on("run", run)
Being embedded in a YAML, this makes it harder for you to use IDE features and testing framework and so on that supports writing code.
Instead, you can use --set-file defaultScript=brigade.js with brigade.js containing:
const { events, Job } = require("brigadier")
function run(e, project) {
console.log("hello default script")
}
events.on("run", run)
More Installation Methods
The helm install command can install from several sources:
- A chart repository (as we’ve seen above)
- A local chart archive (
helm install foo-0.1.1.tgz) - An unpacked chart directory (
helm install path/to/foo) - A full URL (
helm install https://example.com/charts/foo-1.2.3.tgz)
‘helm upgrade’ and ‘helm rollback’: Upgrading a Release, and Recovering on Failure
When a new version of a chart is released, or when you want to change
the configuration of your release, you can use the helm upgrade
command.
An upgrade takes an existing release and upgrades it according to the information you provide. Because Kubernetes charts can be large and complex, Helm tries to perform the least invasive upgrade. It will only update things that have changed since the last release.
$ helm upgrade -f panda.yaml happy-panda stable/mariadb
Fetched stable/mariadb-0.3.0.tgz to /Users/mattbutcher/Code/Go/src/k8s.io/helm/mariadb-0.3.0.tgz
happy-panda has been upgraded.
Last Deployed: Wed Sep 28 12:47:54 2016
Namespace: default
Status: DEPLOYED
...
In the above case, the happy-panda release is upgraded with the same
chart, but with a new YAML file:
mariadbUser: user1
We can use helm get values to see whether that new setting took
effect.
$ helm get values happy-panda
mariadbUser: user1
The helm get command is a useful tool for looking at a release in the
cluster. And as we can see above, it shows that our new values from
panda.yaml were deployed to the cluster.
Now, if something does not go as planned during a release, it is easy to
roll back to a previous release using helm rollback [RELEASE] [REVISION].
$ helm rollback happy-panda 1
The above rolls back our happy-panda to its very first release version.
A release version is an incremental revision. Every time an install,
upgrade, or rollback happens, the revision number is incremented by 1.
The first revision number is always 1. And we can use helm history [RELEASE]
to see revision numbers for a certain release.
Helpful Options for Install/Upgrade/Rollback
There are several other helpful options you can specify for customizing the
behavior of Helm during an install/upgrade/rollback. Please note that this
is not a full list of cli flags. To see a description of all flags, just run
helm <command> --help.
--timeout: A value in seconds to wait for Kubernetes commands to complete This defaults to 300 (5 minutes)--wait: Waits until all Pods are in a ready state, PVCs are bound, Deployments have minimum (DesiredminusmaxUnavailable) Pods in ready state and Services have an IP address (and Ingress if aLoadBalancer) before marking the release as successful. It will wait for as long as the--timeoutvalue. If timeout is reached, the release will be marked asFAILED. Note: In scenario where Deployment hasreplicasset to 1 andmaxUnavailableis not set to 0 as part of rolling update strategy,--waitwill return as ready as it has satisfied the minimum Pod in ready condition.--no-hooks: This skips running hooks for the command--recreate-pods(only available forupgradeandrollback): This flag will cause all pods to be recreated (with the exception of pods belonging to deployments)
‘helm delete’: Deleting a Release
When it is time to uninstall or delete a release from the cluster, use
the helm delete command:
$ helm delete happy-panda
This will remove the release from the cluster. You can see all of your
currently deployed releases with the helm list command:
$ helm list
NAME VERSION UPDATED STATUS CHART
inky-cat 1 Wed Sep 28 12:59:46 2016 DEPLOYED alpine-0.1.0
From the output above, we can see that the happy-panda release was
deleted.
However, Helm always keeps records of what releases happened. Need to
see the deleted releases? helm list --deleted shows those, and helm
list --all shows all of the releases (deleted and currently deployed,
as well as releases that failed):
⇒ helm list --all
NAME VERSION UPDATED STATUS CHART
happy-panda 2 Wed Sep 28 12:47:54 2016 DELETED mariadb-0.3.0
inky-cat 1 Wed Sep 28 12:59:46 2016 DEPLOYED alpine-0.1.0
kindred-angelf 2 Tue Sep 27 16:16:10 2016 DELETED alpine-0.1.0
Because Helm keeps records of deleted releases, a release name cannot be
re-used. (If you really need to re-use a release name, you can use the
--replace flag, but it will simply re-use the existing release and
replace its resources.)
Note that because releases are preserved in this way, you can rollback a deleted resource, and have it re-activate.
‘helm repo’: Working with Repositories
So far, we’ve been installing charts only from the stable repository.
But you can configure helm to use other repositories. Helm provides
several repository tools under the helm repo command.
You can see which repositories are configured using helm repo list:
$ helm repo list
NAME URL
stable https://kubernetes-charts.storage.googleapis.com
local http://localhost:8879/charts
mumoshu https://mumoshu.github.io/charts
And new repositories can be added with helm repo add:
$ helm repo add dev https://example.com/dev-charts
Because chart repositories change frequently, at any point you can make
sure your Helm client is up to date by running helm repo update.
Creating Your Own Charts
The Chart Development Guide explains how to develop your own
charts. But you can get started quickly by using the helm create
command:
$ helm create deis-workflow
Creating deis-workflow
Now there is a chart in ./deis-workflow. You can edit it and create
your own templates.
As you edit your chart, you can validate that it is well-formatted by
running helm lint.
When it’s time to package the chart up for distribution, you can run the
helm package command:
$ helm package deis-workflow
deis-workflow-0.1.0.tgz
And that chart can now easily be installed by helm install:
$ helm install ./deis-workflow-0.1.0.tgz
...
Charts that are archived can be loaded into chart repositories. See the documentation for your chart repository server to learn how to upload.
Note: The stable repository is managed on the Helm Charts
GitHub repository. That project
accepts chart source code, and (after audit) packages those for you.
Tiller, Namespaces and RBAC
In some cases you may wish to scope Tiller or deploy multiple Tillers to a single cluster. Here are some best practices when operating in those circumstances.
- Tiller can be installed into any namespace. By default, it is installed into kube-system. You can run multiple Tillers provided they each run in their own namespace.
- Limiting Tiller to only be able to install into specific namespaces and/or resource types is controlled by Kubernetes RBAC roles and rolebindings. You can add a service account to Tiller when configuring Helm via
helm init --service-account <NAME>. You can find more information about that here. - Release names are unique PER TILLER INSTANCE.
- Charts should only contain resources that exist in a single namespace.
- It is not recommended to have multiple Tillers configured to manage resources in the same namespace.
Conclusion
This chapter has covered the basic usage patterns of the helm client,
including searching, installation, upgrading, and deleting. It has also
covered useful utility commands like helm status, helm get, and
helm repo.
For more information on these commands, take a look at Helm’s built-in
help: helm help.
In the next chapter, we look at the process of developing charts.
The Helm Plugins Guide
Helm 2.1.0 introduced the concept of a client-side Helm plugin. A plugin is a
tool that can be accessed through the helm CLI, but which is not part of the
built-in Helm codebase.
Existing plugins can be found on related section or by searching Github.
This guide explains how to use and create plugins.
An Overview
Helm plugins are add-on tools that integrate seamlessly with Helm. They provide a way to extend the core feature set of Helm, but without requiring every new feature to be written in Go and added to the core tool.
Helm plugins have the following features:
- They can be added and removed from a Helm installation without impacting the core Helm tool.
- They can be written in any programming language.
- They integrate with Helm, and will show up in
helm helpand other places.
Helm plugins live in $(helm home)/plugins.
The Helm plugin model is partially modeled on Git’s plugin model. To that end,
you may sometimes hear helm referred to as the porcelain layer, with
plugins being the plumbing. This is a shorthand way of suggesting that
Helm provides the user experience and top level processing logic, while the
plugins do the “detail work” of performing a desired action.
Installing a Plugin
Plugins are installed using the $ helm plugin install <path|url> command. You can pass in a path to a plugin on your local file system or a url of a remote VCS repo. The helm plugin install command clones or copies the plugin at the path/url given into $ (helm home)/plugins
$ helm plugin install https://github.com/technosophos/helm-template
If you have a plugin tar distribution, simply untar the plugin into the
$(helm home)/plugins directory.
You can also install tarball plugins directly from url by issuing helm plugin install http://domain/path/to/plugin.tar.gz
Building Plugins
In many ways, a plugin is similar to a chart. Each plugin has a top-level
directory, and then a plugin.yaml file.
$(helm home)/plugins/
|- keybase/
|
|- plugin.yaml
|- keybase.sh
In the example above, the keybase plugin is contained inside of a directory
named keybase. It has two files: plugin.yaml (required) and an executable
script, keybase.sh (optional).
The core of a plugin is a simple YAML file named plugin.yaml.
Here is a plugin YAML for a plugin that adds support for Keybase operations:
name: "keybase"
version: "0.1.0"
usage: "Integrate Keybase.io tools with Helm"
description: |-
This plugin provides Keybase services to Helm.
ignoreFlags: false
useTunnel: false
command: "$HELM_PLUGIN_DIR/keybase.sh"
The name is the name of the plugin. When Helm executes it plugin, this is the
name it will use (e.g. helm NAME will invoke this plugin).
name should match the directory name. In our example above, that means the
plugin with name: keybase should be contained in a directory named keybase.
Restrictions on name:
namecannot duplicate one of the existinghelmtop-level commands.namemust be restricted to the characters ASCII a-z, A-Z, 0-9,_and-.
version is the SemVer 2 version of the plugin.
usage and description are both used to generate the help text of a command.
The ignoreFlags switch tells Helm to not pass flags to the plugin. So if a
plugin is called with helm myplugin --foo and ignoreFlags: true, then --foo
is silently discarded.
The useTunnel switch indicates that the plugin needs a tunnel to Tiller. This
should be set to true anytime a plugin talks to Tiller. It will cause Helm
to open a tunnel, and then set $TILLER_HOST to the right local address for that
tunnel. But don’t worry: if Helm detects that a tunnel is not necessary because
Tiller is running locally, it will not create the tunnel.
Finally, and most importantly, command is the command that this plugin will
execute when it is called. Environment variables are interpolated before the plugin
is executed. The pattern above illustrates the preferred way to indicate where
the plugin program lives.
There are some strategies for working with plugin commands:
- If a plugin includes an executable, the executable for a
command:should be packaged in the plugin directory. - The
command:line will have any environment variables expanded before execution.$HELM_PLUGIN_DIRwill point to the plugin directory. - The command itself is not executed in a shell. So you can’t oneline a shell script.
- Helm injects lots of configuration into environment variables. Take a look at the environment to see what information is available.
- Helm makes no assumptions about the language of the plugin. You can write it in whatever you prefer.
- Commands are responsible for implementing specific help text for
-hand--help. Helm will useusageanddescriptionforhelm helpandhelm help myplugin, but will not handlehelm myplugin --help.
Downloader Plugins
By default, Helm is able to fetch Charts using HTTP/S. As of Helm 2.4.0, plugins can have a special capability to download Charts from arbitrary sources.
Plugins shall declare this special capability in the plugin.yaml file (top level):
downloaders:
- command: "bin/mydownloader"
protocols:
- "myprotocol"
- "myprotocols"
If such plugin is installed, Helm can interact with the repository using the specified
protocol scheme by invoking the command. The special repository shall be added
similarly to the regular ones: helm repo add favorite myprotocol://example.com/
The rules for the special repos are the same to the regular ones: Helm must be able
to download the index.yaml file in order to discover and cache the list of
available Charts.
The defined command will be invoked with the following scheme:
command certFile keyFile caFile full-URL. The SSL credentials are coming from the
repo definition, stored in $HELM_HOME/repository/repositories.yaml. Downloader
plugin is expected to dump the raw content to stdout and report errors on stderr.
Environment Variables
When Helm executes a plugin, it passes the outer environment to the plugin, and also injects some additional environment variables.
Variables like KUBECONFIG are set for the plugin if they are set in the
outer environment.
The following variables are guaranteed to be set:
HELM_PLUGIN: The path to the plugins directoryHELM_PLUGIN_NAME: The name of the plugin, as invoked byhelm. Sohelm myplugwill have the short namemyplug.HELM_PLUGIN_DIR: The directory that contains the plugin.HELM_BIN: The path to thehelmcommand (as executed by the user).HELM_HOME: The path to the Helm home.HELM_PATH_*: Paths to important Helm files and directories are stored in environment variables prefixed byHELM_PATH.TILLER_HOST: Thedomain:portto Tiller. If a tunnel is created, this will point to the local endpoint for the tunnel. Otherwise, it will point to$HELM_HOST,--host, or the default host (according to Helm’s rules of precedence).
While HELM_HOST may be set, there is no guarantee that it will point to the
correct Tiller instance. This is done to allow plugin developer to access
HELM_HOST in its raw state when the plugin itself needs to manually configure
a connection.
A Note on useTunnel
If a plugin specifies useTunnel: true, Helm will do the following (in order):
- Parse global flags and the environment
- Create the tunnel
- Set
TILLER_HOST - Execute the plugin
- Close the tunnel
The tunnel is removed as soon as the command returns. So, for example, a
command cannot background a process and assume that process will be able
to use the tunnel.
A Note on Flag Parsing
When executing a plugin, Helm will parse global flags for its own use. Some of these flags are not passed on to the plugin.
--debug: If this is specified,$HELM_DEBUGis set to1--home: This is converted to$HELM_HOME--host: This is converted to$HELM_HOST--kube-context: This is simply dropped. If your plugin usesuseTunnel, this is used to set up the tunnel for you.
Plugins should display help text and then exit for -h and --help. In all
other cases, plugins may use flags as appropriate.
Role-based Access Control
In Kubernetes, granting a role to an application-specific service account is a best practice to ensure that your application is operating in the scope that you have specified. Read more about service account permissions in the official Kubernetes docs.
Bitnami also has a fantastic guide for configuring RBAC in your cluster that takes you through RBAC basics.
This guide is for users who want to restrict Tiller’s capabilities to install resources to certain namespaces, or to grant a Helm client running access to a Tiller instance.
Tiller and Role-based Access Control
You can add a service account to Tiller using the --service-account <NAME> flag while you’re configuring Helm. As a prerequisite, you’ll have to create a role binding which specifies a role and a service account name that have been set up in advance.
Once you have satisfied the pre-requisite and have a service account with the correct permissions, you’ll run a command like this: helm init --service-account <NAME>
Example: Service account with cluster-admin role
In rbac-config.yaml:
apiVersion: v1
kind: ServiceAccount
metadata:
name: tiller
namespace: kube-system
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
name: tiller
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: cluster-admin
subjects:
- kind: ServiceAccount
name: tiller
namespace: kube-system
Note: The cluster-admin role is created by default in a Kubernetes cluster, so you don’t have to define it explicitly.
$ kubectl create -f rbac-config.yaml
serviceaccount "tiller" created
clusterrolebinding "tiller" created
$ helm init --service-account tiller --history-max 200
Example: Deploy Tiller in a namespace, restricted to deploying resources only in that namespace
In the example above, we gave Tiller admin access to the entire cluster. You are not at all required to give Tiller cluster-admin access for it to work. Instead of specifying a ClusterRole or a ClusterRoleBinding, you can specify a Role and RoleBinding to limit Tiller’s scope to a particular namespace.
$ kubectl create namespace tiller-world
namespace "tiller-world" created
$ kubectl create serviceaccount tiller --namespace tiller-world
serviceaccount "tiller" created
Define a Role that allows Tiller to manage all resources in tiller-world like in role-tiller.yaml:
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: tiller-manager
namespace: tiller-world
rules:
- apiGroups: ["", "batch", "extensions", "apps"]
resources: ["*"]
verbs: ["*"]
$ kubectl create -f role-tiller.yaml
role "tiller-manager" created
In rolebinding-tiller.yaml,
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: tiller-binding
namespace: tiller-world
subjects:
- kind: ServiceAccount
name: tiller
namespace: tiller-world
roleRef:
kind: Role
name: tiller-manager
apiGroup: rbac.authorization.k8s.io
$ kubectl create -f rolebinding-tiller.yaml
rolebinding "tiller-binding" created
Afterwards you can run helm init to install Tiller in the tiller-world namespace.
$ helm init --service-account tiller --tiller-namespace tiller-world
$HELM_HOME has been configured at /Users/awesome-user/.helm.
Tiller (the Helm server side component) has been installed into your Kubernetes Cluster.
$ helm install nginx --tiller-namespace tiller-world --namespace tiller-world
NAME: wayfaring-yak
LAST DEPLOYED: Mon Aug 7 16:00:16 2017
NAMESPACE: tiller-world
STATUS: DEPLOYED
RESOURCES:
==> v1/Pod
NAME READY STATUS RESTARTS AGE
wayfaring-yak-alpine 0/1 ContainerCreating 0 0s
Example: Deploy Tiller in a namespace, restricted to deploying resources in another namespace
In the example above, we gave Tiller admin access to the namespace it was deployed inside. Now, let’s limit Tiller’s scope to deploy resources in a different namespace!
For example, let’s install Tiller in the namespace myorg-system and allow Tiller to deploy resources in the namespace myorg-users.
$ kubectl create namespace myorg-system
namespace "myorg-system" created
$ kubectl create serviceaccount tiller --namespace myorg-system
serviceaccount "tiller" created
Define a Role that allows Tiller to manage all resources in myorg-users like in role-tiller.yaml:
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: tiller-manager
namespace: myorg-users
rules:
- apiGroups: ["", "batch", "extensions", "apps"]
resources: ["*"]
verbs: ["*"]
$ kubectl create -f role-tiller.yaml
role "tiller-manager" created
Bind the service account to that role. In rolebinding-tiller.yaml,
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: tiller-binding
namespace: myorg-users
subjects:
- kind: ServiceAccount
name: tiller
namespace: myorg-system
roleRef:
kind: Role
name: tiller-manager
apiGroup: rbac.authorization.k8s.io
$ kubectl create -f rolebinding-tiller.yaml
rolebinding "tiller-binding" created
We’ll also need to grant Tiller access to read configmaps in myorg-system so it can store release information. In role-tiller-myorg-system.yaml:
kind: Role
apiVersion: rbac.authorization.k8s.io/v1
metadata:
namespace: myorg-system
name: tiller-manager
rules:
- apiGroups: ["", "extensions", "apps"]
resources: ["configmaps"]
verbs: ["*"]
$ kubectl create -f role-tiller-myorg-system.yaml
role "tiller-manager" created
And the respective role binding. In rolebinding-tiller-myorg-system.yaml:
kind: RoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
name: tiller-binding
namespace: myorg-system
subjects:
- kind: ServiceAccount
name: tiller
namespace: myorg-system
roleRef:
kind: Role
name: tiller-manager
apiGroup: rbac.authorization.k8s.io
$ kubectl create -f rolebinding-tiller-myorg-system.yaml
rolebinding "tiller-binding" created
Helm and Role-based Access Control
When running a Helm client in a pod, in order for the Helm client to talk to a Tiller instance, it will need certain privileges to be granted. Specifically, the Helm client will need to be able to create pods, forward ports and be able to list pods in the namespace where Tiller is running (so it can find Tiller).
Example: Deploy Helm in a namespace, talking to Tiller in another namespace
In this example, we will assume Tiller is running in a namespace called tiller-world and that the Helm client is running in a namespace called helm-world. By default, Tiller is running in the kube-system namespace.
In helm-user.yaml:
apiVersion: v1
kind: ServiceAccount
metadata:
name: helm
namespace: helm-world
---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
name: tiller-user
namespace: tiller-world
rules:
- apiGroups:
- ""
resources:
- pods/portforward
verbs:
- create
- apiGroups:
- ""
resources:
- pods
verbs:
- list
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
name: tiller-user-binding
namespace: tiller-world
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: Role
name: tiller-user
subjects:
- kind: ServiceAccount
name: helm
namespace: helm-world
$ kubectl create -f helm-user.yaml
serviceaccount "helm" created
role "tiller-user" created
rolebinding "tiller-user-binding" created
Using SSL Between Helm and Tiller
This document explains how to create strong SSL/TLS connections between Helm and Tiller. The emphasis here is on creating an internal CA, and using both the cryptographic and identity functions of SSL.
Support for TLS-based auth was introduced in Helm 2.3.0
Configuring SSL is considered an advanced topic, and knowledge of Helm and Tiller is assumed.
Overview
The Tiller authentication model uses client-side SSL certificates. Tiller itself verifies these certificates using a certificate authority. Likewise, the client also verifies Tiller’s identity by certificate authority.
There are numerous possible configurations for setting up certificates and authorities, but the method we cover here will work for most situations.
As of Helm 2.7.2, Tiller requires that the client certificate be validated by its CA. In prior versions, Tiller used a weaker validation strategy that allowed self-signed certificates.
In this guide, we will show how to:
- Create a private CA that is used to issue certificates for Tiller clients and servers.
- Create a certificate for Tiller
- Create a certificate for the Helm client
- Create a Tiller instance that uses the certificate
- Configure the Helm client to use the CA and client-side certificate
By the end of this guide, you should have a Tiller instance running that will only accept connections from clients who can be authenticated by SSL certificate.
Generating Certificate Authorities and Certificates
One way to generate SSL CAs is via the openssl command line tool. There are many
guides and best practices documents available online. This explanation is focused
on getting ready within a small amount of time. For production configurations,
we urge readers to read the official documentation and
consult other resources.
There are other alternative ways to generating SSL CAs in addition to openssl, for example Terraform. They are not documented here but you can find links to these alternative means in Related Projects and Documentation.
Generate a Certificate Authority
The simplest way to generate a certificate authority is to run two commands:
$ openssl genrsa -out ./ca.key.pem 4096
$ openssl req -key ca.key.pem -new -x509 -days 7300 -sha256 -out ca.cert.pem -extensions v3_ca
Enter pass phrase for ca.key.pem:
You are about to be asked to enter information that will be incorporated
into your certificate request.
What you are about to enter is what is called a Distinguished Name or a DN.
There are quite a few fields but you can leave some blank
For some fields there will be a default value,
If you enter '.', the field will be left blank.
-----
Country Name (2 letter code) [AU]:US
State or Province Name (full name) [Some-State]:CO
Locality Name (eg, city) []:Boulder
Organization Name (eg, company) [Internet Widgits Pty Ltd]:tiller
Organizational Unit Name (eg, section) []:
Common Name (e.g. server FQDN or YOUR name) []:tiller
Email Address []:tiller@example.com
Note that the data input above is sample data. You should customize to your own specifications.
The above will generate both a secret key and a CA. Note that these two files are very important. The key in particular should be handled with particular care.
Often, you will want to generate an intermediate signing key. For the sake of brevity, we will be signing keys with our root CA.
Generating Certificates
We will be generating two certificates, each representing a type of certificate:
- One certificate is for Tiller. You will want one of these per tiller host that you run.
- One certificate is for the user. You will want one of these per helm user.
Since the commands to generate these are the same, we’ll be creating both at the same time. The names will indicate their target.
First, the Tiller key:
$ openssl genrsa -out ./tiller.key.pem 4096
Generating RSA private key, 4096 bit long modulus
..........................................................................................................................................................................................................................................................................................................................++
............................................................................++
e is 65537 (0x10001)
Enter pass phrase for ./tiller.key.pem:
Verifying - Enter pass phrase for ./tiller.key.pem:
Next, generate the Helm client’s key:
$ openssl genrsa -out ./helm.key.pem 4096
Generating RSA private key, 4096 bit long modulus
.....++
......................................................................................................................................................................................++
e is 65537 (0x10001)
Enter pass phrase for ./helm.key.pem:
Verifying - Enter pass phrase for ./helm.key.pem:
Again, for production use you will generate one client certificate for each user.
Next we need to create certificates from these keys. For each certificate, this is a two-step process of creating a CSR, and then creating the certificate.
$ openssl req -key tiller.key.pem -new -sha256 -out tiller.csr.pem
Enter pass phrase for tiller.key.pem:
You are about to be asked to enter information that will be incorporated
into your certificate request.
What you are about to enter is what is called a Distinguished Name or a DN.
There are quite a few fields but you can leave some blank
For some fields there will be a default value,
If you enter '.', the field will be left blank.
-----
Country Name (2 letter code) [AU]:US
State or Province Name (full name) [Some-State]:CO
Locality Name (eg, city) []:Boulder
Organization Name (eg, company) [Internet Widgits Pty Ltd]:Tiller Server
Organizational Unit Name (eg, section) []:
Common Name (e.g. server FQDN or YOUR name) []:tiller-server
Email Address []:
Please enter the following 'extra' attributes
to be sent with your certificate request
A challenge password []:
An optional company name []:
And we repeat this step for the Helm client certificate:
$ openssl req -key helm.key.pem -new -sha256 -out helm.csr.pem
# Answer the questions with your client user's info
(In rare cases, we’ve had to add the -nodes flag when generating the request.)
Now we sign each of these CSRs with the CA certificate we created (adjust the days parameter to suit your requirements):
$ openssl x509 -req -CA ca.cert.pem -CAkey ca.key.pem -CAcreateserial -in tiller.csr.pem -out tiller.cert.pem -days 365
Signature ok
subject=/C=US/ST=CO/L=Boulder/O=Tiller Server/CN=tiller-server
Getting CA Private Key
Enter pass phrase for ca.key.pem:
And again for the client certificate:
$ openssl x509 -req -CA ca.cert.pem -CAkey ca.key.pem -CAcreateserial -in helm.csr.pem -out helm.cert.pem -days 365
At this point, the important files for us are these:
# The CA. Make sure the key is kept secret.
ca.cert.pem
ca.key.pem
# The Helm client files
helm.cert.pem
helm.key.pem
# The Tiller server files.
tiller.cert.pem
tiller.key.pem
Now we’re ready to move on to the next steps.
Creating a Custom Tiller Installation
Helm includes full support for creating a deployment configured for SSL. By specifying
a few flags, the helm init command can create a new Tiller installation complete
with all of our SSL configuration.
To take a look at what this will generate, run this command:
$ helm init --dry-run --debug --tiller-tls --tiller-tls-cert ./tiller.cert.pem --tiller-tls-key ./tiller.key.pem --tiller-tls-verify --tls-ca-cert ca.cert.pem
The output will show you a Deployment, a Secret, and a Service. Your SSL information will be preloaded into the Secret, which the Deployment will mount to pods as they start up.
If you want to customize the manifest, you can save that output to a file and then
use kubectl create to load it into your cluster.
We strongly recommend enabling RBAC on your cluster and adding service accounts with RBAC.
Otherwise, you can remove the --dry-run and --debug flags. We also recommend
putting Tiller in a non-system namespace (--tiller-namespace=something) and enable
a service account (--service-account=somename). But for this example we will stay
with the basics:
$ helm init --tiller-tls --tiller-tls-cert ./tiller.cert.pem --tiller-tls-key ./tiller.key.pem --tiller-tls-verify --tls-ca-cert ca.cert.pem
In a minute or two it should be ready. We can check Tiller like this:
$ kubectl -n kube-system get deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
... other stuff
tiller-deploy 1 1 1 1 2m
If there is a problem, you may want to use kubectl get pods -n kube-system to
find out what went wrong. With the SSL/TLS support, the most common problems all
have to do with improperly generated TLS certificates or accidentally swapping the
cert and the key.
At this point, you should get a failure when you run basic Helm commands:
$ helm ls
Error: transport is closing
This is because your Helm client does not have the correct certificate to authenticate to Tiller.
Configuring the Helm Client
The Tiller server is now running with TLS protection. It’s time to configure the Helm client to also perform TLS operations.
For a quick test, we can specify our configuration manually. We’ll run a normal
Helm command (helm ls), but with SSL/TLS enabled.
helm ls --tls --tls-ca-cert ca.cert.pem --tls-cert helm.cert.pem --tls-key helm.key.pem
This configuration sends our client-side certificate to establish identity, uses the client key for encryption, and uses the CA certificate to validate the remote Tiller’s identity.
Typing a line that is cumbersome, though. The shortcut is to move the key,
cert, and CA into $HELM_HOME:
$ cp ca.cert.pem $(helm home)/ca.pem
$ cp helm.cert.pem $(helm home)/cert.pem
$ cp helm.key.pem $(helm home)/key.pem
With this, you can simply run helm ls --tls to enable TLS.
Troubleshooting
Running a command, I get Error: transport is closing
This is almost always due to a configuration error in which the client is missing
a certificate (--tls-cert) or the certificate is bad.
I’m using a certificate, but get Error: remote error: tls: bad certificate
This means that Tiller’s CA cannot verify your certificate. In the examples above, we used a single CA to generate both the client and server certificates. In these examples, the CA has signed the client’s certificate. We then load that CA up to Tiller. So when the client certificate is sent to the server, Tiller checks the client certificate against the CA.
If I use --tls-verify on the client, I get Error: x509: certificate is valid for tiller-server, not localhost
If you plan to use --tls-verify on the client, you will need to make sure that
the host name that Helm connects to matches the host name on the certificate. In
some cases this is awkward, since Helm will connect over localhost, or the FQDN is
not available for public resolution.
If I use --tls-verify on the client, I get Error: x509: cannot validate certificate for 127.0.0.1 because it doesn't contain any IP SANs
By default, the Helm client connects to Tiller via tunnel (i.e. kube proxy) at 127.0.0.1. During the TLS handshake, a target, usually provided as a hostname (e.g. example.com), is checked against the subject and subject alternative names of the certificate (i.e. hostname verification). However, because of the tunnel, the target is an IP address. Therefore, to validate the certificate, the IP address 127.0.0.1 must be listed as an IP subject alternative name (IP SAN) in the Tiller certificate.
For example, to list 127.0.0.1 as an IP SAN when generating the Tiller certificate:
$ echo subjectAltName=IP:127.0.0.1 > extfile.cnf
$ openssl x509 -req -CA ca.cert.pem -CAkey ca.key.pem -CAcreateserial -in tiller.csr.pem -out tiller.cert.pem -days 365 -extfile extfile.cnf
Alternatively, you can override the expected hostname of the tiller certificate using the --tls-hostname flag.
If I use --tls-verify on the client, I get Error: x509: certificate has expired or is not yet valid
Your helm certificate has expired, you need to sign a new certificate using your private key and the CA (and consider increasing the number of days)
If your tiller certificate has expired, you’ll need to sign a new certificate, base64 encode it and update the Tiller Secret:
kubectl edit secret tiller-secret
References
https://github.com/denji/golang-tls https://www.openssl.org/docs/ https://jamielinux.com/docs/openssl-certificate-authority/sign-server-and-client-certificates.html
Securing your Helm Installation
Helm is a powerful and flexible package-management and operations tool for Kubernetes. Installing it using the default installation command – helm init – quickly and easily installs Tiller, the server-side component with which Helm corresponds.
This default installation applies no security configurations, however. It’s completely appropriate to use this type of installation when you are working against a cluster with no or very few security concerns, such as local development with Minikube or with a cluster that is well-secured in a private network with no data-sharing or no other users or teams. If this is the case, then the default installation is fine, but remember: With great power comes great responsibility. Always use due diligence when deciding to use the default installation.
Who Needs Security Configurations?
For the following types of clusters we strongly recommend that you apply the proper security configurations to Helm and Tiller to ensure the safety of the cluster, the data in it, and the network to which it is connected.
- Clusters that are exposed to uncontrolled network environments: either untrusted network actors can access the cluster, or untrusted applications that can access the network environment.
- Clusters that are for many people to use – multitenant clusters – as a shared environment
- Clusters that have access to or use high-value data or networks of any type
Often, environments like these are referred to as production grade or production quality because the damage done to any company by misuse of the cluster can be profound for either customers, the company itself, or both. Once the risk of damage becomes high enough, you need to ensure the integrity of your cluster no matter what the actual risk.
To configure your installation properly for your environment, you must:
- Understand the security context of your cluster
- Choose the Best Practices you should apply to your helm installation
The following assumes you have a Kubernetes configuration file (a kubeconfig file) or one was given to you to access a cluster.
Understanding the Security Context of your Cluster
helm init installs Tiller into the cluster in the kube-system namespace and without any RBAC rules applied. This is appropriate for local development and other private scenarios because it enables you to be productive immediately. It also enables you to continue running Helm with existing Kubernetes clusters that do not have role-based access control (RBAC) support until you can move your workloads to a more recent Kubernetes version.
There are four main areas to consider when securing a tiller installation:
- Role-based access control, or RBAC
- Tiller’s gRPC endpoint and its usage by Helm
- Tiller release information
- Helm charts
RBAC
Recent versions of Kubernetes employ a role-based access control (or RBAC) system (as do modern operating systems) to help mitigate the damage that can be done if credentials are misused or bugs exist. Even where an identity is hijacked, the identity has only so many permissions to a controlled space. This effectively adds a layer of security to limit the scope of any attack with that identity.
Helm and Tiller are designed to install, remove, and modify logical applications that can contain many services interacting together. As a result, often its usefulness involves cluster-wide operations, which in a multitenant cluster means that great care must be taken with access to a cluster-wide Tiller installation to prevent improper activity.
Specific users and teams – developers, operators, system and network administrators – will need their own portion of the cluster in which they can use Helm and Tiller without risking other portions of the cluster. This means using a Kubernetes cluster with RBAC enabled and Tiller configured to enforce them. For more information about using RBAC in Kubernetes, see Using RBAC Authorization.
Tiller and User Permissions
Tiller in its current form does not provide a way to map user credentials to specific permissions within Kubernetes. When Tiller is running inside of the cluster, it operates with the permissions of its service account. If no service account name is supplied to Tiller, it runs with the default service account for that namespace. This means that all Tiller operations on that server are executed using the Tiller pod’s credentials and permissions.
To properly limit what Tiller itself can do, the standard Kubernetes RBAC mechanisms must be attached to Tiller, including Roles and RoleBindings that place explicit limits on what things a Tiller instance can install, and where.
This situation may change in the future. While the community has several methods that might address this, at the moment performing actions using the rights of the client, instead of the rights of Tiller, is contingent upon the outcome of the Pod Identity Working Group, which has taken on the task of solving the problem in a general way.
The Tiller gRPC Endpoint
In the default installation the gRPC endpoint that Tiller offers is available inside the cluster (not external to the cluster) without authentication configuration applied. Without applying authentication, any process in the cluster can use the gRPC endpoint to perform operations inside the cluster. In a local or secured private cluster, this enables rapid usage and is normal. (When running outside the cluster, Helm authenticates through the Kubernetes API server to reach Tiller, leveraging existing Kubernetes authentication support.)
The following two sub-sections describe options of how to setup Tiller so there isn’t an unauthenticated endpoint (i.e. gRPC) in your cluster.
Enabling TLS
(Note that out of the two options, this is the recommended one for Helm 2.)
Shared and production clusters – for the most part – should use Helm 2.7.2 at a minimum and configure TLS for each Tiller gRPC endpoint to ensure that within the cluster usage of gRPC endpoints is only for the properly authenticated identity for that endpoint (i.e. configure each endpoint to use a separate TLS certificate). Doing so enables any number of Tiller instances to be deployed in any number of namespaces and yet no unauthenticated usage of any gRPC endpoint is possible. Finally, use Helm init with the --tiller-tls-verify option to install Tiller with TLS enabled and to verify remote certificates, and all other Helm commands should use the --tls option.
For more information about the proper steps to configure Tiller and use Helm properly with TLS configured, see the Best Practices section below, and Using SSL between Helm and Tiller.
When Helm clients are connecting from outside of the cluster, the security between the Helm client and the API server is managed by Kubernetes itself. You may want to ensure that this link is secure. Note that if you are using the TLS configuration recommended above, not even the Kubernetes API server has access to the encrypted messages between the client and Tiller.
Running Tiller Locally
Contrary to the previous Enabling TLS section, this section does not involve running a tiller server pod in your cluster (for what it’s worth, that lines up with the current helm v3 proposal), thus there is no gRPC endpoint (and thus there’s no need to create & manage TLS certificates to secure each gRPC endpoint).
Steps:
- Fetch the latest helm release tarball from the GitHub release page, and extract and move
helmandtillersomewhere on your$PATH. - “Server”: Run
tiller --storage=secret. (Note thattillerhas a default value of “:44134” for the--listenargument.) - Client: In another terminal (and on the same host that the aforementioned
tillercommand was run for the previous bullet): Runexport HELM_HOST=:44134, and then runhelmcommands as usual.
Tiller’s Release Information
For historical reasons, Tiller stores its release information in ConfigMaps. We suggest changing the default to Secrets.
Secrets are the Kubernetes accepted mechanism for saving configuration data that is considered sensitive. While secrets don’t themselves offer many protections, Kubernetes cluster management software often treats them differently than other objects. Thus, we suggest using secrets to store releases.
Enabling this feature currently requires setting the --storage=secret flag in the tiller-deploy deployment. This entails directly modifying the deployment or using helm init --override 'spec.template.spec.containers[0].command'='{/tiller,--storage=secret}', as no helm init flag is currently available to do this for you.
Thinking about Charts
Because of the relative longevity of Helm, the Helm chart ecosystem evolved without the immediate concern for cluster-wide control, and especially in the developer space this makes complete sense. However, charts are a kind of package that not only installs containers you may or may not have validated yourself, but it may also install into more than one namespace.
As with all shared software, in a controlled or shared environment you must validate all software you install yourself before you install it. If you have secured Tiller with TLS and have installed it with permissions to only one or a subset of namespaces, some charts may fail to install – but in these environments, that is exactly what you want. If you need to use the chart, you may have to work with the creator or modify it yourself in order to use it securely in a multitenant cluster with proper RBAC rules applied. The helm template command renders the chart locally and displays the output.
Once vetted, you can use Helm’s provenance tools to ensure the provenance and integrity of charts that you use.
gRPC Tools and Secured Tiller Configurations
Many very useful tools use the gRPC interface directly, and having been built against the default installation – which provides cluster-wide access – may fail once security configurations have been applied. RBAC policies are controlled by you or by the cluster operator, and either can be adjusted for the tool, or the tool can be configured to work properly within the constraints of specific RBAC policies applied to Tiller. The same may need to be done if the gRPC endpoint is secured: the tools need their own secure TLS configuration in order to use a specific Tiller instance. The combination of RBAC policies and a secured gRPC endpoint configured in conjunction with gRPC tools enables you to control your cluster environment as you should.
Best Practices for Securing Helm and Tiller
The following guidelines reiterate the Best Practices for securing Helm and Tiller and using them correctly.
- Create a cluster with RBAC enabled
- Configure each Tiller gRPC endpoint to use a separate TLS certificate
- Release information should be a Kubernetes Secret
- Install one Tiller per user, team, or other organizational entity with the
--service-accountflag, Roles, and RoleBindings - Use the
--tiller-tls-verifyoption withhelm initand the--tlsflag with other Helm commands to enforce verification
If these steps are followed, an example helm init command might look something like this:
$ helm init \
--override 'spec.template.spec.containers[0].command'='{/tiller,--storage=secret}' \
--tiller-tls \
--tiller-tls-verify \
--tiller-tls-cert=cert.pem \
--tiller-tls-key=key.pem \
--tls-ca-cert=ca.pem \
--service-account=accountname
This command will start Tiller with strong authentication over gRPC, release information stored in a Kubernetes Secret, and a service account to which RBAC policies have been applied.
