luxolus
/
open-consul
2
0
Fork 0
Code Issues 1 Pull Requests Packages Releases Wiki Activity

Initial Helm Chart/K8S Docs (#4653) 

* website: initial Kubernetes section with Helm information

* website: extraConfig for clients

* website: add more helm fields

* website: document extraVolumes

* website: document Consul DNS

* website: fix typos and show example of downward API
This commit is contained in:
Mitchell Hashimoto 2018-09-12 08:44:30 -07:00 committed by GitHub
parent 5ecf9823d2
commit b69342f0c1
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
5 changed files with 696 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

102
website/source/docs/platform/k8s/dns.html.md Normal file
View File

@ -0,0 +1,102 @@
---
layout: "docs"
page_title: "Consul DNS - Kubernetes"
sidebar_current: "docs-platform-k8s-dns"
description: |-
One of the primary query interfaces to Consul is the DNS interface. The Consul DNS interface can be exposed for all pods in Kubernetes using a stub-domain configuration.
---
# Consul DNS on Kubernetes
One of the primary query interfaces to Consul is the
[DNS interface](/docs/agent/dns.html). The Consul DNS interface can be
exposed for all pods in Kubernetes using a
[stub-domain configuration](https://kubernetes.io/docs/tasks/administer-cluster/dns-custom-nameservers/#configure-stub-domain-and-upstream-dns-servers).
The stub-domain configuration must point to a static IP of a DNS resolver.
The [Helm chart](/docs/platform/k8s/helm.html) creates a `consul-dns` service
by default that exports Consul DNS. The cluster IP of this service can be used
to configure a stub-domain with kube-dns. While the `kube-dns` configuration
lives in the `kube-system` namepace, the IP just has to be routable so the
service can live in a different namespace.
```
cat <<EOF | kubectl apply -f -
apiVersion: v1
kind: ConfigMap
metadata:
labels:
addonmanager.kubernetes.io/mode: EnsureExists
name: kube-dns
namespace: kube-system
data:
stubDomains: |
{"consul": ["$(kubectl get svc consul-dns -o jsonpath='{.spec.clusterIP}')"]}
EOF
```
-> **Note:** The `stubDomain` can only point to a static IP. If the cluster IP
of the `consul-dns` service changes, then it must be updated to continue
working. This can happen if the service is deleted and recreated, such as
in full cluster rebuilds.
## Verifying DNS Works
To verify DNS works, run a simple job to query DNS. Save the following
job to the file `job.yaml` and run it:
```yaml
apiVersion: batch/v1
kind: Job
metadata:
name: dns
spec:
template:
spec:
containers:
- name: dns
image: anubhavmishra/tiny-tools
command: ["dig", "consul.service.consul"]
restartPolicy: Never
backoffLimit: 4
```
```sh
$ kubectl apply -f job.yaml
```
Then query the pod name for the job and check the logs. You should see
output similar to the following showing a successful DNS query. If you see
any errors, then DNS is not configured properly.
```
$ kubectl get pods --show-all | grep dns
dns-lkgzl 0/1 Completed 0 6m
$ kubectl logs dns-lkgzl
; <<>> DiG 9.11.2-P1 <<>> consul.service.consul
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 4489
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 4
;; OPT PSEUDOSECTION:
; EDNS: version: 0, flags:; udp: 4096
;; QUESTION SECTION:
;consul.service.consul. IN A
;; ANSWER SECTION:
consul.service.consul. 0 IN A 10.36.2.23
consul.service.consul. 0 IN A 10.36.4.12
consul.service.consul. 0 IN A 10.36.0.11
;; ADDITIONAL SECTION:
consul.service.consul. 0 IN TXT "consul-network-segment="
consul.service.consul. 0 IN TXT "consul-network-segment="
consul.service.consul. 0 IN TXT "consul-network-segment="
;; Query time: 5 msec
;; SERVER: 10.39.240.10#53(10.39.240.10)
;; WHEN: Wed Sep 12 02:12:30 UTC 2018
;; MSG SIZE rcvd: 206
```

261
website/source/docs/platform/k8s/helm.html.md Normal file
View File

@ -0,0 +1,261 @@
---
layout: "docs"
page_title: "Helm - Kubernetes"
sidebar_current: "docs-platform-k8s-helm"
description: |-
The Consul Helm chart is the recommended way to install and configure Consul on Kubernetes. In addition to running Consul itself, the Helm chart is the primary method for installing and configuring Consul integrations with Kubernetes such as catalog syncing, Connect injection, and more.
---
# Helm Chart
The [Consul Helm chart](https://github.com/hashicorp/consul-helm)
is the recommended way to install and configure Consul on Kubernetes.
In addition to running Consul itself, the Helm chart is the primary
method for installing and configuring Consul integrations with
Kubernetes such as catalog syncing, Connect injection, and more.
This page assumes general knowledge of [Helm](https://helm.sh/) and
how to use it. Using Helm to install Consul will require that Helm is
properly installed and configured with your Kubernetes cluster.
-> **Important:** The Helm chart is new and
may still change significantly over time. Please always run Helm with
`--dry-run` before any install or upgrade to verify changes.
## Using the Helm Chart
To use the Helm chart, you must download or clone the
[consul-helm GitHub repository](https://github.com/hashicorp/consul-helm)
and run Helm against the directory. We plan to transition to using a real
Helm repository soon. When running Helm, we highly recommend you always
checkout a specific tagged release of the chart to avoid any
instabilities from master.
Prior to this, you must have Helm installed and configured both in your
Kubernetes cluster and locally on your machine. The steps to do this are
out of the scope of this document, please read the
[Helm documentation](https://helm.sh/) for more information.
Example chart usage:
```sh
# Clone the chart repo
$ git clone https://github.com/hashicorp/consul-helm.git
$ cd consul-helm
# Checkout a tagged version
$ git checkout v0.1.0
# Run Helm
$ helm install --dry-run ./
```
~> **Warning:** By default, the chart will install _everything_: a
Consul server cluster, client agents on all nodes, feature components, etc.
This provides a nice out-of-box experience for new users, but may not be
appropriate for a production setup. Considering setting the `global.enabled`
value to `false` and opt-in to the various components.
## Configuration (Values)
The chart is highly customizable using
[Helm configuration values](https://docs.helm.sh/using_helm/#customizing-the-chart-before-installing).
Each value has a sane default tuned for an optimal getting started experience
with Consul. Before going into production, please review the parameters below
and consider if they're appropriate for your deployment.
* <a name="v-global" href="#v-global">`global`</a> - These values are global values that affect multiple components of the chart.
- <a name="v-global-enabled" href="#v-global-enabled">`enabled`</a> (`boolean: true`) -
The master enabled/disabled configuration. If this is true, most components
will be installed by default. If this is false, no components will be installed
by default and manually opt-in is required, such as by setting
<a href="#v-">`server.enabled`</a> to true.
- <a name="v-global-domain" href="#v-global-domain">`domain`</a> (`string: "consul"`) -
The domain Consul uses for DNS queries. This is used to configure agents both
for DNS listening but also to know what domain to join to join the cluster.
This should be consistent throughout the chart, but can be overridden
per-component as well.
- <a name="v-global-image" href="#v-global-image">`image`</a> (`string: "consul:latest"`) -
The name of the Docker image (including any tag) for the containers running
Consul agents. **This should be pinned to a specific version when running
in production.** Otherwise, other changes to the chart may inadvertently
upgrade your Consul version.
- <a name="v-global-datacenter" href="#v-global-datacenter">`datacenter`</a> (`string: "dc1"`) -
The name of the datacenter that the agent cluster should register as.
This must not be changed once the cluster is bootstrapped and running,
since Consul doesn't yet support an automatic way to change this value.
* <a name="v-server" href="#v-server">`server`</a> - Values that configure
running a Consul server cluster within Kubernetes.
- <a name="v-server-enabled" href="#v-server-enabled">`enabled`</a> (`boolean: global.enabled`) -
If true, the chart will install all the resources necessary for a Consul
server cluster. If you're running Consul externally and want agents within
Kubernetes to join that cluster, this should probably be false.
- <a name="v-server-image" href="#v-server-image">`image`</a> (`string: global.image`) -
The name of the Docker image (including any tag) for the containers running
Consul server agents.
- <a name="v-server-replicas" href="#v-server-replicas">`replicas`</a> (`integer: 3`) -
The number of server agents to run. This determines the fault tolerance
of the cluster. Please see the [deployment table](/docs/internals/consensus.html#deployment-table)
for more information.
- <a name="v-server-bootstrapexpect" href="#v-server-bootstrapexpect">`bootstrapExpect`</a> (`integer: 3`) -
For new clusters, this is the number of servers to wait for before
performing the initial leader election and bootstrap of the cluster. This
must be less than or equal to `server.replicas`. This value is only
used when bootstrapping new clusters, it has no effect during ongoing cluster
maintenance.
- <a name="v-server-connect" href="#v-server-connect">`connect`</a> (`boolean: true`) -
This will enable/disable [Connect](/docs/connect/index.html). Setting this
to true _will not_ automatically secure pod communication, this setting will
only enable usage of the feature. Consul will automatically initialize a new
CA and set of certificates. Additional Connect settings can be configured
by setting the `server.extraConfig` value.
- <a name="v-server-disruptionbudget" href="#v-server-disruptionbudget">`disruptionBudget`</a> -
This configures the
[PodDisruptionBudget](https://kubernetes.io/docs/tasks/run-application/configure-pdb/)
for the server cluster.
* <a name="v-server-disruptionbudget-enabled" href="#v-server-disruptionbudget-enabled">`enabled`</a> (`boolean: true`) -
This will enable/disable registering a PodDisruptionBudget for
the server cluster. If this is enabled, it will only register the
budget so long as the server cluster is enabled.
* <a name="v-server-disruptionbudget-maxunavailable" href="#v-server-disruptionbudget-maxunavailable">`maxUnavailable`</a> (`integer: null`) -
The maximum number of unavailable pods. By default, this will be automatically
computed based on the `server.replicas` value to be `(n/2)-1`.
- <a name="v-server-extraconfig" href="#v-server-extraconfig">`extraConfig`</a> (`string: "{}"`) -
A raw string of extra JSON or HCL configuration for Consul servers. This
will be saved as-is into a ConfigMap that is read by the Consul server agents.
This can be used to add additional configuration that isn't directly exposed
by the chart.
- <a name="v-server-extravolumes" href="#v-server-extravolumes">`extraVolumes`</a> (`array: []`) -
A list of extra volumes to mount for server agents. This is useful for bringing
in extra data that can be referenced by other configurations at a well known
path, such as TLS certificates or Gossip encryption keys.
The value of this should be a list of objects. Each object has the following
supports the following keys:
* <a name="v-server-extravolumes-type" href="#v-server-extravolumes-type">`type`</a> (`string: required`) -
Type of the volume, must be one of "configMap" or "secret". Case sensitive.
* <a name="v-server-extravolumes-name" href="#v-server-extravolumes-name">`name`</a> (`string: required`) -
Name of the configMap or secret to be mounted. This also controls the path
that it is mounted to. The volume will be mounted to `/config/userconfig/<name>`.
* <a name="v-server-extravolumes-load" href="#v-server-extravolumes-load">`load`</a> (`boolean: false`) -
If true, then the agent will be configured to automatically load HCL/JSON
configuration files from this volume with `-config-dir`. This defaults
to false.
- <a name="v-server-resources" href="#v-server-resources">`resources`</a> (`object: {}`) -
The resource requests (CPU, memory, etc.) for each of the server agents.
This should be an object mapping directly to a Kubernetes
[ResourceRequirements](https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.11/#resourcerequirements-v1-core) object. If this isn't specified, then the pods
won't request any specific amount of resources. **Setting this is highly
recommended.**
- <a name="v-server-storage" href="#v-server-storage">`storage`</a> (`string: "10Gi"`) -
The amount of storage to request for the persistant volume backing the
Consul server data.
- <a name="v-server-updatepartition" href="#v-server-updatepartition">`updatePartition`</a> (`integer: 0`) -
This value is used to carefully control a rolling update of Consul server
agents. This value specifies the
[partition](https://kubernetes.io/docs/concepts/workloads/controllers/statefulset/#partitions)
for performing a rolling update. Please read the linked Kubernetes
documentation for more information.
* <a name="v-client" href="#v-client">`client`</a> - Values that configure
running a Consul client agent on Kubernetes nodes.
- <a name="v-client-enabled" href="#v-client-enabled">`enabled`</a> (`boolean: global.enabled`) -
If true, the chart will install all the resources necessary for a Consul
client on every Kubernetes node. This _does not_ require `server.enabled`,
since the agents can be configured to join an external cluster.
- <a name="v-client-image" href="#v-client-image">`image`</a> (`string: global.image`) -
The name of the Docker image (including any tag) for the containers running
Consul client agents.
- <a name="v-client-extraconfig" href="#v-client-extraconfig">`extraConfig`</a> (`string: "{}"`) -
A raw string of extra JSON or HCL configuration for Consul clients. This
will be saved as-is into a ConfigMap that is read by the Consul agents.
This can be used to add additional configuration that isn't directly exposed
by the chart.
- <a name="v-client-extravolumes" href="#v-client-extravolumes">`extraVolumes`</a> (`array: []`) -
A list of extra volumes to mount for client agents. This is useful for bringing
in extra data that can be referenced by other configurations at a well known
path, such as TLS certificates or Gossip encryption keys.
The value of this should be a list of objects. Each object has the following
supports the following keys:
* <a name="v-client-extravolumes-type" href="#v-client-extravolumes-type">`type`</a> (`string: required`) -
Type of the volume, must be one of "configMap" or "secret". Case sensitive.
* <a name="v-client-extravolumes-name" href="#v-client-extravolumes-name">`name`</a> (`string: required`) -
Name of the configMap or secret to be mounted. This also controls the path
that it is mounted to. The volume will be mounted to `/config/userconfig/<name>`.
* <a name="v-client-extravolumes-load" href="#v-client-extravolumes-load">`load`</a> (`boolean: false`) -
If true, then the agent will be configured to automatically load HCL/JSON
configuration files from this volume with `-config-dir`. This defaults
to false.
- <a name="v-client-join" href="#v-client-join">`join`</a> (`array<string>: null`) -
A list of values to specify to `-retry-join`. If this is `null` (default),
then the clients will attempt to automatically join the server cluster
running within Kubernetes. This means that with `server.enabled` set to true,
clients will automatically join that cluster. If `server.enabled` is not
true, then a value must be specified so the clients can join a valid cluster.
- <a name="v-client-resources" href="#v-client-resources">`resources`</a> (`object: {}`) -
The resource requests (CPU, memory, etc.) for each of the client agents.
This should be an object mapping directly to a Kubernetes
[ResourceRequirements](https://kubernetes.io/docs/reference/generated/kubernetes-api/v1.11/#resourcerequirements-v1-core) object. If this isn't specified, then the pods
won't request any specific amount of resources.
* <a name="v-dns" href="#v-dns">`dns`</a> - Values that configure the Consul DNS service.
- <a name="v-dns-enabled" href="#v-dns-enabled">`enabled`</a> (`boolean: global.enabled`) -
If true, a `consul-dns` service will be created that exposes port 53 for
TCP and UDP to the running Consul agents (servers and clients). This can
then be used to [configure kube-dns](/docs/platform/k8s/dns.html). The Helm
chart _does not_ automatically configure kube-dns.
* <a name="v-ui" href="#v-ui">`ui`</a> - Values that configure the Consul UI.
- <a name="v-ui-enabled" href="#v-ui-enabled">`enabled`</a> (`boolean: global.enabled`) -
If true, the UI will be enabled. This will only _enable_ the UI, it doesn't
automatically register any service for external access. The UI will only
be enabled on server agents. If `server.enabled` is false, then this setting
has no effect. To expose the UI in some way, you must configure
`ui.service`.
- <a name="v-ui-service" href="#v-ui-service">`service`</a> -
This configures the `Service` resource registered for the Consul UI.
* <a name="v-ui-service-enabled" href="#v-ui-service-enabled">`enabled`</a> (`boolean: true`) -
This will enable/disable registering a Kubernetes Service for the Consul UI.
This value only takes effect if `ui.enabled` is true and taking effect.
* <a name="v-ui-service-type" href="#v-ui-service-type">`type`</a> (`string: null`) -
The service type to register. This defaults to `null` which doesn't set
an explicit service type, which typically is defaulted to "ClusterIP"
by Kubernetes. The available service types are documented on
[the Kubernetes website](https://kubernetes.io/docs/concepts/services-networking/service/#publishing-services-service-types).

44
website/source/docs/platform/k8s/index.html.md Normal file
View File

@ -0,0 +1,44 @@
---
layout: "docs"
page_title: "Kubernetes"
sidebar_current: "docs-platform-k8s-index"
description: |-
Consul has many integrations with Kubernetes. You can deploy Consul to Kubernetes using the Helm Chart, sync services between Consul and Kubernetes, automatically secure Pod communication with Connect, and more. This section documents the official integrations between Consul and Kubernetes.
---
# Kubernetes
Consul has many integrations with Kubernetes. You can deploy Consul
to Kubernetes using the Helm Chart, sync services between Consul and
Kubernetes, automatically secure Pod communication with Connect, and more.
This section documents the official integrations between Consul and Kubernetes.
## Use Cases
**Running a Consul server cluster:** The Consul server cluster can run directly
on Kubernetes. This can be used by both nodes within Kubernetes as well as
nodes external to Kubernetes, as long as they can communicate to the server
nodes via the network.
**Running Consul clients:** Consul clients can run as pods on every node
and expose the Consul API to running pods. This enables many Consul tools
such as envconsul, consul-template, and more to work on Kubernetes since a
local agent is available. This will also register each Kubernetes node with
the Consul catalog for full visibility into your infrastructure
**Service sync to enable Kubernetes and non-Kubernetes services to communicate:**
Consul can sync Kubernetes services with its own service registry. This allows
Kubernetes services to use native Kubernetes service discovery to discover
and connect to external services, and for external services to use Consul
service discovery to discover and connect to Kubernetes services.
**Automatic encryption and authorization for pod network connections:**
Consul can automatically inject the [Connect](/docs/connect/index.html)
sidecar into pods so that they can accept and establish encrypted
and authorized network connections via mutual TLS. And because Connect
can run anywhere, pods can also communicate with external services (and
vice versa) over a fully encrypted connection.
**And more!** Consul can run directly on Kubernetes, so in addition to the
native integrations provided by Consul itself, any other tool built for
Kubernetes can choose to leverage Consul.

274
website/source/docs/platform/k8s/run.html.md Normal file
View File

@ -0,0 +1,274 @@
---
layout: "docs"
page_title: "Running Consul - Kubernetes"
sidebar_current: "docs-platform-k8s-run"
description: |-
Consul can run directly on Kubernetes, both in server or client mode. For pure-Kubernetes workloads, this enables Consul to also exist purely within Kubernetes. For heterogeneous workloads, Consul agents can join a server running inside or outside of Kubernetes.
---
# Running Consul on Kubernetes
Consul can run directly on Kubernetes, both in server or client mode.
For pure-Kubernetes workloads, this enables Consul to also exist purely
within Kubernetes. For heterogeneous workloads, Consul agents can join
a server running inside or outside of Kubernetes.
This page starts with a large how-to section for various specific tasks.
To learn more about the general architecture of Consul on Kubernetes, scroll
down to the [architecture](/docs/platform/k8s/run.html#architecture) section.
## Helm Chart
The recommended way to run Consul on Kubernetes is via the
[Helm Chart](/docs/platform/k8s/helm.html). This will install and configure
all the necessary components to run Consul. The configuration enables you
to run just a server cluster, just a client cluster, or both. Using the Helm
Chart, you can have a full Consul deployment up and running in seconds.
While the Helm Chart exposes dozens of useful configurations and automatically
sets up complex resources, it **does not automatically operate Consul.**
You are still reponsible for learning how to monitor, backup,
upgrade, etc. the Consul cluster.
The Helm Chart has no required configuration and will install a Consul
cluster with sane defaults out of the box. Prior to going to production,
it is highly recommended that you
[learn about the configuration options](/docs/platform/k8s/helm.html#configuration-values-).
## How-To
### Installing Consul
To install Consul, clone the consul-helm repository, checkout the latest release, and install
Consul. You can run `helm install` with the `--dry-run` flag to see the
resources it would configure. In a production environment, you should always
use the `--dry-run` flag prior to making any changes to the Consul cluster
via Helm.
```sh
# Clone the chart repo
$ git clone https://github.com/hashicorp/consul-helm.git
$ cd consul-helm
# Checkout a tagged version
$ git checkout v0.1.0
# Run Helm
$ helm install --name consul ./
...
```
_That's it._ The Helm chart does everything to setup a recommended
Consul-on-Kubernetes deployment.
In a couple minutes, a Consul cluster will be formed and a leader
elected and every node will have a running Consul agent.
The defaults will install both server and client agents. To install
only one or the other, see the
[chart configuration values](/docs/platform/k8s/helm.html#configuration-values-).
### Viewing the Consul UI
The Consul UI is enabled by default when using the Helm chart.
For security reasons, it isn't exposed via a Service by default so you must
use `kubectl port-forward` to visit the UI. Once the port is forwarded as
shown below, navigate your browser to `http://localhost:8500`.
```
$ kubectl port-forward consul-server-0 8500:8500
...
```
The UI can also be exposed via a Kubernetes Service. To do this, configure
the [`ui.service` chart values](/docs/platform/k8s/helm.html#v-ui-service).
### Joining an Existing Consul Cluster
If you have a Consul cluster already running, you can configure your
Kubernetes nodes to join this existing cluster.
```yaml
global:
enabled: false
client:
enabled: true
join:
- "provider=my-cloud config=val ..."
```
The `values.yaml` file to configure the Helm chart sets the proper
configuration to join an existing cluster.
The `global.enabled` value first disables all chart components by default
so that each component is opt-in. This allows us to _only_ setup the client
agents. We then opt-in to the client agents by setting `client.enabled` to
`true`.
Next, `client.join` is set to an array of valid
[`-retry-join` values](/docs/agent/options.html#retry-join). In the
example above, a fake [cloud auto-join](/docs/agent/cloud-auto-join.html)
value is specified. This should be set to resolve to the proper addresses of
your existing Consul cluster.
-> **Networking:** Note that for the Kubernetes nodes to join an existing
cluster, the nodes (and specifically the agent pods) must be able to connect
to all other server and client agents inside and _outside_ of Kubernetes.
If this isn't possible, consider running the Kubernetes agents as a separate
DC or adopting Enterprise for
[network segments](/docs/enterprise/network-segments/index.html).
### Accessing the Consul HTTP API
The Consul HTTP API should be accessed by communicating to the local agent
running on the same node. While technically any listening agent (client or
server) can respond to the HTTP API, communicating with the local agent
has important caching behavior, and allows you to use the simpler
[`/agent` endpoints for services and checks](/api/agent.html).
For Consul installed via the Helm chart, a client agent is installed on
each Kubernetes node. This is explained in the [architecture](/docs/platform/k8s/run.html#client-agents)
section. To access the agent, you may use the
[downward API](https://kubernetes.io/docs/tasks/inject-data-application/downward-api-volume-expose-pod-information/).
An example pod specification is shown below:
```yaml
apiVersion: v1
kind: Pod
metadata:
name: consul-example
spec:
containers:
- name: example
image: "consul:latest"
env:
- name: HOST_IP
valueFrom:
fieldRef:
fieldPath: status.hostIP
command:
- "/bin/sh"
- "-ec"
- |
export CONSUL_HTTP_ADDR="${HOST_IP}:8500"
consul kv put hello world
restartPolicy: Never
```
### Upgrading Consul on Kubernetes
To upgrade Consul on Kubernetes, we follow the same pattern as
[generally upgrading Consul](/docs/upgrading.html), except we can use
the Helm chart to step through a rolling deploy. It is important to understand
how to [generally upgrade Consul](/docs/upgrading.html) before reading this
section.
Upgrading Consul on Kubernetes will follow the same pattern: each server
will be updated one-by-one. After that is successful, the clients will
be updated in batches.
#### Upgrading Consul Servers
To initiate the upgrade, change the `server.image` value to the
desired Consul version. For illustrative purposes, the example below will
use `consul:123.456`. Also set the `server.updatePartition` value
_equal to the number of server replicas_:
```yaml
server:
image: "consul:123.456"
replicas: 3
updatePartition: 3
```
The `updatePartition` value controls how many instances of the server
cluster are updated. Only instances with an index _greater than_ the
`updatePartition` value are updated (zero-indexed). Therefore, by setting
it equal to replicas, none should update yet.
Next, run the upgrade. You should run this with `--dry-run` first to verify
the changes that will be sent to the Kubernetes cluster.
```
$ helm upgrade consul ./
...
```
This should cause no changes (although the resource will be updated). If
everything is stable, begin by decreasing the `updatePartition` value by one,
and running `helm upgrade` again. This should cause the first Consul server
to be stopped and restarted with the new image.
Wait until the Consul server cluster is healthy again (30s to a few minutes)
then decrease `updatePartition` and upgrade again. Continue until
`updatePartition` is `0`. At this point, you may remove the
`updatePartition` configuration. Your server upgrade is complete.
#### Upgrading Consul Clients
With the servers upgraded, it is time to upgrade the clients. To upgrade
the clients, set the `client.image` value to the desired Consul version.
Then, run `helm upgrade`. This will upgrade the clients in batches, waiting
until the clients come up healthy before continuing.
## Architecture
We recommend running Consul on Kubernetes with the same
[general architecture](/docs/internals/architecture.html)
as running it anywhere else. There are some benefits Kubernetes can provide
that eases operating a Consul cluster and we document those below. The standard
[production deployment guide](/docs/guides/deployment.html) is still an
important read even if running Consul within Kubernetes.
Each section below will outline the different components of running Consul
on Kubernetes and an overview of the resources that are used within the
Kubernetes cluster.
### Server Agents
The server agents are run as a **StatefulSet**, using persistent volume
claims to store the server state. This also ensures that the
[node ID](/docs/agent/options.html#_node_id) is persisted so that servers
can be rescheduled onto new IP addresses without causing issues. The server agents
are configured with
[anti-affinity](https://kubernetes.io/docs/concepts/configuration/assign-pod-node/#affinity-and-anti-affinity)
rules so that they are placed on different nodes. A readiness probe is
configured that marks the pod as ready only when it has established a leader.
A **Service** is registered to represent the servers and expose the various
ports. The DNS address of this service is used to join the servers to each
other without requiring any other access to the Kubernetes cluster. The
service is configured to publish non-ready endpoints so that it can be used
for joining during bootstrap and upgrades.
Additionally, a **PodDisruptionBudget** is configured so the Consul server
cluster maintains quorum during voluntary operational events. The maximum
unavailable is `(n/2)-1` where `n` is the number of server agents.
### Client Agents
The client agents are run as a **DaemonSet**. This places one agent
(within its own pod) on each Kubernetes node.
The clients expose the Consul HTTP API via a static port (default 8500)
bound to the host port. This enables all other pods on the node to connect
to the node-local agent using the host IP that can be retrieved via the
Kubernetes downward API. See
[accessing the Consul HTTP API](/docs/platform/k8s/run.html#accessing-the-consul-http-api)
for an example.
There is a major limitation to this: there is no way to bind to a local-only
host port. Therefore, any other node can connect to the agent. This should be
considered for security. For a properly production-secured agent with TLS
and ACLs, this is safe.
Some people prefer to run **Consul agent per pod** architectures, since this
makes it easy to register the pod as a service easily. However, this turns
a pod into a "node" in Consul and also causes an explosion of resource usage
since every pod needs a Consul agent. We recommend instead running an
agent (in a dedicated pod) per node, via the DaemonSet. This maintains the
node equivalence in Consul. Service registration should be handled via the
catalog syncing feature with Services rather than pods.
-> **Note:** Due to a limitation of anti-affinity rules with DaemonSets,
a client-mode agent runs alongide server-mode agents in Kubernetes. This
duplication wastes some resources, but otherwise functions perfectly fine.

15
website/source/layouts/docs.erb
View File

@ -296,6 +296,21 @@
</ul>
</li>
<li<%= sidebar_current("docs-platform-k8s") %>>
<a href="/docs/platform/k8s/index.html">Kubernetes</a>
<ul class="nav">
<li<%= sidebar_current("docs-platform-k8s-run") %>>
<a href="/docs/platform/k8s/run.html">Running Consul</a>
</li>
<li<%= sidebar_current("docs-platform-k8s-helm") %>>
<a href="/docs/platform/k8s/helm.html">Helm Chart</a>
</li>
<li<%= sidebar_current("docs-platform-k8s-dns") %>>
<a href="/docs/platform/k8s/dns.html">Consul DNS</a>
</li>
</ul>
</li>
<hr>
<li<%= sidebar_current("docs-guides") %>>