docs/ucp/high-availability/replicate-cas.md

103 lines
3.4 KiB
Markdown

---
description: Docker Universal Control plane has support for high availability. Learn
how to set up your installation to ensure it tolerates failures.
keywords:
- replica, controller, availability, high, ucp
menu:
main:
identifier: ucp_replicate_cas
parent: mn_ucp_high_availability
weight: 10
title: Replicate CAs for high availability
---
# Replicate CAs for high availability
Internally, UCP uses two Certificate Authorities:
* `ucp-cluster-root-ca` signs the certificates that are used by new nodes
joining the cluster. This CA also generates the certificates for admin client
certificate bundles.
* `ucp-client-root-ca` signs non-admin client certificate bundles.
When configuring UCP for high-availability, you need to ensure the CAs running
on each UCP controller node are interchangeable. This is done by using the same
root certificates and keys for every CA on the cluster.
## Replicating the CAs across controller nodes
There are two ways to ensure the same root certificates and keys are used on
every controller node:
* When joining a controller node:
When you install the first controller node, root certificates and keys are
automatically generated.
After installing the first controller node, you can backup the certificates
and keys used by that controller using.
Then, when joining new controller nodes to the cluster, you use the use
the backup you've created to make the CAs in both nodes use the same root
certificates and keys.
* After joining all controller nodes:
Alternatively, you can replicate root certificates and keys through all
controller nodes after you've joined them to the cluster.
After the installation, you backup the root certificates and keys of the
controller that had the valid CA material and restore all other controllers
with that backup archive, one by one.
## Backup the certificates and keys
To create a backup of the root certificates and keys used by the CAs, use the
backup command. Notice that this command temporarily stops the UCP CA
containers, so you should use it outside business peak hours.
Log into the node using ssh, and run:
```bash
$ docker run --rm -i --name ucp \
-v /var/run/docker.sock:/var/run/docker.sock \
docker/ucp backup --root-ca-only --interactive \
--passphrase "secret" > /tmp/backup.tar
```
Where:
* `root-ca-only` specifies to only backup the CA certificates and keys.
* `interactive` makes the command prompt for any information it needs.
* `passphrase` encrypts the backup with a given passphrase.
* `> backup.tar` streams the backup output to a file.
## Restore the certificate and keys
Once you have a backup archive of the certificates and keys used by the CAs
of a controller node, you can use it to make CAs in other controller nodes
use the same certificate and private key.
Log into the node using ssh, and run:
```bash
$ docker run --rm -i --name ucp \
-v /var/run/docker.sock:/var/run/docker.sock \
docker/ucp restore --root-ca-only --interactive \
--passphrase "secret" < /tmp/backup.tar
```
Where:
* `root-ca-only` specifies to only restore the volumes used by the CAs.
* `interactive` makes the command prompt for any information it needs.
* `passphrase` specifies the passphrase to decrypt the backup archive.
* `< backup.tar`, reads input from the backup.tar file.
## Where to go next
* [Set up high availability](set-up-high-availability.md)
* [Backups and disaster recovery](backups-and-disaster-recovery.md)