Add fernet token info to the README.md

This adds some details to the README for the new Fernet support and
how to configure it.

Change-Id: Ib90179583dc024c5a74fd2812240f8cf6955c1d9
This commit is contained in:
Alex Kavanagh 2018-08-29 16:17:25 +01:00
parent 8da8061687
commit c45f580211
1 changed files with 128 additions and 25 deletions

153
README.md
View File

@ -1,8 +1,8 @@
Overview
========
This charm provides Keystone, the Openstack identity service. Its target
platform is (ideally) Ubuntu LTS + Openstack.
This charm provides Keystone, the OpenStack identity service. Its target
platform is (ideally) Ubuntu LTS + OpenStack.
Usage
=====
@ -11,28 +11,27 @@ The following interfaces are provided:
- nrpe-external-master: Used to generate Nagios checks.
- identity-service: Openstack API endpoints request an entry in the
Keystone service catalog + endpoint template catalog. When a relation
is established, Keystone receives: `service_name`, `region`, `public_url`,
`admin_url` and `internal_url`. It first checks that the requested service
is listed as a supported service. This list should stay updated to
support current Openstack core services. If the service is supported,
an entry in the service catalog is created, an endpoint template is
created and a admin token is generated. The other end of the relation
receives the token as well as info on which ports Keystone is listening
on.
- identity-service: OpenStack API endpoints request an entry in the
Keystone service catalog + endpoint template catalog. When a relation is
established, Keystone receives: `service_name`, `region`, `public_url`,
`admin_url` and `internal_url`. It first checks that the requested service is
listed as a supported service. This list should stay updated to support
current OpenStack core services. If the service is supported, an entry in the
service catalog is created, an endpoint template is created and a admin token
is generated. The other end of the relation receives the token as well as
info on which ports Keystone is listening on.
- keystone-service: This is currently only used by Horizon/dashboard
as its interaction with Keystone is different from other Openstack API
services. That is, Horizon requests a Keystone role and token exists.
During a relation, Horizon requests its configured default role and
Keystone responds with a token and the auth + admin ports on which
Keystone is listening.
as its interaction with Keystone is different from other OpenStack API
services. That is, Horizon requests a Keystone role and token exists. During
a relation, Horizon requests its configured default role and Keystone
responds with a token and the auth + admin ports on which Keystone is
listening.
- identity-admin: Charms use this relation to obtain the credentials
for the admin user. This is intended for charms that automatically
provision users, tenants, etc. or that otherwise automate using the
Openstack cluster deployment.
for the admin user. This is intended for charms that automatically provision
users, tenants, etc. or that otherwise automate using the OpenStack cluster
deployment.
- identity-notifications: Used to broadcast messages to any services
listening on the interface.
@ -128,17 +127,22 @@ need to provide the `ssl_ca` configuration option.
Network Space support
---------------------
This charm supports the use of Juju Network Spaces, allowing the charm to be bound to network space configurations managed directly by Juju. This is only supported with Juju 2.0 and above.
This charm supports the use of Juju Network Spaces, allowing the charm to be
bound to network space configurations managed directly by Juju. This is only
supported with Juju 2.0 and above.
API endpoints can be bound to distinct network spaces supporting the network separation of public, internal and admin endpoints.
API endpoints can be bound to distinct network spaces supporting the network
separation of public, internal and admin endpoints.
Access to the underlying MySQL instance can also be bound to a specific space using the shared-db relation.
Access to the underlying MySQL instance can also be bound to a specific space
using the shared-db relation.
To use this feature, use the --bind option when deploying the charm:
juju deploy keystone --bind "public=public-space internal=internal-space admin=admin-space shared-db=internal-space"
Alternatively these can also be provided as part of a juju native bundle configuration:
Alternatively these can also be provided as part of a juju native bundle
configuration:
keystone:
charm: cs:xenial/keystone
@ -151,4 +155,103 @@ Alternatively these can also be provided as part of a juju native bundle configu
NOTE: Spaces must be configured in the underlying provider prior to attempting to use them.
NOTE: Existing deployments using `os\-\*-network` configuration options will continue to function; these options are preferred over any network space binding provided if set.
NOTE: Existing deployments using `os\-\*-network` configuration options will
continue to function; these options are preferred over any network space
binding provided if set.
Token Support
-------------
As the keystone-charm supports multiple releases of the OpenStack software, it
also supports two keystone token systems: UUID and Fernet. The capabilities are:
- pre 'ocata': UUID tokens only.
- ocata and pike: UUID or Fernet tokens, configured via the 'token-provider'
configuration parameter.
- rocky and later: Fernet tokens only.
Fernet tokens were added to OpenStack to solve the problem of keystone being
required to persist tokens to a common database (cluster) like UUID tokens,
and solve the problem of size for PKI or PKIZ tokens.
For further information please see [Fernet - Frequently Asked
Questions](https://docs.openstack.org/keystone/pike/admin/identity-fernet-token-faq.html).
### Theory of Operation
In order to generate tokens, Fernet keys are used. These are generated by
keystone and have an expiry date. The key repository is a directory, and each
key is an integer number, with the highest number being the primary key. Key
'0' is the staged key, that will be the next primary. Other keys are secondary
keys.
New tokens are only ever generated from the primary key, whilst they secondary
keys are used to validate existing tokens. The staging key is not used to
generate tokens, but can be used to validate tokens as the staging key might be
the new primary key on the master due to a rotation and the keys have not yet
been synchronised across all the units.
Fernet keys need to be rotated at periodic intervals, and the keys need to be
synchronised to each of the other keystone units. Keys should only be rotated
on the master keystone unit, and must be synchronised *before* they are rotated
again. *Over rotation* occurs if a unit rotates its keys such that there is
no suitable decoding key on another unit that can decode a token that has been
generated on the master. This happens if two key rotations are done on the
master before a synchronisation has been successfully performed. This should
be avoided. Over rotations can also cause validation keys to be removed
*before* a token's expiration which would result in failed validations.
There are 3 parts to the **Key Rotation Strategy**:
1. The rotation frequency
2. The token lifespan
3. The number of active keys
There needs to be at least 3 keys as a minimum. The actual number of keys is
determined by the *token lifespan* and the *rotation frequency*. The
*max_active_keys* must be one greater than the *token lifespan* / *rotation
frequency*
To quote from the [FAQ](https://docs.openstack.org/keystone/queens/admin/identity-fernet-token-faq.html):
The number of max_active_keys for a deployment can be determined by
dividing the token lifetime, in hours, by the frequency of rotation in
hours and adding two. Better illustrated as:
### Configuring key lifetime
In the keystone-charm, the _rotation frequency_ is calculated
automatically from the `token-expiration` and the `fernet-max-active-keys`
configuration parameters. For example, with an expiration of 24 hours and
6 active keys, the rotation frequency is calculated as:
```python
token_expiration = 24 # actually 3600, as it's in seconds
max_active_keys = 6
rotation_frequency = token_expiration / (max_active_keys - 2)
```
Thus, the `fernet-max-active-keys` can never be less than 3 (which is
enforced in the charm), which would make the rotation frequency the *same*
as the token expiration time.
Note: that to increase the rotation frequency, _either_ increase
`fernet-max-active-keys` or reduce `token-expiration`, and to decrease
rotation frequency, do the opposite.
Note: if the configuration parameters are used to significantly reduce the
key lifetime, then it is possible to over-rotate the verification keys
such that services will hold tokens that cannot be verified but haven't
yet expired. This should be avoided by only making small changes and
verifying that current tokens will still be able to be verified. In
particular, `fernet-max-active-keys` affects the rotation time.
### Upgrades
When an older keystone-charm is upgraded to this version, NO change will
occur to the token system. That is, an ocata system will continue to use
UUID tokens. In order to change the token system to Fernet, change the
`token-provider` configuration item to `fernet`. This will switch the
token system over. There may be a small outage in the _control plane_,
but the running instances will be unaffected.