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

Website: GH-781 and cleanup for docs/internals/acl.html.

This commit is contained in:
Ryan Breen 2015-03-28 11:26:26 -04:00
parent f5d3e5653f
commit 9d71844965
1 changed files with 74 additions and 62 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

136
website/source/docs/internals/acl.html.markdown
View File

@ -9,78 +9,84 @@ description: |-
# ACL System
Consul provides an optional Access Control List (ACL) system which can be used to control
access to data and APIs. The ACL system is a
[Capability-based system](http://en.wikipedia.org/wiki/Capability-based_security) that relies
access to data and APIs. The ACL is
[Capability-based](http://en.wikipedia.org/wiki/Capability-based_security), relying
on tokens to which fine grained rules can be applied. It is very similar to
[AWS IAM](http://aws.amazon.com/iam/) in many ways.
## Scope
When the ACL system was launched in Consul 0.4, it was only possible to specify
policies for the KV store. In Consul 0.5, ACL policies were extended to service
registrations.
## ACL Design
The ACL system is designed to be easy to use, fast to enforce, flexible to new
policies, all while providing administrative insight. It has been modeled on
the AWS IAM system, as well as the more general object-capability model. The system
is modeled around "tokens".
The ACL system is designed to be easy to use, fast to enforce, and flexible to new
policies, all while providing administrative insight.
Every token has an ID, name, type and rule set. The ID is a randomly generated
UUID, making it unfeasible to guess. The name is opaque and human readable.
The type is either "client" meaning it cannot modify ACL rules, and is restricted
by the provided rules, or is "management" and is allowed to perform all actions.
Every token has an ID, name, type, and rule set. The ID is a randomly generated
UUID, making it unfeasible to guess. The name is opaque to Consul and human readable.
The type is either "client" (meaning the token cannot modify ACL rules) or "management"
(meaning the token is allowed to perform all actions).
The token ID is passed along with each RPC request to the servers. Agents
can be configured with an [`acl_token`](/docs/agent/options.html#acl_token) property
to provide a default token, but the token can also be specified by a client on a
[per-request basis](/docs/agent/http.html). ACLs are new as of Consul 0.4, meaning
[per-request basis](/docs/agent/http.html). ACLs were added in Consul 0.4, meaning
prior versions do not provide a token. This is handled by the special "anonymous"
token. Anytime there is no token provided, the rules defined by that token are
token. If no token provided, the rules associated with the anonymous token are
automatically applied. This allows policy to be enforced on legacy clients.
Enforcement is always done by the server nodes. All servers must be configured
to provide an [`acl_datacenter`](/docs/agent/options.html#acl_datacenter) which
enables ACL enforcement but also specifies the authoritative datacenter. Consul does not
replicate data cross-WAN, and instead relies on [RPC forwarding](/docs/internal/architecture.html)
to support Multi-Datacenter configurations. However, because requests can be
made across datacenter boundaries, ACL tokens must be valid globally. To avoid
replicate data cross-WAN and instead relies on [RPC forwarding](/docs/internal/architecture.html)
to support Multi-Datacenter configurations. However, because requests can be made
across datacenter boundaries, ACL tokens must be valid globally. To avoid
replication issues, a single datacenter is considered authoritative and stores
all the tokens.
When a request is made to any non-authoritative server with a token, it must
be resolved into the appropriate policy. This is done by reading the token
from the authoritative server and caching a configurable
When a request is made to a server in a non-authoritative datacenter server, it
must be resolved into the appropriate policy. This is done by reading the token
from the authoritative server and caching the result for a configurable
[`acl_ttl`](/docs/agent/options.html#acl_ttl). The implication
of caching is that the cache TTL is an upper bound on the staleness of policy
that is enforced. It is possible to set a zero TTL, but this has adverse
performance impacts, as every request requires refreshing the policy.
performance impacts, as every request requires refreshing the policy via a
cross-datacenter WAN call.
Another possible issue is an outage of the
The Consul ACL center is also built to accommodate for an outage of the
[`acl_datacenter`](/docs/agent/options.html#acl_datacenter) or networking
issues preventing access. In this case, it may be impossible for non-authoritative
servers to resolve tokens. Consul provides a number of configurable
[`acl_down_policy`](/docs/agent/options.html#acl_down_policy)
choices to tune behavior. It is possible to deny or permit all actions, or to ignore
cache TTLs and enter a fail-safe mode.
issues preventing access to it. In this case, it may be impossible for
servers in non-authoritative datacenters to resolve tokens. Consul provides
a number of configurable [`acl_down_policy`](/docs/agent/options.html#acl_down_policy)
choices to tune behavior. It is possible to deny or permit all actions or to ignore
cache TTLs and enter a fail-safe mode. The default is to ignore cache TTLs
for any previously resolved tokens and to deny any uncached tokens.
ACLs can also act in either a whitelist or blacklist mode depending
on the configuration of
[`acl_default_policy`](/docs/agent/options.html#acl_default_policy). If the default
policy is to deny all actions, then token rules can be set to allow or whitelist
actions. In the inverse, the allow all default behavior is a blacklist,
where rules are used to prohibit actions.
actions. In the inverse, the allow all default behavior is a blacklist
where rules are used to prohibit actions. By default, Consul will allow all
actions.
### Blacklist mode and `consul exec`
If you set [`acl_default_policy`](/docs/agent/options.html#acl_default_policy)
to `deny`, the `anonymous` token won't have the permission to read the default
`_rexec` prefix, and therefore token-less Consul agents (using the `anonymous` token)
to `deny`, the `anonymous` token won't have permission to read the default
`_rexec` prefix; therefore, Consul agents using the `anonymous` token
won't be able to perform [`consul exec`](/docs/commands/exec.html) actions.
There is a subtle interaction there. The agents will need permission to
read/write to the `_rexec` prefix for [`consul exec`](/docs/commands/exec.html) to
work properly. They use that as the transport for most data, only the edge trigger
uses the event system.
Here's why: the agents need read/write permission to the `_rexec` prefix for
[`consul exec`](/docs/commands/exec.html) to work properly. They use that prefix
as the transport for most data.
You can do this by allowing the `anonymous` token to access that prefix, or by
providing tokens to the agents that enable it. The former can be done by giving
this rule to the `anonymous` token:
You can enable [`consul exec`](/docs/commands/exec.html) from agents that are not
configured with a token by allowing the `anonymous` token to access that prefix.
This can be done by giving this rule to the `anonymous` token:
```javascript
key "_rexec/" {
@ -88,10 +94,14 @@ key "_rexec/" {
}
```
Alternatively, you can, of course, add an explicit
[`acl_token`](/docs/agent/options.html#acl_token) to each agent, giving it access
to that prefix.
### Bootstrapping ACLs
Bootstrapping the ACL system is done by providing an initial [`acl_master_token`
configuration](/docs/agent/options.html#acl_master_token), which will be created
configuration](/docs/agent/options.html#acl_master_token) which will be created
as a "management" type token if it does not exist. Note that the [`acl_master_token`
](/docs/agent/options.html#acl_master_token) is only installed when a server acquires
cluster leadership. If you would like to install or change the
@ -101,14 +111,33 @@ for all servers. Once this is done, restart the current leader to force a leader
## Rule Specification
A core part of the ACL system is a rule language which is used
to describe the policy that must be enforced. We make use of
A core part of the ACL system is a rule language which is used to describe the policy
that must be enforced. Consul supports ACLs for both [K/Vs](/intro/getting-started/kv.html)
and [services](/intro/getting-started/services.html).
Key policies are defined by coupling a prefix with a policy. The rules are enforced
using a longest-prefix match policy; Consul picks the most specific policy possible. The
policy is either "read", "write" or "deny". A "write" policy implies "read", and there is no
way to specify write-only. If there is no applicable rule, the
[`acl_default_policy`](/docs/agent/options.html#acl_default_policy) is applied.
Service policies are defined by coupling a service name and a policy. The rules are
enforced using an exact match policy. The default rule, applied to any service that doesn't
have a matching policy, is provided using the empty string. The policy is either "read",
"write", or "deny". A "write" policy implies "read", and there is no way to specify write-only.
If there is no applicable rule, the
[`acl_default_policy`](/docs/agent/options.html#acl_default_policy) is
applied. Currently, only the "write" level is enforced for registration of
services; services can always be read.
The policy for the "consul" service is always "write" as it is managed internally by Consul.
We make use of
the [HashiCorp Configuration Language (HCL)](https://github.com/hashicorp/hcl/)
to specify policy. This language is human readable and interoperable
with JSON making it easy to machine generate.
with JSON making it easy to machine-generate.
As of Consul 0.4, it is only possible to specify policies for the
KV store. Specification in the HCL format looks like:
Specification in the HCL format looks like:
```javascript
# Default all keys to read-only
@ -119,17 +148,17 @@ key "foo/" {
policy = "write"
}
key "foo/private/" {
# Deny access to the private dir
# Deny access to the dir "foo/private"
policy = "deny"
}
# Default all services to allowing registration
# Default all services to allow registration
service "" {
policy = "write"
}
service "secure" {
# Deny registration access to secure service
# Deny registration access to service named "secure"
policy = "read"
}
```
@ -160,20 +189,3 @@ This is equivalent to the following JSON input:
}
```
Key policies provide both a prefix and a policy. The rules are enforced
using a longest-prefix match policy. This means we pick the most specific
policy possible. The policy is either "read", "write" or "deny". A "write"
policy implies "read", and there is no way to specify write-only. If there
is no applicable rule, the
[`acl_default_policy`](/docs/agent/options.html#acl_default_policy) is applied.
Services policies provide both a service name and a policy. The rules are
enforced using an exact match policy. The default rule is provided using
the empty string. The policy is either "read", "write", or "deny". A "write"
policy implies "read", and there is no way to specify write-only. If there
is no applicable rule, the
[`acl_default_policy`](/docs/agent/options.html#acl_default_policy) is
applied. Currently, only the "write" level is enforced for registration of
services. The policy for the "consul" service is always "write" as it is
managed internally.