ACL tokens are granted permissions either by direct policy links
or via ACL role links. Callers should therefore be able to read
policies directly assigned to the caller token or indirectly by
ACL role links.
This changes adds ACL role creation and deletion to the event
stream. It is exposed as a single topic with two types; the filter
is primarily the role ID but also includes the role name.
While conducting this work it was also discovered that the events
stream has its own ACL resolution logic. This did not account for
ACL tokens which included role links, or tokens with expiry times.
ACL role links are now resolved to their policies and tokens are
checked for expiry correctly.
* One-time tokens are not replicated between regions, so we don't want to enforce
that the version check across all of serf, just members in the same region.
* Scheduler: Disconnected clients handling is specific to a single region, so we
don't want to enforce that the version check across all of serf, just members in
the same region.
* Variables: enforce version check in Apply RPC
* Cleans up a bunch of legacy checks.
This changeset is specific to 1.4.x and the changes for previous versions of
Nomad will be manually backported in a separate PR.
An ACL roles name must be unique, however, a bug meant multiple
roles of the same same could be created. This fixes that problem
with checks in the RPC handler and state store.
Making the ACL Role listing return object a stub future-proofs the
endpoint. In the event the role object grows, we are not bound by
having to return all fields within the list endpoint or change the
signature of the endpoint to reduce the list return size.
ACL Roles along with policies and global token will be replicated
from the authoritative region to all federated regions. This
involves a new replication loop running on the federated leader.
Policies and roles may be replicated at different times, meaning
the policies and role references may not be present within the
local state upon replication upsert. In order to bypass the RPC
and state check, a new RPC request parameter has been added. This
is used by the replication process; all other callers will trigger
the ACL role policy validation check.
There is a new ACL RPC endpoint to allow the reading of a set of
ACL Roles which is required by the replication process and matches
ACL Policies and Tokens. A bug within the ACL Role listing RPC has
also been fixed which returned incorrect data during blocking
queries where a deletion had occurred.
ACL tokens can now utilize ACL roles in order to provide API
authorization. Each ACL token can be created and linked to an
array of policies as well as an array of ACL role links. The link
can be provided via the role name or ID, but internally, is always
resolved to the ID as this is immutable whereas the name can be
changed by operators.
When resolving an ACL token, the policies linked from an ACL role
are unpacked and combined with the policy array to form the
complete auth set for the token.
The ACL token creation endpoint handles deduplicating ACL role
links as well as ensuring they exist within state.
When reading a token, Nomad will also ensure the ACL role link is
current. This handles ACL roles being deleted from under a token
from a UX standpoint.
New ACL Role RPC endpoints have been created to allow the creation,
update, read, and deletion of ACL roles. All endpoints require a
management token; in the future readers will also be allowed to
view roles associated to their ACL token.
The create endpoint in particular is responsible for deduplicating
ACL policy links and ensuring named policies are found within
state. This is done within the RPC handler so we perform a single
loop through the links for slight efficiency.
When applying a raft log to expire ACL tokens, we need to use a
timestamp provided by the leader so that the result is deterministic
across servers. Use leader's timestamp from RPC call
The ACL token state schema has been updated to utilise two new
indexes which track expiration of tokens that are configured with
an expiration TTL or time. A new state function allows listing
ACL expired tokens which will be used by internal garbage
collection.
The ACL endpoint has been modified so that all validation happens
within a single function call. This is easier to understand and
see at a glance. The ACL token validation now also includes logic
for expiry TTL and times. The ACL endpoint upsert tests have been
condensed into a single, table driven test.
There is a new token canonicalize which provides a single place
for token canonicalization, rather than logic spread in the RPC
handler.
Fix a panic in handling one-time auth tokens, used to support `nomad ui
--authenticate`.
If the nomad leader is a 1.1.x with some servers running as 1.0.x, the
pre-1.1.0 servers risk crashing and the cluster may lose quorum. That
can happen when `nomad authenticate -ui` command is issued, or when the
leader scans for expired tokens every 10 minutes.
Fixed#10943 .
RPC endpoints for the user-driven APIs (`UpsertOneTimeToken` and
`ExchangeOneTimeToken`) and token expiration (`ExpireOneTimeTokens`).
Includes adding expiration to the periodic core GC job.
If ACL Request is unauthenticated, we should honor the anonymous token.
This PR makes few changes:
* `GetPolicy` endpoints may return policy if anonymous policy allows it,
or return permission denied otherwise.
* `ListPolicies` returns an empty policy list, or one with anonymous
policy if one exists.
Without this PR, the we return an incomprehensible error.
Before:
```
$ curl http://localhost:4646/v1/acl/policy/doesntexist; echo
acl token lookup failed: index error: UUID must be 36 characters
$ curl http://localhost:4646/v1/acl/policies; echo
acl token lookup failed: index error: UUID must be 36 characters
```
After:
```
$ curl http://localhost:4646/v1/acl/policy/doesntexist; echo
Permission denied
$ curl http://localhost:4646/v1/acl/policies; echo
[]
```
Noticed that ACL endpoints return 500 status code for user errors. This
is confusing and can lead to false monitoring alerts.
Here, I introduce a concept of RPCCoded errors to be returned by RPC
that signal a code in addition to error message. Codes for now match
HTTP codes to ease reasoning.
```
$ nomad acl bootstrap
Error bootstrapping: Unexpected response code: 500 (ACL bootstrap already done (reset index: 9))
$ nomad acl bootstrap
Error bootstrapping: Unexpected response code: 400 (ACL bootstrap already done (reset index: 9))
```
The ForceLeaveRequest endpoint may only be called on servers, but the
code was using a Client to resolve tokens. This would cause a panic when
an agent wasn't both a Server and a Client.
James Rasell