The observed bug was that a full restart of a consul datacenter (servers
and clients) in conjunction with a restart of a connect-flavored
application with bring-your-own-service-registration logic would very
frequently cause the envoy sidecar service check to never reflect the
aliased service.
Over the course of investigation several bugs and unfortunate
interactions were corrected:
(1)
local.CheckState objects were only shallow copied, but the key piece of
data that gets read and updated is one of the things not copied (the
underlying Check with a Status field). When the stock code was run with
the race detector enabled this highly-relevant-to-the-test-scenario field
was found to be racy.
Changes:
a) update the existing Clone method to include the Check field
b) copy-on-write when those fields need to change rather than
incrementally updating them in place.
This made the observed behavior occur slightly less often.
(2)
If anything about how the runLocal method for node-local alias check
logic was ever flawed, there was no fallback option. Those checks are
purely edge-triggered and failure to properly notice a single edge
transition would leave the alias check incorrect until the next flap of
the aliased check.
The change was to introduce a fallback timer to act as a control loop to
double check the alias check matches the aliased check every minute
(borrowing the duration from the non-local alias check logic body).
This made the observed behavior eventually go away when it did occur.
(3)
Originally I thought there were two main actions involved in the data race:
A. The act of adding the original check (from disk recovery) and its
first health evaluation.
B. The act of the HTTP API requests coming in and resetting the local
state when re-registering the same services and checks.
It took awhile for me to realize that there's a third action at work:
C. The goroutines associated with the original check and the later
checks.
The actual sequence of actions that was causing the bad behavior was
that the API actions result in the original check to be removed and
re-added _without waiting for the original goroutine to terminate_. This
means for brief windows of time during check definition edits there are
two goroutines that can be sending updates for the alias check status.
In extremely unlikely scenarios the original goroutine sees the aliased
check start up in `critical` before being removed but does not get the
notification about the nearly immediate update of that check to
`passing`.
This is interlaced wit the new goroutine coming up, initializing its
base case to `passing` from the current state and then listening for new
notifications of edge triggers.
If the original goroutine "finishes" its update, it then commits one
more write into the local state of `critical` and exits leaving the
alias check no longer reflecting the underlying check.
The correction here is to enforce that the old goroutines must terminate
before spawning the new one for alias checks.
* First conversion
* Use serf 0.8.2 tag and associated updated deps
* * Move freeport and testutil into internal/
* Make internal/ its own module
* Update imports
* Add replace statements so API and normal Consul code are
self-referencing for ease of development
* Adapt to newer goe/values
* Bump to new cleanhttp
* Fix ban nonprintable chars test
* Update lock bad args test
The error message when the duration cannot be parsed changed in Go 1.12
(ae0c435877d3aacb9af5e706c40f9dddde5d3e67). This updates that test.
* Update another test as well
* Bump travis
* Bump circleci
* Bump go-discover and godo to get rid of launchpad dep
* Bump dockerfile go version
* fix tar command
* Bump go-cleanhttp
Prevent race between register and deregister requests by saving them
together in the local state on registration.
Also adds more cleaning in case of failure when registering services
/ checks.
This PR adds two features which will be useful for operators when ACLs are in use.
1. Tokens set in configuration files are now reloadable.
2. If `acl.enable_token_persistence` is set to `true` in the configuration, tokens set via the `v1/agent/token` endpoint are now persisted to disk and loaded when the agent starts (or during configuration reload)
Note that token persistence is opt-in so our users who do not want tokens on the local disk will see no change.
Some other secondary changes:
* Refactored a bunch of places where the replication token is retrieved from the token store. This token isn't just for replicating ACLs and now it is named accordingly.
* Allowed better paths in the `v1/agent/token/` API. Instead of paths like: `v1/agent/token/acl_replication_token` the path can now be just `v1/agent/token/replication`. The old paths remain to be valid.
* Added a couple new API functions to set tokens via the new paths. Deprecated the old ones and pointed to the new names. The names are also generally better and don't imply that what you are setting is for ACLs but rather are setting ACL tokens. There is a minor semantic difference there especially for the replication token as again, its no longer used only for ACL token/policy replication. The new functions will detect 404s and fallback to using the older token paths when talking to pre-1.4.3 agents.
* Docs updated to reflect the API additions and to show using the new endpoints.
* Updated the ACL CLI set-agent-tokens command to use the non-deprecated APIs.
This way we can avoid unnecessary panics which cause other tests not to run.
This doesn't remove all the possibilities for panics causing other tests not to run, it just fixes the TestAgent
Fixes: https://github.com/hashicorp/consul/issues/3676
This fixes a bug were registering an agent with a non-existent ACL token can prevent other
services registered with a good token from being synced to the server when using
`acl_enforce_version_8 = false`.
## Background
When `acl_enforce_version_8` is off the agent does not check the ACL token validity before
storing the service in its state.
When syncing a service registered with a missing ACL token we fall into the default error
handling case (https://github.com/hashicorp/consul/blob/master/agent/local/state.go#L1255)
and stop the sync (https://github.com/hashicorp/consul/blob/master/agent/local/state.go#L1082)
without setting its Synced property to true like in the permission denied case.
This means that the sync will always stop at the faulty service(s).
The order in which the services are synced is random since we iterate on a map. So eventually
all services with good ACL tokens will be synced, this can however take some time and is influenced
by the cluster size, the bigger the slower because retries are less frequent.
Having a service in this state also prevent all further sync of checks as they are done after
the services.
## Changes
This change modify the sync process to continue even if there is an error.
This fixes the issue described above as well as making the sync more error tolerant: if the server repeatedly refuses
a service (the ACL token could have been deleted by the time the service is synced, the servers
were upgraded to a newer version that has more strict checks on the service definition...).
Then all services and check that can be synced will, and those that don't will be marked as errors in
the logs instead of blocking the whole process.
This PR is almost a complete rewrite of the ACL system within Consul. It brings the features more in line with other HashiCorp products. Obviously there is quite a bit left to do here but most of it is related docs, testing and finishing the last few commands in the CLI. I will update the PR description and check off the todos as I finish them over the next few days/week.
Description
At a high level this PR is mainly to split ACL tokens from Policies and to split the concepts of Authorization from Identities. A lot of this PR is mostly just to support CRUD operations on ACLTokens and ACLPolicies. These in and of themselves are not particularly interesting. The bigger conceptual changes are in how tokens get resolved, how backwards compatibility is handled and the separation of policy from identity which could lead the way to allowing for alternative identity providers.
On the surface and with a new cluster the ACL system will look very similar to that of Nomads. Both have tokens and policies. Both have local tokens. The ACL management APIs for both are very similar. I even ripped off Nomad's ACL bootstrap resetting procedure. There are a few key differences though.
Nomad requires token and policy replication where Consul only requires policy replication with token replication being opt-in. In Consul local tokens only work with token replication being enabled though.
All policies in Nomad are globally applicable. In Consul all policies are stored and replicated globally but can be scoped to a subset of the datacenters. This allows for more granular access management.
Unlike Nomad, Consul has legacy baggage in the form of the original ACL system. The ramifications of this are:
A server running the new system must still support other clients using the legacy system.
A client running the new system must be able to use the legacy RPCs when the servers in its datacenter are running the legacy system.
The primary ACL DC's servers running in legacy mode needs to be a gate that keeps everything else in the entire multi-DC cluster running in legacy mode.
So not only does this PR implement the new ACL system but has a legacy mode built in for when the cluster isn't ready for new ACLs. Also detecting that new ACLs can be used is automatic and requires no configuration on the part of administrators. This process is detailed more in the "Transitioning from Legacy to New ACL Mode" section below.
* Proxy Config Manager
This component watches for local state changes on the agent and ensures that each service registered locally with Kind == connect-proxy has it's state being actively populated in the cache.
This serves two purposes:
1. For the built-in proxy, it ensures that the state needed to accept connections is available in RAM shortly after registration and likely before the proxy actually starts accepting traffic.
2. For (future - next PR) xDS server and other possible future proxies that require _push_ based config discovery, this provides a mechanism to subscribe and be notified about updates to a proxy instance's config including upstream service discovery results.
* Address review comments
* Better comments; Better delivery of latest snapshot for slow watchers; Embed Config
* Comment typos
* Add upstream Stringer for funsies
- A new endpoint `/v1/agent/service/:service_id` which is a generic way to look up the service for a single instance. The primary value here is that it:
- **supports hash-based blocking** and so;
- **replaces `/agent/connect/proxy/:proxy_id`** as the mechanism the built-in proxy uses to read its config.
- It's not proxy specific and so works for any service.
- It has a temporary shim to call through to the existing endpoint to preserve current managed proxy config defaulting behaviour until that is removed entirely (tested).
- The built-in proxy now uses the new endpoint exclusively for it's config
- The built-in proxy now has a `-sidecar-for` flag that allows the service ID of the _target_ service to be specified, on the condition that there is exactly one "sidecar" proxy (that is one that has `Proxy.DestinationServiceID` set) for the service registered.
- Several fixes for edge cases for SidecarService
- A fix for `Alias` checks - when running locally they didn't update their state until some external thing updated the target. If the target service has no checks registered as below, then the alias never made it past critical.
* Refactor Service Definition ProxyDestination.
This includes:
- Refactoring all internal structs used
- Updated tests for both deprecated and new input for:
- Agent Services endpoint response
- Agent Service endpoint response
- Agent Register endpoint
- Unmanaged deprecated field
- Unmanaged new fields
- Managed deprecated upstreams
- Managed new
- Catalog Register
- Unmanaged deprecated field
- Unmanaged new fields
- Managed deprecated upstreams
- Managed new
- Catalog Services endpoint response
- Catalog Node endpoint response
- Catalog Service endpoint response
- Updated API tests for all of the above too (both deprecated and new forms of register)
TODO:
- config package changes for on-disk service definitions
- proxy config endpoint
- built-in proxy support for new fields
* Agent proxy config endpoint updated with upstreams
* Config file changes for upstreams.
* Add upstream opaque config and update all tests to ensure it works everywhere.
* Built in proxy working with new Upstreams config
* Command fixes and deprecations
* Fix key translation, upstream type defaults and a spate of other subtele bugs found with ned to end test scripts...
TODO: tests still failing on one case that needs a fix. I think it's key translation for upstreams nested in Managed proxy struct.
* Fix translated keys in API registration.
≈
* Fixes from docs
- omit some empty undocumented fields in API
- Bring back ServiceProxyDestination in Catalog responses to not break backwards compat - this was removed assuming it was only used internally.
* Documentation updates for Upstreams in service definition
* Fixes for tests broken by many refactors.
* Enable travis on f-connect branch in this branch too.
* Add consistent Deprecation comments to ProxyDestination uses
* Update version number on deprecation notices, and correct upstream datacenter field with explanation in docs
* Implementation of Weights Data structures
Adding this datastructure will allow us to resolve the
issues #1088 and #4198
This new structure defaults to values:
```
{ Passing: 1, Warning: 0 }
```
Which means, use weight of 0 for a Service in Warning State
while use Weight 1 for a Healthy Service.
Thus it remains compatible with previous Consul versions.
* Implemented weights for DNS SRV Records
* DNS properly support agents with weight support while server does not (backwards compatibility)
* Use Warning value of Weights of 1 by default
When using DNS interface with only_passing = false, all nodes
with non-Critical healthcheck used to have a weight value of 1.
While having weight.Warning = 0 as default value, this is probably
a bad idea as it breaks ascending compatibility.
Thus, we put a default value of 1 to be consistent with existing behaviour.
* Added documentation for new weight field in service description
* Better documentation about weights as suggested by @banks
* Return weight = 1 for unknown Check states as suggested by @banks
* Fixed typo (of -> or) in error message as requested by @mkeeler
* Fixed unstable unit test TestRetryJoin
* Fixed unstable tests
* Fixed wrong Fatalf format in `testrpc/wait.go`
* Added notes regarding DNS SRV lookup limitations regarding number of instances
* Documentation fixes and clarification regarding SRV records with weights as requested by @banks
* Rephrase docs
- Improve resilience of testrpc.WaitForLeader()
- Add additionall retry to CI
- Increase "go test" timeout to 8m
- Add wait for cluster leader to several tests in the agent package
- Add retry to some tests in the api and command packages
There are also a lot of small bug fixes found when testing lots of things end-to-end for the first time and some cleanup now it's integrated with real CA code.
Intention de-duplication in previously merged PR actualy failed some tests that were not caught be me or CI. I ran the test files for state changes but they happened not to trigger this case so I made sure they did first and then fixed. That fixed some upstream intention endpoint tests that I'd not run as part of testing the previous fix.
The anti-entropy tests relied on the side-effect of the StartSync()
method to perform a full sync instead of a partial sync. This lead to
multiple anti-entropy go routines being started unnecessary retry loops.
This change changes the behavior to perform synchronous full syncs when
necessary removing the need for all of the time.Sleep and most of the
retry loops.
The state of the service and health check records was spread out over
multiple maps guarded by a single lock. Access to the maps has to happen
in a coordinated effort and the tests often violated this which made
them brittle and racy.
This patch replaces the multiple maps with a single one for both checks
and services to make the code less fragile.
This is also necessary since moving the local state into its own package
creates circular dependencies for the tests. To avoid this the tests can
no longer access internal data structures which they should not be doing
in the first place.
The tests still don't compile but this is a ncessary step in that
direction.
