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.
* Add function to wait for serfHealth in api tests
* Disable connect when creating semaphore test clients
* Wait for serfHealth when creating sessions in their tests
* Add helper functions to create lock/semaphore sessions without checks
* Log passing tests to prevent timeout in Travis due to lack of output
- 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
Improvements:
- More modular
- Building within docker doesn’t use volumes so can be run on a remote docker host
- Build containers include only minimal context so they only rarely need to be rebuilt and most of the time can be used from the cache.
- 3 build containers instead of 1. One based off of the upstream golang containers for building go stuff with all our required GOTOOLS installed. One like the old container based off ubuntu bionic for building the old UI (didn’t bother creating a much better container as this shouldn’t be needed once we completely remove the legacy UI). One for building the new UI. Its alpine based with all the node, ember, yarn stuff installed.
- Top level makefile has the ability to do a container based build without running make dist
- Can build for arbitrary platforms at the top level using: make consul-docker XC_OS=… XC_ARCH=…
- overridable functionality to allow for customizations to the enterprise build (like to generate multiple binaries)
- unified how we compile our go. always use gox even for dev-builds or rather always use the tooling around our scripts which will make sure things get copied to the correct places throughout the filesystem.
* Move settings to use the same service/route API as the rest of the app
* Put some ideas down for unit testing on adapters
* Favour `Model` over `Entity`
* Move away from using `reopen` to using Mixins
* Amend messages, comment/document some usage
* Make sure the returns are consistent in normalizePayload, also
Add some todo's in to remind me to think consider this further at a
later date. For example, is normalizePayload to be a hook or an
overridable method
* Start stripping back the HTML to semantics
* Use a variable rather than chaining
* Remove unused helpers
* Start picking through the new designs, start with listing pages
* First draft HTML for every page
* Making progress on the CSS
* Keep plugging away at the catalog css
* Looking at scrolling
* Wire up filtering
* Sort out filter counting, more or less done a few outstanding
* Start knocking the forms into shape
* Add in codemirror
* Keep moving forwards with the form like layouts
* Start looking at ACL editing page, add footer in
* Pull the filters back in, look at an autoresizer for scroll views
* First draft toggles
* 2nd draft healthcheck icons
* Tweak node healthcheck icons
* Looking at healthcheck detail icons
* Tweak the filter-bar and add selections to the in content tabs
* Add ACL create, pill-like acl type highlight
* Tweaking the main nav some more
* Working on the filter-bar and freetext-filter
* Masonry layout
* Stick with `checks` instead of healthy/unhealthy
* Fix up the filter numbers/counts
* Use the thead for a measure
* First draft tomography back in
* First draft DC dropdown
* Add a temporary create buttong to kv's
* Move KV and ACL to use a create page
* Move tags
* Run through old tests
* Injectable server
* Start adding test attributes
* Add some page objects
* More test attributes and pages
* Acl filter objects
* Add a page.. page object
* Clickable items in lists
* Add rest/spread babel plugin, remove mirage for now
* Add fix for ember-collection
* Keep track of acl filters
* ember-cli-page-object
* ember-test-selectors
* ui: update version of ui compile deps
* Update static assets
* Centralize radiogroup helper
* Rejig KV's and begin to clean it up
* Work around lack of Tags for the moment..
* Some little css tweaks and start to remove possibles
* Working on the dc page and incidentals
1. Sort the datacenter-picker list
2. Add a selected state to the datacenter-picker
3. Make dc an {Name: dc}
4. Add an env helper to get to 'env vars' from within templates
* Click outside stuff for the datacenter-picker, is-active on nav
* Make sure the dropdown CTA can be active
* Bump ember add pluralize helper
* Little try at sass based custom queries
* Rejig tablular collection so it deals with resizing, actions
1. WIP: start building actions dropdowns
2. Move tabular collection to deal with resizing to rule out differences
* First draft actions dropdowns
* Add ports, selectable IP's
* Flash messages, plus general cleanup/consistency
1. Add ember-cli-flash for flash messages
2. Move everything to get() instead of item.get
3. Spotted a few things that weren't consistent
* DOn't go lower than zero
* First draft vertical menu
* Missed a get, tweak dropmenu tick
* Big cleanup
1. this.get(), this.set() > get(), set()
2. assign > {...{}, ...{}}
3. Seperator > separator
* WIP: settings
* Moved things into a ui-v2 folder
* Decide on a way to do the settings page whilst maintaining the url + dc's
* Start some error pages
* Remove base64 polyfill
* Tie in settings, fix atob bug, tweak layout css
* Centralize confirmations into a component
* Allow switching between the old and new UI with the CONSUL_UI_BETA env var
Currently all the assets are packaged into a single AssetFS and a prefix is configured to switch between the two.
* Attempt at some updates to integrate the v2 ui build into the main infrastructure
* Add redirect to index.html for unknown paths
* Allow redictor to /index.html for new ui when using -ui-dir
* Take ACLs to the correct place on save
* First pass breadcrumbs
* Remove datacenter selector on the index page
* Tweak overall layout
* Make buttons 'resets'
* Tweak last DC stuff
* Validations plus kv keyname viewing tweaks
* Pull sessions back in
* Tweak the env vars to be more reusable
* Move isAnon to the view
* No items and disabled acl css
* ACL and KV details
1. Unauthorized page
2. Make sure the ACL is always selected when it needs it
3. Check record deletion with a changeset
* Few more acl tweaks/corrections
* Add no items view to node > services
* Tags for node > services
* Make sure we have tags
* Fix up the labels on the tomography graph
* Add node link (agent) to kv sessions
* Duplicate up `create` for KV 'root creation'
* Safety check for health checks
* Fix up the grids
* Truncate td a's, fix kv columns
* Watch for spaces in KV id's
* Move actions to their own mixins for now at least
* Link reset to settings incase I want to type it in
* Tweak error page
* Cleanup healthcheck icons in service listing
* Centralize errors and make getting back easier
* Nice numbers
* Compact buttons
* Some incidental css cleanups
* Use 'Key / Value' for root
* Tweak tomography layout
* Fix single healthcheck unhealthy resource
* Get loading screen ready
* Fix healthy healthcheck tick
* Everything in header starts white
* First draft loader
* Refactor the entire backend to use proper unique keys, plus..
1. Make unique keys form dc + slug (uid)
2. Fun with errors...
* Tweak header colors
* Add noopener noreferrer to external links
* Add supers to setupController
* Implement cloning, using ember-data...
* Move the more expensive down the switch order
* First draft empty record cleanup..
* Add the cusomt store test
* Temporarily use the htmlSafe prototype to remove the console warning
* Encode hashes in urls
* Go back to using title for errors for now
* Start removing unused bulma
* Lint
* WIP: Start looking at failing tests
* Remove single redirect test
* Finish off error message styling
* Add full ember-data cache invalidation to avoid stale data...
* Add uncolorable warning icons
* More info icon
* Rearrange single service, plus tag printing
* Logo
* No quotes
* Add a simple startup logo
* Tweak healthcheck statuses
* Fix border-color for healthchecks
* Tweak node tabs
* Catch 401 ACL errors and rethrow with the provided error message
* Remove old acl unauth and error routes
* Missed a super
* Make 'All' refer to number of checks, not services
* Remove ember-resizer, add autoprefixer
* Don't show tomography if its not worth it, viewify it more also
* Little model cleanup
* Chevrons
* Find a way to reliably set the class of html from the view
* Consistent html
* Make sure session id's are visible as long as possible
* Fix single service check count
* Add filters and searchs to the query string
* Don't remember the selected tab
* Change text
* Eror tweaking
* Use chevrons on all breadcrumbs even in kv's
* Clean up a file
* Tweak some messaging
* Makesure the footer overlays whats in the page
* Tweak KV errors
* Move json toggle over to the right
* feedback-dialog along with copy buttons
* Better confirmation dialogs
* Add git sha comment
* Same title as old UI
* Allow defaults
* Make sure value is a string
* WIP: Scrolling dropdowns/confirmations
* Add to kv's
* Remove set
* First pass trace
* Better table rows
* Pull over the hashi code editor styles
* Editor tweaks
* Responsive tabs
* Add number formatting to tomography
* Review whats left todo
* Lint
* Add a coordinate ember data triplet
* Bump in a v2.0.0
* Update old tests
* Get coverage working again
* Make sure query keys are also encoded
* Don't test console.error
* Unit test some more utils
* Tweak the size of the tabular collections
* Clean up gitignore
* Fix copy button rollovers
* Get healthcheck 'icon icons' onto the text baseline
* Tweak healthcheck padding and alignment
* Make sure commas kick in in rtt, probably never get to that
* Improve vertical menu
* Tweak dropdown active state to not have a bg
* Tweak paddings
* Search entire string not just 'startsWith'
* Button states
* Most buttons have 1px border
* More button tweaks
* You can only view kv folders
* CSS cleanup reduction
* Form input states and little cleanup
* More CSS reduction
* Sort checks by importance
* Fix click outside on datacenter picker
* Make sure table th's also auto calculate properly
* Make sure `json` isn't remembered in KV editing
* Fix recursive deletion in KV's
* Centralize size
* Catch updateRecord
* Don't double envode
* model > item consistency
* Action loading and ACL tweaks
* Add settings dependencies to acl tests
* Better loading
* utf-8 base64 encode/decode
* Don't hang off a prototype for htmlSafe
* Missing base64 files...
* Get atob/btoa polyfill right
* Shadowy rollovers
* Disabled button styling for primaries
* autofocuses only onload for now
* Fix footer centering
* Beginning of 'notices'
* Remove the isLocked disabling as we are letting you do what the API does
* Don't forget the documentation link for sessions
* Updates are more likely
* Use exported constant
* Dont export redirectFS and a few other PR updates
* Remove the old bootstrap config which was used for the old UI skin
* Use curlies for multiple properties
This was reported in #3868. We make a HashiCorp hard fork of the
jteeuwen/go-bindata hard fork that was replaced and diffed the code
against a Dec 1, 2015 copy of the original repository we had as a
cross-check of that hard fork.
This replaces references to jteeuwen/go-bindata to point to the
HashiCorp fork.
Fixes#3868
This patch removes the porter tool which hands out free ports from a
given range with a library which does the same thing. The challenge for
acquiring free ports in concurrent go test runs is that go packages are
tested concurrently and run in separate processes. There has to be some
inter-process synchronization in preventing processes allocating the
same ports.
freeport allocates blocks of ports from a range expected to be not in
heavy use and implements a system-wide mutex by binding to the first
port of that block for the lifetime of the application. Ports are then
provided sequentially from that block and are tested on localhost before
being returned as available.
* new config parser for agent
This patch implements a new config parser for the consul agent which
makes the following changes to the previous implementation:
* add HCL support
* all configuration fragments in tests and for default config are
expressed as HCL fragments
* HCL fragments can be provided on the command line so that they
can eventually replace the command line flags.
* HCL/JSON fragments are parsed into a temporary Config structure
which can be merged using reflection (all values are pointers).
The existing merge logic of overwrite for values and append
for slices has been preserved.
* A single builder process generates a typed runtime configuration
for the agent.
The new implementation is more strict and fails in the builder process
if no valid runtime configuration can be generated. Therefore,
additional validations in other parts of the code should be removed.
The builder also pre-computes all required network addresses so that no
address/port magic should be required where the configuration is used
and should therefore be removed.
* Upgrade github.com/hashicorp/hcl to support int64
* improve error messages
* fix directory permission test
* Fix rtt test
* Fix ForceLeave test
* Skip performance test for now until we know what to do
* Update github.com/hashicorp/memberlist to update log prefix
* Make memberlist use the default logger
* improve config error handling
* do not fail on non-existing data-dir
* experiment with non-uniform timeouts to get a handle on stalled leader elections
* Run tests for packages separately to eliminate the spurious port conflicts
* refactor private address detection and unify approach for ipv4 and ipv6.
Fixes#2825
* do not allow unix sockets for DNS
* improve bind and advertise addr error handling
* go through builder using test coverage
* minimal update to the docs
* more coverage tests fixed
* more tests
* fix makefile
* cleanup
* fix port conflicts with external port server 'porter'
* stop test server on error
* do not run api test that change global ENV concurrently with the other tests
* Run remaining api tests concurrently
* no need for retry with the port number service
* monkey patch race condition in go-sockaddr until we understand why that fails
* monkey patch hcl decoder race condidtion until we understand why that fails
* monkey patch spurious errors in strings.EqualFold from here
* add test for hcl decoder race condition. Run with go test -parallel 128
* Increase timeout again
* cleanup
* don't log port allocations by default
* use base command arg parsing to format help output properly
* handle -dc deprecation case in Build
* switch autopilot.max_trailing_logs to int
* remove duplicate test case
* remove unused methods
* remove comments about flag/config value inconsistencies
* switch got and want around since the error message was misleading.
* Removes a stray debug log.
* Removes a stray newline in imports.
* Fixes TestACL_Version8.
* Runs go fmt.
* Adds a default case for unknown address types.
* Reoders and reformats some imports.
* Adds some comments and fixes typos.
* Reorders imports.
* add unix socket support for dns later
* drop all deprecated flags and arguments
* fix wrong field name
* remove stray node-id file
* drop unnecessary patch section in test
* drop duplicate test
* add test for LeaveOnTerm and SkipLeaveOnInt in client mode
* drop "bla" and add clarifying comment for the test
* split up tests to support enterprise/non-enterprise tests
* drop raft multiplier and derive values during build phase
* sanitize runtime config reflectively and add test
* detect invalid config fields
* fix tests with invalid config fields
* use different values for wan sanitiziation test
* drop recursor in favor of recursors
* allow dns_config.udp_answer_limit to be zero
* make sure tests run on machines with multiple ips
* Fix failing tests in a few more places by providing a bind address in the test
* Gets rid of skipped TestAgent_CheckPerformanceSettings and adds case for builder.
* Add porter to server_test.go to make tests there less flaky
* go fmt
Build all packages together with a global timeout. Locally,
the tests pass within 40 sec. On travis, we complete within
3-4 min.
travis truncates the logs after 4MB which we are hitting so
show the relevant information first and then whatever is left.
* capture all output to test.log only and print only failures
* always print output on travis
* test github.com/hashicorp/consul/consul separately until we've
found the interactions with the other packages during testing.
This patch runs 'go install' instead of gox which runs 'go build' for
'make dev' and copies the binary into ./bin and ./pkg/${GOOS}_${GOARCH}
to mimick the previous behavior. This reduces the roundtrip times
for a dev build from 11 sec to 500ms if there weren't any changes.
Matt Keeler