Fixes a bug whereby servers present in multiple network areas would be
properly segmented in the Router, but not in the gRPC mirror. This would
lead servers in the current datacenter leaving from a network area
(possibly during the network area's removal) from deleting their own
records that still exist in the standard WAN area.
The gRPC client stack uses the gRPC server tracker to execute all RPCs,
even those targeting members of the current datacenter (which is unlike
the net/rpc stack which has a bypass mechanism).
This would manifest as a gRPC method call never opening a socket because
it would block forever waiting for the current datacenter's pool of
servers to be non-empty.
types: add TLS constants
types: distinguish between human and Envoy serialization for TLSVersion constants
types: add DeprecatedAgentTLSVersions for backwards compatibility
types: add methods for printing TLSVersion as strings
types: add TLSVersionInvalid error value
types: add a basic test for TLSVersion comparison
types: add TLS cihper suite mapping using IANA constant names and values
types: adding ConsulAutoConfigTLSVersionStrings
changelog: add entry for TLSVersion and TLSCipherSuite types
types: initialize TLSVerison constants starting at zero
types: remove TLSVersionInvalid < 0 test
types: update note for ConsulAutoConfigTLSVersionStrings
types: programmatically invert TLSCipherSuites for HumanTLSCipherSuiteStrings lookup map
Co-authored-by: Dan Upton <daniel@floppy.co>
types: add test for TLSVersion zero-value
types: remove unused EnvoyTLSVersionStrings
types: implement MarshalJSON for TLSVersion
types: implement TLSVersionUnspecified as zero value
types: delegate TLS.MarshalJSON to json.Marshal, use ConsulConfigTLSVersionStrings as default String() values
Co-authored-by: Dan Upton <daniel@floppy.co>
The test added in this commit shows the problem. Previously the
SigningKeyID was set to the RootCert not the local leaf signing cert.
This same bug was fixed in two other places back in 2019, but this last one was
missed.
While fixing this bug I noticed I had the same few lines of code in 3
places, so I extracted a new function for them.
There would be 4 places, but currently the InitializeCA flow sets this
SigningKeyID in a different way, so I've left that alone for now.
We were not adding the local signing cert to the CARoot. This commit
fixes that bug, and also adds support for fixing existing CARoot on
upgrade.
Also update the tests for both primary and secondary to be more strict.
Check the SigningKeyID is correct after initialization and rotation.
This commit uses all our new ways of doing things to Lock Sessions and their interactions with KV and Nodes. This is mostly around are new under-the-hood things, but also I took the opportunity to upgrade some of the CSS to reuse some of our CSS utils that have been made over the past few months (%csv-list and %horizontal-kv-list).
Also added (and worked on existing) documentation for Lock Session related components.
- Moves where they appear up to the <App /> component.
- Instead of a <Notification /> wrapping component to move whatever you use for a notification up to where they need to appear (via ember-cli-flash), we now use a {{notification}} modifier now we have modifiers.
- Global notifications/flashes are no longer special styles of their own. You just use the {{notification}} modifier to hoist whatever component/element you want up to the top of the page. This means we can re-use our existing <Notice /> component for all our global UI notifications (this is the user visible change here)
* Upgrade AuthForm and document current state a little better
* Hoist SSO out of the AuthForm
* Bare minimum admin partitioned SSO
also:
ui: Tabbed Login with Token or SSO interface (#11619)
- I upgraded our super old, almost the first ember component I wrote, to use glimmer/almost template only. This should use slots/contextual components somehow, but thats a bigger upgrade so I didn't go that far.
- I've been wanting to upgrade the shape of our StateChart component for a very long while now, here its very apparent that it would be much better to do this sooner rather than later. I left it as is for now, but there will be a PR coming soon with a slight reshaping of this component.
- Added a did-upsert modifier which is a mix of did-insert/did-update
- Documentation added/amended for all the new things.
* Support vault auth methods for the Vault connect CA provider
* Rotate the token (re-authenticate to vault using auth method) when the token can no longer be renewed
For our dc, nspace and partition 'bucket' menus, sometimes when selecting one 'bucket' we need to reset a different 'bucket' back to the one that your token has by default (or the default if not). For example when switching to a different partition whilst you are in a non-default namespace of another partition, we need to switch you to the token default namespace of the partition you are switching to.
Fixes an issue where the code editor would not resizing to the full extent of the browser window plus CodeEditor restyling/refactoring
- :label named block
- :tools named block
- :content named block
- code and CSS cleanup
- CodeEditor.mdx
Signed-off-by: Alessandro De Blasis <alex@deblasis.net>
Co-authored-by: John Cowen <johncowen@users.noreply.github.com>
Most HTTP API calls will use the default namespace of the calling token to additionally filter/select the data used for the response if one is not specified by the frontend.
The internal permissions/authorize endpoint does not do this (you can ask for permissions from different namespaces in on request).
Therefore this PR adds the tokens default namespace in the frontend only to our calls to the authorize endpoint. I tried to do it in a place that made it feel like it's getting added in the backend, i.e. in a place which was least likely to ever require changing or thinking about.
Note: We are probably going to change this internal endpoint to also inspect the tokens default namespace on the backend. At which point we can revert this commit/PR.
* Add the same support for the tokens default partition
* command/redirect_traffic: add rules to redirect DNS to Consul. Currently uses a hack to get the consul dns service ip, and this hack only works when the service is deployed in the same namespace as consul.
* command/redirect_traffic: redirect DNS to Consul when -consul-dns-ip is passed in
* Add unit tests to Consul DNS IP table redirect rules
Co-authored-by: Ashwin Venkatesh <ashwin@hashicorp.com>
Co-authored-by: Iryna Shustava <ishustava@users.noreply.github.com>
Temporarily revert to pre-1.10 UI functionality by overwriting frontend
permissions. These are used to hide certain UI elements, but they are
still enforced on the backend.
This temporary measure should be removed again once https://github.com/hashicorp/consul/issues/11098
has been resolved
The duo of `makeUpstreamFilterChainForDiscoveryChain` and `makeListenerForDiscoveryChain` were really hard to reason about, and led to concealing a bug in their branching logic. There were several issues here:
- They tried to accomplish too much: determining filter name, cluster name, and whether RDS should be used.
- They embedded logic to handle significantly different kinds of upstream listeners (passthrough, prepared query, typical services, and catch-all)
- They needed to coalesce different data sources (Upstream and CompiledDiscoveryChain)
Rather than handling all of those tasks inside of these functions, this PR pulls out the RDS/clusterName/filterName logic.
This refactor also fixed a bug with the handling of [UpstreamDefaults](https://www.consul.io/docs/connect/config-entries/service-defaults#defaults). These defaults get stored as UpstreamConfig in the proxy snapshot with a DestinationName of "*", since they apply to all upstreams. However, this wildcard destination name must not be used when creating the name of the associated upstream cluster. The coalescing logic in the original functions here was in some situations creating clusters with a `*.` prefix, which is not a valid destination.
* ui: Filter global intentions list by namespace and partition
Filters global intention listing by the current partition rather than trying to use a wildcard.
Fixes an issue described in #10132, where if two DCs are WAN federated
over mesh gateways, and the gateway in the non-primary DC is terminated
and receives a new IP address (as is commonly the case when running them
on ephemeral compute instances) the primary DC is unable to re-establish
its connection until the agent running on its own gateway is restarted.
This was happening because we always preferred gateways discovered by
the `Internal.ServiceDump` RPC (which would fail because there's no way
to dial the remote DC) over those discovered in the federation state,
which is replicated as long as the primary DC's gateway is reachable.
* Support Vault Namespaces explicitly in CA config
If there is a Namespace entry included in the Vault CA configuration,
set it as the Vault Namespace on the Vault client
Currently the only way to support Vault namespaces in the Consul CA
config is by doing one of the following:
1) Set the VAULT_NAMESPACE environment variable which will be picked up
by the Vault API client
2) Prefix all Vault paths with the namespace
Neither of these are super pleasant. The first requires direct access
and modification to the Consul runtime environment. It's possible and
expected, not super pleasant.
The second requires more indepth knowledge of Vault and how it uses
Namespaces and could be confusing for anyone without that context. It
also infers that it is not supported
* Add changelog
* Remove fmt.Fprint calls
* Make comment clearer
* Add next consul version to website docs
* Add new test for default configuration
* go mod tidy
* Add skip if vault not present
* Tweak changelog text
* Remove some usage of md5 from the system
OSS side of https://github.com/hashicorp/consul-enterprise/pull/1253
This is a potential security issue because an attacker could conceivably manipulate inputs to cause persistence files to collide, effectively deleting the persistence file for one of the colliding elements.
Signed-off-by: Mark Anderson <manderson@hashicorp.com>
Port of: Ensure we check intention service prefix permissions for per service (#11270)
Previously, when showing some action buttons for 'per service intentions' we used a global 'can I do something with any intention' permission to decide whether to show a certain button or not. If a user has a token that does not have 'global' intention permissions, but does have intention permissions on one or more specific services (for example via service / service_prefix), this meant that we did not show them certain buttons required to create/edit the intentions for this specific service.
This PR adds that extra permissions check so we now check the intentions permissions per service instead of using the 'global' "can I edit intentions" question/request.
**Notes:**
- If a HTML button is `disabled` this means tippy.js doesn't adopt the
popover properly and subsequently hide it from the user, so aswell as
just disabling the button so you can't active the popover, we also don't
even put the popover on the page
- If `ability.item` or `ability.item.Resources` are empty then assume no access
**We don't try to disable service > right hand side intention actions here**
Whether you can create intentions for a service depends on the
_destination_ of the intention you would like to create. For the
topology view going from the LHS to the center, this is straightforwards
as we only need to know the permissions for the central service, as when
you are going from the LHS to the center, the center is the
_destination_.
When going from the center to the RHS the _destination[s]_ are on the
RHS. This means we need to know the permissions for potentially 1000s of
services all in one go in order to know when to show a button or not.
We can't realistically discover the permissions for service > RHS
services as we'd have either make a HTTP request per right hand service,
or potentially make an incredibly large POST request for all the
potentially 1000s of services on the right hand side (more preferable to
1000s of HTTP requests).
Therefore for the moment at least we keep the old functionality (thin client)
for the middle to RHS here. If you do go to click on the button and you
don't have permissions to update the intention you will still not be
able to update it, only you won't know this until you click the button
(at which point you'll get a UI visible 403 error)
Note: We reversed the conditional here between 1.10 and 1.11
So this make 100% sense that the port is different here to 1.11
* add root_cert_ttl option for consul connect, vault ca providers
Signed-off-by: FFMMM <FFMMM@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Chris S. Kim <ckim@hashicorp.com>
* add changelog, pr feedback
Signed-off-by: FFMMM <FFMMM@users.noreply.github.com>
* Update .changelog/11428.txt, more docs
Co-authored-by: Daniel Nephin <dnephin@hashicorp.com>
* Update website/content/docs/agent/options.mdx
Co-authored-by: Kyle Havlovitz <kylehav@gmail.com>
Co-authored-by: Chris S. Kim <ckim@hashicorp.com>
Co-authored-by: Daniel Nephin <dnephin@hashicorp.com>
Co-authored-by: Kyle Havlovitz <kylehav@gmail.com>
* ui: Ensure dc selector correctly shows the currently selected dc
* ui: Restrict access to non-default partitions in non-primaries (#11420)
This PR restricts access via the UI to only the default partition when in a non-primary datacenter i.e. you can only have multiple (non-default) partitions in the primary datacenter.
* Add `is` and `test` helpers in a similar vein to `can`
Adds 2 new helpers in a similar vein to ember-cans can:
- `is` allows you to use vocab/phrases such as (is "something model") which calls isSomething() on the models ability.
- `test` allows you to use vocab/phrases such as (test "is something model") or (test "can something model")which calls isSomething() / canSomething() on the models ability. Mostly using the is helper and the can helper. It's basically the is/can helper combined.
* Adds TextInput component + related modifiers/helpers/machines/services (#11189)
Adds a few new components/modifiers/helpers to aid building forms.
- state-chart helper, used in lieu of a more generic approach for requiring our statecharts.
- A few modifications to our existing disabled modifier.
- A new 'validation' modifier, a super small form validation approach built to make use of state charts (optionally). Eventually we should be able to replace our current validation approach (ember-changeset-validations + extra deps) with this.
- A new TextInput component, which is the first of our new components specifically to make it easy to build forms with validations. This is still a WIP, I left some comments in pointing out where this one would be progressed, but as we don't need the planned functionality yet, I left it where it was. All of this will be fleshed out more at a later date.
Documentation is included for all of ^
* ui: Adds initial CRUD for partitions (#11190)
Adds basic CRUD support for partitions. Engineering-wise probably the biggest takeaway here is that we needed to write very little javascript code to add this entire feature, and the little javascript we did need to write was very straightforwards. Everything is pretty much just HTML. Another note to make is that both ember-changeset and ember-data (model layer things) are now completely abstracted away from the view layer of the application.
New components:
- Consul::Partition::Form
- Consul::Partition::List
- Consul::Partition::Notifications
- Consul::Partition::SearchBar
- Consul::Partition::Selector
See additional documentation here for more details
New Route templates:
- index.hbs partition listing/searching/filtering
- edit.hbs partition editing and creation
Additionally:
There is some additional debug work here for better observability and to prevent any errors regarding our href-to usage when a dc is not available in our documentation site.
Our softDelete functionality has been DRYed out a little to be used across two repos.
isLinkable was removed from our ListCollection component for lists like upstream and service listing, and instead use our new is helper from within the ListCollection, meaning we've added a few more lighterweight templateOnly components.
* ui: Exclude all debug-like files from the build (#11211)
This PR adds **/*-debug.* to our test/prod excluded files (realised I needed to add test-support.js also so added that here as its more or less the same thing). Conditionally juggling ES6 static imports (specifically debug ones) for this was also getting a little hairy, so I moved it all to use the same approach as our conditional routes. All in all it brings the vendor build back down to ~430kb gzipped.
From an engineers perspective, whenever specifying colors from now on we should use the form:
```
color: rgb(var(--tone-red-500));
```
Please note:
- Use rgb. This lets us do this like rgb(var(--tone-red-500) / 10%) so we can use a 10% opacity red-500 if we ever need to whilst still making use of our color tokens.
- Use --tone-colorName-000 (so the prefix tone). Previously we could use a mix of --gray-500: $gray-500 (note the left hand CSS prop and right hand SASS var) for the things we need to theme currently. As we no longer use SASS we can't do --gray-500: --gray-500, so we now do --tone-gray-500: --gray-500.
Just for clarity after that, whenever specifying a color anywhere, use rgb and --tone. There is only one reason where you might not use tone, and that is if you never want a color to be affected by a theme (for example a background shadow probably always should use --black)
There are a 2 or 3 left for the code editor, plus our custom-query values
> In the future, this should all be moved to each individual repository now, which will mean we can finally get rid of this service.
This PR moves reconciliation to 'each individual repository'. I stopped short of getting rid of the service, but its so small now we pretty much don't need it. I'd rather wait until I look at the equivalent DataSink service and see if we can get rid of both equivalent services together (this also currently dependant on work soon to be merged)
Reconciliation of models (basically doing the extra work to clean up the ember-data store and bring our frontend 'truth' into line with the actual backend truth) when blocking/long-polling on different views/filters of data is slightly more complicated due to figuring out what should be cleaned up and what should be left in the store. This is especially apparent for KVs.
I built in a such a way to hopefully make sure it will all make sense for the future. I also checked that this all worked nicely with all our models, even KV which has never supported blocking queries. I left all that work in so that if we want to enable blocking queries/live updates for KV it now just involves deleting a couple of lines of code.
There is a tonne of old stuff that we can clean up here now (our 'fake headers' that we pass around) and I've added that to my list of thing for a 'Big Cleanup PR' that will remove lots of code that we no longer require.
Add changelog to document what changed.
Add entry to telemetry section of the website to document what changed
Add docs to the usagemetric endpoint to help document the metrics in code
Our DataSource came in very iteratively, when we first started using it we specifically tried not to use it for things that would require portions of the @src="" attribute to be URL encoded (so things like service names couldn't be used, but dc etc would be fine). We then gradually added an easy way to url encode the @src="" attributes with a uri helper and began to use the DataSource component more and more. This meant that some DataSource usage continued to be used without our uri helper.
Recently we hit #10901 which was a direct result of us not encoding @src values/URIs (I didn't realise this was one of the places that required URL encoding) and not going back over things to finish things off once we had implemented our uri helper, resulting in ~half of the codebase using it and ~half of it not.
Now that almost all of the UI uses our DataSource component, this PR makes it even harder to not use the uri helper, by wrapping the string that it requires in a private URI class/object, that is then expected/asserted within the DataSource component/service. This means that as a result of this PR you cannot pass a plain string to the DataSource component without seeing an error in your JS console, which in turn means you have to use the uri helper, and it's very very hard to not URL encode any dynamic/user provided values, which otherwise could lead to bugs/errors similar to the one mentioned above.
The error that you see when you don't use the uri helper is currently a 'soft' dev time only error, but like our other functionality that produces a soft error when you mistakenly pass an undefined value to a uri, at some point soon we will make these hard failing "do not do this" errors.
Both of these 'soft error' DX features have been used this to great effect to implement our Admin Partition feature and these kind of things will minimize the amount of these types of bugs moving forwards in a preventative rather than curative manner. Hopefully these are the some of the kinds of things that get added to our codebase that prevent a multitude of problems and therefore are often never noticed/appreciated.
Additionally here we moved the remaining non-uri using DataSources to use uri (that were now super easy to find), and also fixed up a place where I noticed (due to the soft errors) where we were sometimes passing undefined values to a uri call.
The work here also led me to find another couple of non-important 'bugs' that I've PRed already separately, one of which is yet to be merged (#11105), hence the currently failing tests here. I'll rebase that once that PR is in and the tests here should then pass 🤞
Lastly, I didn't go the whole hog here to make DataSink also be this strict with its uri usage, there is a tiny bit more work on DataSink as a result of recently work, so I may (or may not) make DataSink equally as strict as part of that work in a separate PR.
This PR adds a check to policy, role and namespace list pages to make sure the user has can write those things before offering to create them via a button. (The create page/form would then be a read-only form)
* ui: Don't show the CRD menu for read-only intentions
The UI bug here manifests itself only when a user/token is configured to have read-only access to intentions. Instead of only letting folks click to see a read only page of the intention, we would show an additional message saying that the intention was read-only due to it being 'Managed by [a kubernetes] CRD'. Whilst the intention was still read only, this extra message was still confusing for users.
This PR fixes up the conditional logic and further moves the logic to use ember-can - looking at the history of the files in question, this bug snuck itself in partly due to it being 'permission-y type stuff' previous to using ember-can and when something being editable or not was nothing to do with ACLs. Then we moved to start using ember-can without completely realising what IsEditable previously meant. So overall the code here is a tiny bit clearer/cleaner by adding a proper can view CRD intention instead of overloading the idea of 'editability'.
* ui: Gracefully recover from non-existent DC errors
This PR fixes what happens in the UI if you try to navigate to a non-existing DC.
When we received a 500 error from an API response due to a non-existent DC, previously we would show a 404 error, which is what we were trying to convey. But in the spirit of the UI being a 'thin client', its probably best to just show the 500 error from the API response, which may help folks to debug any issues better.
* Automatically set the CONSUL_DATACENTER_LOCAL env var for testing
* ui: Ignore response from API for KV permissions
Currently there is no way for us to use our HTTP authorization API
endpoint to tell us whether a user has access to any KVs (including the
case where a user may not have access to the root KV store, but do have
access to a sub item)
This is a little weird still as in the above case the user would click
on this link and still get a 403 for the root, and then have to manually
type in the URL for the KV they do have access to.
Despite this we think this change makes sense as at least something about KV is
visible in the main navigation.
Once we have the ability to know if any KVs are accessible, we can add
this guard back in.
We'd initially just removed the logic around the button, but then
noticed there may be further related KV issues due to the nested nature
of KVs so we finally decided on simply ignoring the responses from the
HTTP API, essentially reverting the KV area back to being a thin client.
This means when things are revisited in the backend we can undo this
easily change in one place.
* Move acceptance tests to use ACLs perms instead of KV ones
Daniel Nephin
Dan Upton
freddygv
R.B. Boyer
John Cowen
Freddy
Mike Morris
Kyle Havlovitz
Matt Keeler
FFMMM
Dhia Ayachi
Iryna Shustava
Alessandro De Blasis
Kenia
99
John Cowen
Nitya Dhanushkodi
Mark Anderson
Giulio Micheloni
Connor
Jared Kirschner
Evan Culver
Konstantine
Paul Banks
Bisakh Mondal
Paul Banks
radiantly
tarat44
Evan Culver
Chris S. Kim