netconf.git
4 weeks agoRetain Encoding in endpoint configuration 79/113779/6
Robert Varga [Tue, 1 Oct 2024 08:38:44 +0000 (10:38 +0200)]
Retain Encoding in endpoint configuration

Do not peel Encoding, just forward requests for default encoding to it.

JIRA: NETCONF-1379
Change-Id: I8df2f190b030f505023ac6a5e28aeca54eb74343
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
4 weeks agoAdd response body example for ROOT operations GET 21/113121/5
Yaroslav Lastivka [Fri, 16 Aug 2024 10:02:00 +0000 (13:02 +0300)]
Add response body example for ROOT operations GET

Hardcoded top-level container "ietf-restconf:operations" for
the OpenAPI-generated example.
Implemented fixed structure to ensure proper schema generation
for both JSON and XML.
Updated unit tests (UT) accordingly to reflect
the new changes.

JIRA: NETCONF-1343
Change-Id: I4aa9f4d782eee929cd49b445913683c3cd1ea33b
Signed-off-by: Yaroslav Lastivka <yaroslav.lastivka@pantheon.tech>
4 weeks agoUpdate Netty RESTCONF API root path configuration 51/113751/12
Robert Varga [Fri, 27 Sep 2024 18:44:48 +0000 (20:44 +0200)]
Update Netty RESTCONF API root path configuration

The 'restconf' configuration property defaults to "rests" for historical
reasons and controls the {+restconf} expansion on the JAX-RS endpoint.

In the Netty endpoint we want to retain the ability to control the path,
but want to start with the default of "restconf", to make the
out-of-the-box experience match RFC8040.

This patch introduces a separate "api-root-path" property, with a
well-documented syntax and semantics.

We propagate this property via NettyEndpointConfiguration in a decoded
form, so there is no confusion as to what the actual semantics are --
which is always a problem with naked Strings.

JIRA: NETCONF-1379
Change-Id: I14ae912082c9407cac796971686e75f0c23cf04f
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
4 weeks agoAdd response body example for ROOT data GET 73/113173/10
Oleksandr Zharov [Wed, 21 Aug 2024 19:43:09 +0000 (21:43 +0200)]
Add response body example for ROOT data GET

Added refs to root data GET requests to fill the response
for this path. Controller and devices have different
models wired up to them depending on what you can
actually see after executing GET request to them.

All UT test resources have been adapted for this change.

JIRA: NETCONF-1342
Change-Id: I38ec5df3183809e062fed1af09ba43cf59e1a034
Signed-off-by: Oleksandr Zharov <oleksandr.zharov@pantheon.tech>
5 weeks agoAdd parameters for ROOT data GET 14/112414/6
Yaroslav Lastivka [Tue, 2 Jul 2024 07:42:25 +0000 (10:42 +0300)]
Add parameters for ROOT data GET

Added parameters for GET ROOT path object,
provides possibility to select config or nonconfig data.
Added possibility to set 'Accept' parameter inside header,
included empty content container for proper OpenAPI parsing.

JIRA: NETCONF-1338
Change-Id: I6584e58d1d8857b6abe202b63e14c03d6fd584d9
Signed-off-by: Yaroslav Lastivka <yaroslav.lastivka@pantheon.tech>
5 weeks agoFix request path interpretation 22/113722/11
Robert Varga [Thu, 26 Sep 2024 16:16:53 +0000 (18:16 +0200)]
Fix request path interpretation

We are decoding the request path before processing it, which leads to
our inability to properly target values.

This patch introduces SegmentPeeler, which aids our traversal of the
path -- thus we can examine decoded segments while we still retain the
ability to pick up the remainder of the path in encoded form.

JIRA: NETCONF-1399
Change-Id: Ib7f7b08db8a97da7e9f60528f9834c50e6ce1ee4
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoOpenApi: Avoid leaf names duplication 25/113425/3
Oleksandr Zharov [Wed, 4 Sep 2024 12:45:53 +0000 (14:45 +0200)]
OpenApi: Avoid leaf names duplication

Slightly modified schema generation logic to add the name
of the module before augmentation nodes to avoid
duplication error for augmentations with the same name as
the augmented node. This should help to distinguish
between augmented and not augmented nodes.

JIRA: NETCONF-1361
Change-Id: Iddae59522949a147ef3a6877bef30cfba1d24b20
Signed-off-by: Oleksandr Zharov <oleksandr.zharov@pantheon.tech>
Signed-off-by: Ivan Hrasko <ivan.hrasko@pantheon.tech>
5 weeks agoRemove NettyEndpointConfiguration.baseUri() 58/113658/5
Robert Varga [Thu, 26 Sep 2024 20:52:49 +0000 (22:52 +0200)]
Remove NettyEndpointConfiguration.baseUri()

The idea that the user needs to configure a hostname for Netty endpoint
is quite wrong: we get an absolute URI with each request and hence can
trivially derive restconfURI that the caller can use to reach us.

Remove the host-configuration knob and streamline
NettyEndpointConfiguration by propagating only the restconf() prefix.

JIRA: NETCONF-1379
Change-Id: Ia4857e74e109e04339865fec31f4261343ab953d
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoAdd String-based StorageEntry constructor 48/113748/2
Robert Varga [Fri, 27 Sep 2024 15:02:18 +0000 (17:02 +0200)]
Add String-based StorageEntry constructor

We have two places instantiating StorageEntry from strings:
- TestUtils.storageEntry()
- PlaintextLocalFileStorage.readData()

Expose the appropriate constructor and eliminate and let tests use that.

Change-Id: I4f0225ded5cc925c6197a3b12e36f84f36284856
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoDo not allow StorageEntry to be subverted 47/113747/1
Robert Varga [Fri, 27 Sep 2024 14:53:08 +0000 (16:53 +0200)]
Do not allow StorageEntry to be subverted

Perform a defensive clone on {insert,put}Entry(), so that these cannot
be modified later.

Change-Id: Id41b1a157e3ec1ca30f26b0e37c20d59d7fb983e
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoStorageEntry should not be a Map.Entry 44/113744/4
Robert Varga [Fri, 27 Sep 2024 11:02:10 +0000 (13:02 +0200)]
StorageEntry should not be a Map.Entry

StorageEntry is internal storage and as such it does not follow the
Map.Entry contract -- most notably it has a different hashCode().

Stop claiming we are implementing Map.Entry and instead use Map.entry()
to construct the one to give out to users.

Change-Id: I6ef9900f32d65d129edbbc7aa84e89177ca853e3
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoUse a iterator mapping 43/113743/2
Robert Varga [Fri, 27 Sep 2024 10:43:01 +0000 (12:43 +0200)]
Use a iterator mapping

Each of our entries snapshot is immutable, hence we can use iterator
transforms instead of linearizing to array.

Change-Id: Ic41fbfc35a39db704c78de1ccea5381dfd3ecb8b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoCleanup plaintext-localfile dependencies 42/113742/2
Robert Varga [Fri, 27 Sep 2024 10:30:08 +0000 (12:30 +0200)]
Cleanup plaintext-localfile dependencies

We have superfluous dependencies specified and have them misordered. Fix
both up.

Change-Id: I3101f9f581c3eefeb1e92b4b5673028e3936c980
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoRemove superfluous nullness annotations 41/113741/2
Robert Varga [Fri, 27 Sep 2024 09:52:12 +0000 (11:52 +0200)]
Remove superfluous nullness annotations

Annotations are specified in interface contract, there is no point in
repeating them.

Change-Id: I0c346d62b7c1bde2b5978f7c80b707d601c5fe8a
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoDo not use Optional during lookup 40/113740/2
Robert Varga [Fri, 27 Sep 2024 09:48:27 +0000 (11:48 +0200)]
Do not use Optional during lookup

We are using Optional only to remap it internally. Rename findByKey() to
lookupEntry() and return a nullable instead.

While we are here, also make sure we reject null keys consistently.

Change-Id: I404247e93e34f3e05405e6e0517d942f6b8a5489
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoSpecialize TransportChannelListeners 19/113719/6
Robert Varga [Thu, 26 Sep 2024 14:27:53 +0000 (16:27 +0200)]
Specialize TransportChannelListeners

Take advantage of the ability to cap which TransportChannels a
particular listeners supports. This allows us to eliminate a cast in
RestconfTransportChannelListener.

JIRA: NETCONF-1400
Change-Id: I1f7000b724eb9f94bc28f9cec51f098165d5f722
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoFix private TransportChanneListener references 20/113720/5
Robert Varga [Thu, 26 Sep 2024 14:18:23 +0000 (16:18 +0200)]
Fix private TransportChanneListener references

Our switch making TransportChanneListener a parameterized interface is
causing quite a few warnings.

This patch addresses warnings private to transport implementations,
without an ABI impact.

JIRA: NETCONF-1400
Change-Id: I5d3aa813f9ef4c859b72e774376031facf614ee5
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoParameterize TransportChannelListener 17/113717/6
Robert Varga [Thu, 26 Sep 2024 12:01:01 +0000 (14:01 +0200)]
Parameterize TransportChannelListener

TransportChannel types are exposed to users and some of them provide
useful information. Allow the TransportChannel type to be specified
for TransportChannelListener.

JIRA: NETCONF-1400
Change-Id: Ia5d589ff2496153ee6909350217bb31a4152b4ad
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoPropagate Target URI to dispatcher 16/113716/2
Robert Varga [Thu, 26 Sep 2024 11:30:52 +0000 (13:30 +0200)]
Propagate Target URI to dispatcher

We need the Target URI in RestconfRequestDispatcher, so that we can
propagate the correct {+restconf} base, as seen by the client, to the
RestconfServer.

JIRA: NETCONF-1379
Change-Id: Iac3fb12bc708f72496f576260c7de7a0fb09c5c7
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoRework resource lookup 08/113708/5
Robert Varga [Wed, 25 Sep 2024 10:36:44 +0000 (12:36 +0200)]
Rework resource lookup

RestconfRequestDispatcher is using indirection via PathParameters to
interpret the requested path.

There are multiple problems here:
- the logic is quite undocumented
- there is an overload of PathArguments.EMPTY to report at least two
  distinct paths
- we are reporting 409 where we should be reporting 404, as noted in
  RFC8040 section 4.3
- the splitting works on decoded path, whereas the ApiPath.parse()
  should always we presented with an encoded path

This patch deals with the first three problems, documenting how we are
peeling the path and what gaps there are relative to the specification.

The last problem is marked with a FIXME and will be addressed in a
follow-up patch.

JIRA: NETCONF-1379
Change-Id: I408626a5e328bb938f727d9a549ca8297c9df3a9
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoCorrect well-known error status code 07/113707/3
Robert Varga [Wed, 25 Sep 2024 22:59:54 +0000 (00:59 +0200)]
Correct well-known error status code

We are now reporting unimplemented method from RestconfSession, hence
WellKnownResources can report a proper 405 Method Not Allowed by
default.

JIRA: NETCONF-1379
Change-Id: I62c4323e8bc4f4fe0c0f082bffec18dffab55f01
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoReconstruct Target URI, part two 83/113683/9
Robert Varga [Wed, 25 Sep 2024 16:39:57 +0000 (18:39 +0200)]
Reconstruct Target URI, part two

Reconstructing Target URI from its usual origin-form is a tad more
involved:

1. we need access to the URI scheme being used, which we now have
   conveniently available from HTTPTransportChannel
2. we need to ensure we have a Host header, that it contains only one
   value and that it is a valid authority

We employ java.net.URI for most of the parsing, combining and error
reporting functions. It also allows us to determine in the uri is in
absolute-form, in which case we can pick it up as is.

We use this URI to initialize QueryStringDecoder, so that its uri()
method reports the complete thing.

JIRA: NETCONF-1379
Change-Id: I6641a613f2fd4c63332ae29b0a876ca32a325a03
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoReconstruct Target URI, part one 80/113680/7
Robert Varga [Wed, 25 Sep 2024 12:31:06 +0000 (14:31 +0200)]
Reconstruct Target URI, part one

In order to correctly serve our requests, we need to properly
reconstruct target URI, as per HTTP/1.1 spec.

This patch deals with the simplest case, asterisk-form: it is only
allowed for OPTIONS request and reports all methods implemented by
the server. All other methods fails with either status 400.

That necessitates knowledge as to what methods are requested, hence
we have a statically-populated Set.

JIRA: NETCONF-1379
Change-Id: Id37044799460759d1066ad7a17524ee0c0284060
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoExpose HttpScheme from HTTPTransportChannel 05/113705/4
Robert Varga [Wed, 25 Sep 2024 18:48:46 +0000 (20:48 +0200)]
Expose HttpScheme from HTTPTransportChannel

Downstream users may need to inquire as to whether the channel is secure
or not -- for example to follow RFC9110 Target URI construction.

Expose HttpScheme from HTTPTransportChannel, so that this information is
readility available.

JIRA: NETCONF-1379
Change-Id: Iba7f847246ab9c88a9804caba803f22ab704f406
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoSpecialize HTTPClient 84/113684/4
Robert Varga [Wed, 25 Sep 2024 19:13:36 +0000 (21:13 +0200)]
Specialize HTTPClient

We know statically whether a particular client is TLS-enabled, so use
two subclasses, each specialized for the task.

JIRA: NETCONF-1379
Change-Id: Ib46a3c779dd2f3bdf060561a754180f099fdc997
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoMove ResponseUtils.optionsResponse() 73/113673/3
Robert Varga [Tue, 24 Sep 2024 22:56:46 +0000 (00:56 +0200)]
Move ResponseUtils.optionsResponse()

This method is only called from RestconfRequestDispatcher. Move it there
and simplify it by bypassing ResponseBuilder's complexities.

JIRA: NETCONF-1379
Change-Id: Ie7647fcc2eab6b8c1c9112da55283efb1b5819ae
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoClean up NettyServerRequest subclasses 72/113672/4
Robert Varga [Tue, 24 Sep 2024 22:42:19 +0000 (00:42 +0200)]
Clean up NettyServerRequest subclasses

Almost all anonymous NettyServerRequest subclasses are inefficient in
that they retain RequestParameters twice:
- once in NettyServerRequest
- once via reference to 'params' method argument

Make sure we reference NettyServerRequest.requestParams instead of
params, reducing memory overhead of each request.

JIRA: NETCONF-1379
Change-Id: I0f7df86dbd8681f2d4e92f7dabefb6f6820709ef
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoAdd FormattableServerRequest 71/113671/3
Robert Varga [Tue, 24 Sep 2024 22:35:36 +0000 (00:35 +0200)]
Add FormattableServerRequest

We have two distinct NettyServerRequest subclasses doing exactly the
same thing. Share a single FormattableServerRequest class between the
two code paths.

JIRA: NETCONF-1379
Change-Id: Id30306ce9073afdd73d790282d3a023dde2c4c07
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoIntroduce RestconfRequest 70/113670/3
Robert Varga [Tue, 24 Sep 2024 22:03:41 +0000 (00:03 +0200)]
Introduce RestconfRequest

The use of FutureCallback is misleading and not very well controlled.
Introduce RestconfRequest, which shows that the onFailure() path is
never taken.

This also provides us with an anchor point for future evolution, as we
must be mindful of asynchronous processing. The requirements are laid
down in a FIXME, which we will address in future improvements.

JIRA: NETCONF-1379
Change-Id: I3e9a25ce6932498654567c12ea431417e27a2c7b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoDo not pass around RestconfServer 69/113669/4
Robert Varga [Tue, 24 Sep 2024 21:31:57 +0000 (23:31 +0200)]
Do not pass around RestconfServer

Now that we have localized request dispatch, we can access
RestconfServer via RestconfRequestDispatcher's field rather than passing
it around to static methods.

JIRA: NETCONF-1379
Change-Id: Ie1fd87d1f873b86d46ce68e8e127b13c4e13d5d8
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoLimit the number of candidates for string example 06/113706/2
Sangwook Ha [Wed, 25 Sep 2024 20:32:10 +0000 (13:32 -0700)]
Limit the number of candidates for string example

The number of candidates tried when a string example is generated
with regex & length constraints may increase exponentially as the
string length increases and this may cause out-of-memory error.

For example, if there are 10 transition paths at each step, then
the number of candidates becomes 1 million after just 6 characters.

Limit the maximum number of transitions allowed at each step to prevent
the exponential growth in search paths. This reduces the chance of
generating a conformant example a little, but prevents excessive use
of memory and CPU cycles.

JIRA: NETCONF-1398
Change-Id: I2d926a438da14395f56f93d1a98a56507c1c7ff2
Signed-off-by: Sangwook Ha <sangwook.ha@verizon.com>
5 weeks agoDo not use LinkedList in plaintext-* 74/113674/2
Robert Varga [Wed, 25 Sep 2024 04:19:30 +0000 (06:19 +0200)]
Do not use LinkedList in plaintext-*

Use an ArrayList instead.

Change-Id: I95676a3d3db5870d09d40c54151439d2e1667dd0
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoEliminate RequestUtils 68/113668/1
Robert Varga [Tue, 24 Sep 2024 21:12:48 +0000 (23:12 +0200)]
Eliminate RequestUtils

The only production caller is RestconfRequestDispatcher, move these
utility methods there.

Hide them and ensure AbstractRequestProcessorTest uses explicit
calls to ApiPath.parse().

JIRA: NETCONF-1379
Change-Id: I8f5c69f57d23f9bd74e9f49d0519f8e078c2c6da
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoEliminate ModulesRequestProcessor 67/113667/1
Robert Varga [Tue, 24 Sep 2024 21:04:09 +0000 (23:04 +0200)]
Eliminate ModulesRequestProcessor

RestconfRequestDispatcher can host the static methods just as well, and
such organization is actually conducive to structuring dispatch based on
needs, rather than upfront parsing.

Also updates RFC references to point to their canonical web presence.

JIRA: NETCONF-1379
Change-Id: I268ba415f30daf492fe12573a3b09bb9ed8b3a6f
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoEliminate OperationsRequestProcessor 66/113666/1
Robert Varga [Tue, 24 Sep 2024 20:56:50 +0000 (22:56 +0200)]
Eliminate OperationsRequestProcessor

RestconfRequestDispatcher can host the static methods just as well, and
such organization is actually conducive to structuring dispatch based on
needs, rather than upfront parsing.

Also updates RFC references to point to their canonical web presence.

JIRA: NETCONF-1379
Change-Id: Ib51a3c83de9400c61304d99d4a3b21f99f502045
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
5 weeks agoEliminate DataRequestProcessor 65/113665/1
Robert Varga [Tue, 24 Sep 2024 20:45:21 +0000 (22:45 +0200)]
Eliminate DataRequestProcessor

RestconfRequestDispatcher can host the static methods just as well, and
such organization is actually conducive to structuring dispatch based on
needs, rather than upfront parsing.

Also updates RFC references to point to their canonical web presence.

JIRA: NETCONF-1379
Change-Id: I0ec4cdd2634cd707e5bbcef2617e4eebc3173923
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
6 weeks agoClarify FIXME for NETCONF-714 09/113609/1
Ivan Hrasko [Mon, 23 Sep 2024 12:56:53 +0000 (14:56 +0200)]
Clarify FIXME for NETCONF-714

Clarify FIXME in NettyServerRequest to be part of NETCONF-714.

JIRA: NETCONF-714.
Change-Id: Idf7192ff65fd9d48a7d8d00f22b15c294a7faa49
Signed-off-by: Ivan Hrasko <ivan.hrasko@pantheon.tech>
6 weeks agoFix testtool request failure to netty endpoint 51/113451/5
Ruslan Kashapov [Fri, 6 Sep 2024 08:29:53 +0000 (11:29 +0300)]
Fix testtool request failure to netty endpoint

Testtool uses java's HttpClient which includes
HTTP/2 upgrade headers by default when PATCH requests
being sent. This results 413 Entry too large response
from Netty-based server because HttpServerUpgradeHandler
is not expecting non-empty request body with
protocol upgrade request.

The fix limits protocol version for HttpClient to
HTTP 1.1 to omit upgrade headers.

JIRA: NETCONF-1376
Change-Id: I2c07edce78b521417cf073f5c0d757e58f557c91
Signed-off-by: Ruslan Kashapov <ruslan.kashapov@pantheon.tech>
6 weeks agoFix 401 Response headers for HTTP Transport 70/113570/2
Ruslan Kashapov [Fri, 6 Sep 2024 08:29:53 +0000 (11:29 +0300)]
Fix 401 Response headers for HTTP Transport

Auth handler updated to return WWW-Authenticate
header with 401 Unauthorized response.

JIRA: NETCONF-1394
Change-Id: I018b7c37ec08eae4525355001265f4151bafa298
Signed-off-by: Ruslan Kashapov <ruslan.kashapov@pantheon.tech>
6 weeks agoBump versions to 8.0.3-SNAPSHOT 71/113571/1
Robert Varga [Thu, 19 Sep 2024 12:08:56 +0000 (14:08 +0200)]
Bump versions to 8.0.3-SNAPSHOT

This starts the next development iteration.

Change-Id: I02b4ab07c5385f86af946b821f1ede30d58b4ce9
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
6 weeks agoRelease netconf v8.0.2
jenkins-releng [Thu, 19 Sep 2024 10:42:16 +0000 (10:42 +0000)]
Release netconf

6 weeks agoDispatch OPTIONS requests to RestconfServer 61/113561/4
Robert Varga [Wed, 18 Sep 2024 15:18:45 +0000 (17:18 +0200)]
Dispatch OPTIONS requests to RestconfServer

RestconfServer is providing guidance as to what kind of resource a
particular path represents. We take advantage of this, to ensure parity
with JaxRsRestconf.

JIRA: NETCONF-1210
Change-Id: Iab8aa05f2183f5f0f9b5f9d283bea9cdd9579eb6
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
6 weeks agoFixup OptionsResult 65/113565/1
Robert Varga [Thu, 19 Sep 2024 08:43:24 +0000 (10:43 +0200)]
Fixup OptionsResult

We need to differentiate between RPCs and actions, an unfortunate
side-effect of difference of their invocation modeling.

Split OPERATION into ACTION and RPC, and report them as appropriate.

JIRA: NETCONF-1210
Change-Id: Ia9948998ffedb4e7ff1025cb4e268ffeb1a68ad9
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoAdapt docs for netconf-node state encapsulation 45/113545/2
Samuel Schneider [Mon, 16 Sep 2024 15:38:42 +0000 (17:38 +0200)]
Adapt docs for netconf-node state encapsulation

Adapt user guide to changes caused by netconf-node state
encapsulation.

JIRA: NETCONF-1388
Change-Id: I477f14e814c0d4dfa3833454b8803c2a72fdc18e
Signed-off-by: Samuel Schneider <samuel.schneider@pantheon.tech>
7 weeks agoFix stream URLs for Netty RESTCONF 12/113512/3
Ruslan Kashapov [Fri, 13 Sep 2024 09:42:01 +0000 (12:42 +0300)]
Fix stream URLs for Netty RESTCONF

Restconf URI passed for URL construction require to be ended
with slash otherwise last path element will be dropped by URI#resolve.

JIRA: NETCONF-1390
Change-Id: I1707b8460741bd99a971359fd152152b60440c8c
Signed-off-by: Ruslan Kashapov <ruslan.kashapov@pantheon.tech>
Signed-off-by: Ivan Hrasko <ivan.hrasko@pantheon.tech>
7 weeks agoTranslate RpcCredentials with a switch expression 32/113532/1
Robert Varga [Mon, 16 Sep 2024 09:14:14 +0000 (11:14 +0200)]
Translate RpcCredentials with a switch expression

We have an if/else cascade here, use a switch expression instead.

Change-Id: I6d98c2fcd07ab06f9a161ed663864b00347686ee
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoEncapsulate netconf-node state 82/113482/8
Robert Varga [Mon, 9 Sep 2024 19:27:44 +0000 (21:27 +0200)]
Encapsulate netconf-node state

Encapsulate netconf device related configuration into its own container.
So mandatory credentials requirement is not enforced for anyone who is
not writing that container.

This allows using other custom topologies without issues.

JIRA: NETCONF-1382
Change-Id: If015222df4b4a885c832f8de35f3c482e9dfbc2a
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
Signed-off-by: Samuel Schneider <samuel.schneider@pantheon.tech>
7 weeks agoAdd {data,operations}OPTIONS() methods 81/113481/17
Robert Varga [Mon, 9 Sep 2024 15:48:50 +0000 (17:48 +0200)]
Add {data,operations}OPTIONS() methods

JAX-RS endpoint is doing a number of bad things w.r.t OPTIONS:
- it does not produce Accept-Patch header
- it exposes implementation structure via WADL
- it is not accurate for /data/* resources

Enrich RestconfServer API to allow querying the nature of a particular
data item and existence of operations -- which in turn allows us to
report accurate OPTIONS.

JIRA: NETCONF-1210
Change-Id: Iee4171eb9500bb996a63e682c946c20d6e401145
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoSet correct SchemaInferenceStack for list example 07/113507/5
Peter Suna [Thu, 12 Sep 2024 16:32:09 +0000 (18:32 +0200)]
Set correct SchemaInferenceStack for list example

In the method prepareFirstListExample, SchemaInferenceStack points
to the list, which is incorrect as we are processing its child.

This method is called only when the min-elements constraint is set.

JIRA: NETCONF-1389
Change-Id: I4ffb371751a71012ead461837f775a89957fd111
Signed-off-by: Peter Suna <peter.suna@pantheon.tech>
7 weeks agoFix string example generation error 08/113508/2
Sangwook Ha [Fri, 13 Sep 2024 04:03:03 +0000 (21:03 -0700)]
Fix string example generation error

Following a single regex transition path does not guarantee an example
conforming to the regex & length constraints.

Search all transition paths adding a character at a time until the
minimum length requirement & regex pattern are satisfied or the maximum
length limit is reached.

This guarantees the shortest possible example that satisfies the regex
& length constraints.

JIRA: NETCONF-1386
Change-Id: Ib7b7714330b5176541fe928dc5660c20579f8b2d
Signed-off-by: Sangwook Ha <sangwook.ha@verizon.com>
7 weeks agoEliminate Http2SettingsFilter 50/113450/4
Robert Varga [Fri, 6 Sep 2024 09:50:11 +0000 (11:50 +0200)]
Eliminate Http2SettingsFilter

We need Http2SettingsFilter only because Http2Utils.connectionHandler()
configures Http2ToHttpAdapter to send it out.

Do not override the default configuration, leading to HTTPClient's
adapter not sending these -- and hence we can ditch the filter.

JIRA: NETCONF-1379
Change-Id: I0be0ec6b7375b12c183bc2076276c0c6932455ba
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoClean up restconf-nb dependencies 03/113503/3
Robert Varga [Wed, 11 Sep 2024 15:46:23 +0000 (17:46 +0200)]
Clean up restconf-nb dependencies

The dependencies are quite inaccurate now, clean them a up a bit.

Change-Id: I19229a1f8d9027173ea1f42f5fdfa90d336a2acc
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoClean up odl-restconf-nb 05/113505/1
Robert Varga [Thu, 12 Sep 2024 08:01:04 +0000 (10:01 +0200)]
Clean up odl-restconf-nb

Clean up packaging, so we do not pull in duplicate bundles.

Change-Id: I379e1ebbdb203c23c0d8ae21554d04b9e3f32203
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoAdd odl-restconf-(server-)api 04/113504/1
Robert Varga [Wed, 11 Sep 2024 16:11:37 +0000 (18:11 +0200)]
Add odl-restconf-(server-)api

Package restconf-api as a separate feature, so that it can be used from
other places.

Change-Id: I5b40e2814ec15b78b74948c1e2120010425dca7d
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoMove restconf-server-jaxrs tests 01/113501/4
Robert Varga [Wed, 11 Sep 2024 10:39:37 +0000 (12:39 +0200)]
Move restconf-server-jaxrs tests

Move tests into the artifact they are actually testing.

Change-Id: I2d504bcd40c98bb9f0bead309c7e5c7ef6c8e0bf
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoPackage transport-http in odl-netconf-transport 02/113502/1
Robert Varga [Wed, 11 Sep 2024 14:20:38 +0000 (16:20 +0200)]
Package transport-http in odl-netconf-transport

transport-http does not have a home. Make sure to include it in
odl-netconf-transport.

JIRA: NETCONF-1379
Change-Id: Ia6d12e1866a1eb22609586aee578015bb11688d2
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoIntegrate transformation to NettyServerRequest 95/113495/3
Robert Varga [Tue, 10 Sep 2024 14:06:52 +0000 (16:06 +0200)]
Integrate transformation to NettyServerRequest

Make NettyServerRequest abstract, with a transform() method. This makes
it clear what specializations we have and also reduces the amount of
objects we allocate.

JIRA: NETCONF-1379
Change-Id: Ifddc410b929f6b8e64adbf2e94c40bc08856fea5
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoSlice XRD/JRD resource 00/113500/1
Robert Varga [Wed, 11 Sep 2024 08:17:51 +0000 (10:17 +0200)]
Slice XRD/JRD resource

We need to use ByteBuf.slice() so the reader index on shared buffer does
not get updated.

JIRA: NETCONF-1379
Change-Id: Ibd69b480748094fbbe227ef767f47f14b91943cb
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoUse a switch expression with DataPostResult 96/113496/2
Robert Varga [Tue, 10 Sep 2024 14:14:58 +0000 (16:14 +0200)]
Use a switch expression with DataPostResult

We have a sealed hierarchy, hence using a switch expression is
exhaustive and we can eliminate an impossible case. The same thing
applies to both RESTCONF endpoints.

Change-Id: Ia06223bbccb7fda2a33c67cc129c04cd1cc4a335
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoUse OPTIONS * for upgrade 75/113475/2
Robert Varga [Mon, 9 Sep 2024 15:09:50 +0000 (17:09 +0200)]
Use OPTIONS * for upgrade

GET / is a resource operation, which may result a ton of data, be
subject to access control, or actually be unimplemented.

Use OPTIONS *, which is well-documented to be a 'no-op' kind of HTTP
request.

Change-Id: I9e112ffa1817d09b77109171a3a221ed34c9bf6b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoAdvertize "defaults" RESTCONF capability 74/113474/3
Robert Varga [Mon, 9 Sep 2024 14:21:13 +0000 (16:21 +0200)]
Advertize "defaults" RESTCONF capability

Advertizement of "defaults" RESTCONF capability is mandatory, yet we do
not do that.

Advertize it with basic-mode=explicit, which most closely matches the
semantics of the MD-SAL datastore.

JIRA: NETCONF-1380
Change-Id: I48577e606383fc7413a3cd2e95a5c8cbb52fb49f
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoDo not use karaf-services-maven-plugin 94/113494/1
Robert Varga [Tue, 10 Sep 2024 11:31:26 +0000 (13:31 +0200)]
Do not use karaf-services-maven-plugin

We are not really using services -- just publishing Karaf commands.

Add the MANIFEST header and ditch use of karaf-services-maven-plugin --
fixing a build-time warnings and eliminating the unnecessary code it
packages in the bundle.

Change-Id: Ia9dc9cf8f837ac77f0f9998a2943dc1bd73f5712
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoFix truststore-api dependencies 93/113493/1
Robert Varga [Tue, 10 Sep 2024 11:24:05 +0000 (13:24 +0200)]
Fix truststore-api dependencies

We are using nullness annotations, declare them to fix a warning.

Change-Id: I6ed2b81420fecb2023038b773cfaf3db2683ece8
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
7 weeks agoFix dependency issues in keystore/* 92/113492/1
Robert Varga [Tue, 10 Sep 2024 11:23:11 +0000 (13:23 +0200)]
Fix dependency issues in keystore/*

We have two dependency issues reported here, fix them up.

Change-Id: I9bdcc0ae7c7ca4289276e2d445ea6b34e20877bc
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoFix create-notification-stream implementation 87/113487/2
Robert Varga [Mon, 9 Sep 2024 23:13:17 +0000 (01:13 +0200)]
Fix create-notification-stream implementation

We have a mismatch in the output leaf name, as sal-remote.yang has a
day-0 modeling bug: what was supposed to be 'stream-name' is named
'notification-stream-identifier' instead.

Fix the implementation to use the correct container.

JIRA: NETCONF-1295
Change-Id: I9ea3d4b3bb6c70b4c087e597f67456a9a095f789
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoEliminate ResponseUtils.allowHeaderValue() 60/113460/4
Robert Varga [Sat, 7 Sep 2024 20:46:20 +0000 (22:46 +0200)]
Eliminate ResponseUtils.allowHeaderValue()

This creates a static string, from a set of methods. Rather than
performing magic, let's use hand-crafted equivalents.

Also do not share these strings with tests, but rather have tests assert
them separately.

JIRA: NETCONF-1379
Change-Id: Id241430ac633f3bb8c1166dea3e65c34c9266975
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoEliminate restconf.server.Method 59/113459/4
Robert Varga [Sat, 7 Sep 2024 20:33:05 +0000 (22:33 +0200)]
Eliminate restconf.server.Method

This enum might look like it is useful until, but it really is not: we
can equally well use HttpMethod and switch on HttpMethod.name().

JIRA: NETCONF-1379
Change-Id: I30db6a5dca57270f9a774298906d10a327816c3a
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoSplit out WellKnownResources 57/113457/12
Robert Varga [Sat, 7 Sep 2024 11:28:57 +0000 (13:28 +0200)]
Split out WellKnownResources

Well-known resources are extremely simple to service, as they take only
a few inputs and produce an immediate response.

Split them out into their own class and execute them immediately from
RestconfSession. This allows us to report '501 Not Implemented' for
unimplemented methods, along with 'Allow' header hinting at what is
implemented.

JIRA: NETCONF-1379
Change-Id: I91af1aee99e2ffb2bf05d6833d64dd8471a40765
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoTake ownership of request message 49/113449/7
Robert Varga [Fri, 6 Sep 2024 08:39:24 +0000 (10:39 +0200)]
Take ownership of request message

We have asynchronous processing here, hence
SimpleChannelInboundHandler's defaults are working against us.

Switch off autoRelease, hence we do not need to issue retain().

JIRA: NETCONF-1379
Change-Id: If4d86111c97122638f17bed52805468d2c3de240
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoExpose ServerSseHandler 43/113443/9
Robert Varga [Thu, 5 Sep 2024 15:33:35 +0000 (17:33 +0200)]
Expose ServerSseHandler

Rather than hiding this handler behind a simple SseUtils method, expose
it to the outside world, so it can be used directly by NettyRestconf's
initialization code.

JIRA: NETCONF-1379
Change-Id: Ia3d2a19525115323b7d613053f561dd169a1c760
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoDisconnect HTTPServer from RequestDispatcher 36/113436/18
Robert Varga [Wed, 4 Sep 2024 23:52:57 +0000 (01:52 +0200)]
Disconnect HTTPServer from RequestDispatcher

RequestDispatcher is useless layering violation sitting between Netty
pipeline and RESTCONF protocol implementation.

Do not use RequestDispatcher in HTTPServer and instead use
onTransportChannelEstablished() to wire a RestconfSession into the
pipeline.

This means we have centralized protocol-level initialization at the
correct layer and a proper class interfacing between a particular Netty
channel and the RestconfServer.

RestconfRequestDispatcher now becomes a private implementation detail,
acting as a the common router between RestconfSession and
RestconfServer (et al.).

A follow-up patch will deal with exposing RestconfSession from
NettyServerRequest.

JIRA: NETCONF-1379
Change-Id: Ie24e440dbf24a850428902076175fb569981d0d9
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoRename CreateNotificationStreamRpcTest 84/113484/1
Robert Varga [Mon, 9 Sep 2024 22:24:17 +0000 (00:24 +0200)]
Rename CreateNotificationStreamRpcTest

This is really testing CreateDataChangeEventSubscriptionRpc, rename it
accordingly.

Change-Id: Icb0a151c8d2d9b2c5397f5426529a170638e178a
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoClean up CreateNotificationStreamRpc 83/113483/1
Robert Varga [Mon, 9 Sep 2024 22:19:53 +0000 (00:19 +0200)]
Clean up CreateNotificationStreamRpc

Use CreateNotificationStreamOutput instead of
CreateDataChangeEventSubscriptionOutput, just for consistency's sake.

Change-Id: Ia5f4734a2763848592a16bb407f1898a3f272014
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoFix ietf-restconf.yang reference 80/113480/2
Robert Varga [Mon, 9 Sep 2024 15:50:12 +0000 (17:50 +0200)]
Fix ietf-restconf.yang reference

We have a typo in Javadoc, fix that.

Change-Id: I464c854bda2e5a11b986d85979775e234050996f
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoAdvertize :yang-patch capability 73/113473/2
Robert Varga [Mon, 9 Sep 2024 13:32:11 +0000 (15:32 +0200)]
Advertize :yang-patch capability

We support yang-patch media type, hence we should be advertizing that
capability.

JIRA: NETCONF-1365
Change-Id: I0aec76d5e1558b1116d72e8ba60ead34e3b46e75
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoDeprecate capabilityUri() methods 72/113472/2
Robert Varga [Mon, 9 Sep 2024 13:29:25 +0000 (15:29 +0200)]
Deprecate capabilityUri() methods

We have the capability strings centralized in CapabilityURI class, hence
we do not need them in individual parameters. Deprecate them for
removal.

Change-Id: I4ec07a2109ed3967079e140a02c1f92c6f2fd453
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoAdd restconf.api.CapabilityURN 71/113471/3
Robert Varga [Mon, 9 Sep 2024 13:01:18 +0000 (15:01 +0200)]
Add restconf.api.CapabilityURN

We are missing a few key capabilities, simply because we are exposing
only capabilities tied to query parameters.

This patch adds restconf.api.CapabilityURN with current contents of the
corresponding IANA registry, along with ODL-specific capabilities --
mirroring what we have in netconf.api.CapabilityURN.

Change-Id: I418feb6784c732b14f9e2b1c2c91dcfec87699ec
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoFactor out RestconfTransportChannelListener 35/113435/10
Robert Varga [Wed, 4 Sep 2024 22:51:11 +0000 (00:51 +0200)]
Factor out RestconfTransportChannelListener

Pipeline initialization for transport channels sits in an anonymous
class, make it non-obvious as to what its actual role should be.

Split it out into RestconfTransportChannelListener, making it
independent of NettyEndpoint.

JIRA: NETCONF-1379
Change-Id: Ia1a45dfa957f737923aca5e0da3b99004fe727c3
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoFixup restconf-server-mdsal module definition 70/113470/1
Robert Varga [Mon, 9 Sep 2024 13:04:53 +0000 (15:04 +0200)]
Fixup restconf-server-mdsal module definition

We require the ietf-restconf-monitoring module, make sure we advertize
that.

Change-Id: I6c3dd01c14ba92001bbcd59d2f68e9f6bcff8cd5
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoFix host meta template constants 53/113453/1
Robert Varga [Fri, 6 Sep 2024 12:55:14 +0000 (14:55 +0200)]
Fix host meta template constants

The strings should be proper constants, so they can be inlined.

Change-Id: I7ab918b7f7863c70d6f8bb6f286bd2a5934af5c1
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoRefactor HTTPServer pipeline initialization 39/113439/4
Robert Varga [Thu, 5 Sep 2024 12:48:36 +0000 (14:48 +0200)]
Refactor HTTPServer pipeline initialization

The pipeline should be initialized immediately when the channel is
brought up, so that we do not require additional client-side operations
before it is visible the the server side.

This necessitates smarter operations on the pipeline, so that we keep
RequestDispatcher at the last position.

Change-Id: If0f0b0456407ab2b6457c1a34565c87f61c433d5
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoFix heartbeat overflow 34/113434/4
Robert Varga [Wed, 4 Sep 2024 22:24:43 +0000 (00:24 +0200)]
Fix heartbeat overflow

sseHeartbeatIntervalMillis is an Uint32, hence it is wrong to assume its
value can safely fit in an int.

Use .toJava() instead, which returns a proper long, exposing the fact
that ServerSseHandler is using the wrong type.

Change-Id: I309bc441c56d1052c3ea013dd8e33db4c80490cc
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoClean up OSGINorthbound.newNettyEndpointProps() 32/113432/5
Robert Varga [Wed, 4 Sep 2024 19:16:39 +0000 (21:16 +0200)]
Clean up OSGINorthbound.newNettyEndpointProps()

We have a magic check for URI scheme -- i.e. if we grow a different
transport, we might happily be using 'http' without ever realizing it.

Refactor logic to make it absolutely clear which cases we handle, so
that there can be no mistakes.

Also clean up obvious gaps in semantics of objects we create and add
FIXMEs/TODOs for follow-ups.

Change-Id: I5a5b812d137fc4d665cfa457a1df1d5bbe4968e6
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
8 weeks agoUse simple concatenation in SseUtils.chunkOf() 42/113442/2
Robert Varga [Thu, 5 Sep 2024 14:55:54 +0000 (16:55 +0200)]
Use simple concatenation in SseUtils.chunkOf()

Using a simple concatenation improves performance. It also eliminates
the need for @SuppressFBWarnings and making transport-http independent
of spotbugs-annotations.

Change-Id: I2ab11adc679f01b217b21ab847c3e3a96fbdc99b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoEnforce strict stubbing in transport-tcp 37/113437/1
Robert Varga [Thu, 5 Sep 2024 09:04:24 +0000 (11:04 +0200)]
Enforce strict stubbing in transport-tcp

Make sure our mocks are throughly stubbed by pulling in
mockito-configuration.

Change-Id: I2ee6f13fa9adeab187a24698ed313f9f733b733f
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoSeparate out Http2SettingsFilter 30/113430/1
Robert Varga [Wed, 4 Sep 2024 17:43:50 +0000 (19:43 +0200)]
Separate out Http2SettingsFilter

Http2Utils.clientSettingsHandler() is a stateless thing. Separate it out
into a dedicated class, which has a singleton instance shared across all
channels that need it.

Change-Id: I5136b43286c01ad6caa4a3ec11a814e239217dcd
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoSplit up HTTPServer 29/113429/1
Robert Varga [Wed, 4 Sep 2024 17:27:48 +0000 (19:27 +0200)]
Split up HTTPServer

HTTPServer is either backed by TCP or TLS, each of which has a different
flow setup. Specialize the two so we do not need to muck around with
SslHandler.

Change-Id: I91563aad1c3c924a5cb63a47c76df8e706e11a1c
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoClean up HTTPTransportStack, part two 24/113424/3
Robert Varga [Wed, 4 Sep 2024 12:09:44 +0000 (14:09 +0200)]
Clean up HTTPTransportStack, part two

Eliminate HttpChannelInitializer, as both implementations can easily be
integrated into HTTP{Client,Server}.

This also eliminates needless juggling of channel pipeline, as the
channel is always functional by the time we are initializing.

Change-Id: I228de5c6e5bbb301b2984c62a465900e3ecb3a22
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoClean up HTTPTransportStack, part one 23/113423/2
Robert Varga [Wed, 4 Sep 2024 11:30:25 +0000 (13:30 +0200)]
Clean up HTTPTransportStack, part one

A number of cleanups:
- channelInitializer should be private
- naming should be consistent
- superfluous annotations should not be present

Change-Id: I2843f9967ad85c30df366cbc81adc5dcbbb1f89d
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoExpose {SSH,TLS}TransportChannel 26/113426/1
Robert Varga [Wed, 4 Sep 2024 13:00:43 +0000 (15:00 +0200)]
Expose {SSH,TLS}TransportChannel

Knowing the identity of underlying TransportChannel is needed to
accurately report their type. Expose the two implementations, as
{TCP,HTTP}TransportChannel are already exposed.

Change-Id: Iaaca2b92b1df9bedbb278f8c0f0e9d48c617e3d6
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoClean up HTTPTransportChannel 22/113422/1
Robert Varga [Wed, 4 Sep 2024 11:29:33 +0000 (13:29 +0200)]
Clean up HTTPTransportChannel

This class should be final and should only be instantiated from the same
package.

Change-Id: I3261f983b8db52b5405f31b41a3adb69e4c352d6
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoUse enhanced switch for transport dispatch 19/113419/1
Robert Varga [Wed, 4 Sep 2024 09:13:02 +0000 (11:13 +0200)]
Use enhanced switch for transport dispatch

We have two supported possibilities, use a switch to discern them.

Change-Id: Ica6c7529aad0366caf8c413bc58a5ddc63481e5b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoOpenAPI: add metadata endpoint 52/111452/35
lubos-cicut [Thu, 8 Aug 2024 08:57:25 +0000 (10:57 +0200)]
OpenAPI: add metadata endpoint

Added endpoints for pagination metadata.

/single/meta(?offset=0&limit=5)
/mounts/{instance}/meta(?offset=0&limit=5)

Created MetadataStream and MetadataEntity to provide metadata
which would be valuable for end users.

JIRA: NETCONF-1344
Change-Id: I14e476ac4266e48953e9faaca80782fa891d03d3
Signed-off-by: lubos-cicut <lubos.cicut@pantheon.tech>
Signed-off-by: Ivan Hrasko <ivan.hrasko@pantheon.tech>
2 months agoBump versions to 8.0.2-SNAPSHOT 01/113401/1
Robert Varga [Tue, 3 Sep 2024 17:56:36 +0000 (19:56 +0200)]
Bump versions to 8.0.2-SNAPSHOT

This starts the next development iteration.

Change-Id: Ie0a8f46a47852bb05891a6ff422feb7360a110e7
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoRelease netconf v8.0.1
jenkins-releng [Tue, 3 Sep 2024 16:36:25 +0000 (16:36 +0000)]
Release netconf

2 months agoBump upstreams 92/113392/1
Robert Varga [Tue, 3 Sep 2024 12:35:08 +0000 (14:35 +0200)]
Bump upstreams

Adopt:
- odlparent-14.0.3
- infrautils-7.0.3
- yangtools-14.0.4
- mdsal-14.0.2
- controller-10.0.2
- aaa-0.20.1

Change-Id: I035b5e0002f01a54b0ec6971c33aa021aa757ee4
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoFix superfluous nullness annoutations 85/113385/2
Robert Varga [Tue, 3 Sep 2024 08:29:06 +0000 (10:29 +0200)]
Fix superfluous nullness annoutations

We have @NonNullByDefault -- there's no point in @NonNulls.

JIRA: NETCONF-1366
Change-Id: Ib579eddd14dba2ab2f05fdfac61fcb224730a4e2
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoAdd missing return 84/113384/2
Robert Varga [Tue, 3 Sep 2024 08:26:10 +0000 (10:26 +0200)]
Add missing return

We should not be invoking callbacks twice -- make sure to return.

JIRA: NETCONF-1366
Change-Id: If7d5e4c2791187e23b13ae0270e98a44faba5746
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoAdvertize :interleave capability 83/113383/2
Robert Varga [Tue, 3 Sep 2024 08:13:08 +0000 (10:13 +0200)]
Advertize :interleave capability

We support interleave, i.e. concurrent use of RPCs and notifications.
Advertize this fact.

JIRA: NETCONF-1374
Change-Id: I9028814398a8355d8f4a970ab5e7731d7983d78b
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>
2 months agoRevert "Check for netconf interleave capability" 82/113382/2
Robert Varga [Tue, 3 Sep 2024 07:23:30 +0000 (09:23 +0200)]
Revert "Check for netconf interleave capability"

This reverts commit c2cfd0608b41775f9e71904de14b453ad9d2ca59, as it does
not present a complete solution.

JIRA: NETCONF-1375
Change-Id: If7052257472a72f44f0e39401e630c691f13c64c
Signed-off-by: Robert Varga <robert.varga@pantheon.tech>