Age | Commit message (Collapse) | Author | Files | Lines |
|
Issue-ID: POLICY-2514
Change-Id: Ie482926a56a5250be5d78a97aa7c1c19311a292c
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
Adding the actual sample responses to the SO simulator broke the
drools-apps junit for usecases. Fixed (in theory) by updating the
legacy SO actor to properly decode the finishTime. Refactored the
new SO actor, extracting the type adapter into its own class file
so it could be shared between the new and legacy actors.
Issue-ID: POLICY-2570
Change-Id: I061b603172440b1a91da16d09b4f2a0d289dfc41
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
The timestamp in SO responses appears to follow RFC-1123/RFC-822
format instead of ISO format.
Added a type adapter for SO. In Guilin, we may want to consider
combining the two type adapters for LocalDateTime so that either format
is accepted.
Also modified the SO simulator to return responses that are actual
samples from SO.
As part of that work, discovered that the legacy SO actor is not able
to parse these timestamps either. However, as that code is now
deprecated, left it as is.
Issue-ID: POLICY-2570
Change-Id: I322318d1007e36eef47bb8867fd8ed01cb60223a
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
When SO is polled for the result of a previous request, it
does not necessarily include the originally returned request
ID in the response. This causes the SO actor to generate a
"missing request ID in response" exception.
Modified the actor to only extract the request ID from the
first response and cache it for subsequeent responses.
Testing this required the SO simulator to be modified so that
it would return an INCOMPLETE on the initial request, forcing
the actor to poll until it returns a COMPLETE. Made this a
settable flag so that it could be enabled just to test the
SO actor without impacting other components (e.g., drools-apps,
CSITs).
Also fixed a couple of checkstyle issues in the simulators.
Issue-ID: POLICY-2568
Change-Id: Ifad8b3c0c2c0b03cb82da693c2cf5ced44ede105
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
The JAXB unmarshaller used by AaiCqResponse is not thread safe. As a
result, if two responses are decoded at the same time, it will generate
exceptions.
Issue-ID: POLICY-2556
Change-Id: I5e81d5f3c39b9ba321c146dc6f4f73e558aaca3e
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Change-Id: Ic3571df8ed874267fb803d1f384ec5cb02fb80bb
Issue-ID: POLICY-2524
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
Also point to parent, common snapshots
Issue-ID: POLICY-2510
Change-Id: Iad79c341bcfb8cb6bb828ee71e199ba3bb9a838b
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
|
|
Made the following updates:
- added new A&AI get-PNF Operation by refactoring AaiGetOperation,
separating out Tenant and PNF operations
- added PNF support to the CDS actor
- added logging to the CDS Handler
- added get-pnf to the A&AI simulator
Issue-ID: POLICY-2505
Change-Id: Iff140e7c864f762790d8e2ecaba62c161c859e6e
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Like the SDNR Operator, the CDS Operator should work regardless
of the operation name.
Issue-ID: POLICY-2504
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I9a77838462bc5b91e2d95e7dcf1f8822bd45f04c
|
|
Moved the spi file to register the actor and added a test to
ensure that the registration worked.
Also removed some deprecated code from the CDS junits.
Issue-ID: POLICY-2501
Change-Id: I06f50ea8f074cdb7a89c7e387f9f31ea7cffa7e2
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Issue-ID: POLICY-2378
Change-Id: I5461bafd3c2d19f4f457da0816bba117a2d8b550
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
Issue-ID: POLICY-2378
Change-Id: I51c6c57bdcc0299c25b5ee05f47c20f4542bc02d
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
|
|
|
|
Made the following changes:
- Enhanced SDNR Actor to support any operation name specified within
the policy, constructing the same request, but passing a different
"RPC name" and "Action" in the request.
- Added ControlLoopResponse to OperationOutcome
- Modified SDNR Actor to populate ControlLoopResponse
Issue-ID: POLICY-2468
Change-Id: I50ee0674077d975f3cd211454656edc47d78520f
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Addressed the following sonar issues:
- missing assert in junit
- duplicate code blocks
- fixed those in the simulators
- leaving the others in the concept classes for Liam/"G"
Issue-ID: POLICY-2305
Change-Id: Icfe80afb378b4cdec4de3d33d0e33227e04845d8
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Addressed the following sonar issues:
- use RE2 instead of java.util Pattern for "+" and "*"
- don't use deprecated methods
- for Date(long), sonar appeared not to parse the argument's
type correctly. Modified the code slightly to make sonar happy
- duplicate blocks of code
- either log or throw
- missing assert in junit
- for SDNR & VFC, eliminated threads, as they are unnecessary
- duplicate code block in different branches
- useless assignments
- redeclaring abstract methods
- cyclomatic complexity
- used lombok in some cases (e.g., EqualsAndHashCode)
- assert argument order
- actually deleted ControlLoopTargetType, because it is not needed
and sonar complains regardless of which order is used
- add private constructor to utility classes
- use StandardCharsets instead of literals
Also:
- added logback-test.xml to SO to eliminate the voluminous output
from the junit test
Issue-ID: POLICY-2305
Change-Id: I586c331781bedbd54a115a71847d04d293689445
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
|
|
Junits were failing in my local eclipse environment. Fixed the
following:
- more cases of "is marked @NonNull but is null" (I thought Liam had
already fixed all of these, but I'm still seeing errors, so fixed
some more)
- comparing "Instant" values in event time stamps assumes that
the times will be different, but may not be, if the JVM is fast
enough. Modified the test to set the times of the new events to
Instant+1, to ensure that they are different from the original
Issue-ID: POLICY-2305
Change-Id: Ic05bc6a58b559ba589583ce887711ee01f76691c
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Issue-ID: POLICY-2467
Change-Id: If06f5fbf26269f7a34104bb2194258494a6488bf
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
Modified new actor code to generate the sub request ID
before invoking the start callback.
Issue-ID: POLICY-2461
Change-Id: I4adabd6efda2c30c0e2da31f95f01dd5fe546c0a
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Also modified HttpParams to allow "path" to be blank to support any
cases that really have no "path" to append to the context URI base
path.
Issue-ID: POLICY-2405
Change-Id: I49eebde6759659d2804b5a11c1504c37674bd0c4
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Just some trivial sonar bugs for missing argument and a
couple sonar overrides to clear security hotspots.
Issue-ID: POLICY-2165
Change-Id: I392e5fee458465c66313a16a8912e23a3a119848
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
|
|
The new actor for legacy APPC attempts to discard request
messages received on the response topic by discarding those
that have a null response. Unfortunately, when it decodes them,
the Response object creates a status object, even if there isn't
one in the json message, thus the test fails and the actor treats
it as a failed response.
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I26cfb25443ff2ed394781d309f9097d50b8f160b
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Released 2.2.2
Issue-ID: POLICY-2378
Change-Id: I829f0dfc88aeaf1f9b7efad91c983628f2a5d706
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
|
|
|
|
Issue-ID: POLICY-2165
Change-Id: Iaad54af1b1210fe96fa44dcaa609834c46557877
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
|
|
Upgrading io.netty to 4.1.48.Final as the current version has security issues.
Issue-ID: POLICY-2442
Change-Id: I5dbfdc7890c85571ef6b6bcc7a9f95310868f6a0
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
Issue-ID: POLICY-2378
Change-Id: I644e45822154f57d2bf6a81e55df07e2e4ca6966
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Missing various fields within the request structure (e.g.,
ONAPName). Fixed.
In the process, also modified makeGuardPayload() so that it
only constructs the inner "guard" JSON object, making it easier
for invoking code to modify it.
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I7d34a279845bb98425caf66565eab7513d310815
|
|
SDNR v2.0 introduced an extra "input" and "output" layer within
the body of the messages. Added POJOs to match that, and updated
the actor to use the new POJOs so that no extra steps are needed
during serialization or deserialization.
Also modified the SDNR and APPC-LCM simulators to discard request messages
when looking for the response, if the sink and source topics happen to
be the same.
Added the SDNR simulator.
Added more coverage to the SDNR actor.
Fixed a sonar issue in the SDNR junit code:
- use a constant instead of a literal (i.e., for "ModifyConfig")
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I70914a299dd5e9ee8eaccb5d42a2596cab814091
|
|
|
|
|
|
Legacy APPC uses the same topic for both requests and responses.
Unfortunately, when the actor sees the request, thinking it's a
response, it throws an exception because it can't find the status
code. Fixed.
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I6137fb21dbf1149ff2faca2407ba69be0de6d40b
|
|
If the topic processResponse() method throws an exception, then
the actor/operation is left in an incomplete state.
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I6c5d149d4046fbfb970c8dd831fc3938516d1115
|
|
Fixed these issues:
- topics weren't started
- appc topics were reversed
- prevent appc simulator from responding to a response (i.e., infinite
loop)
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I1f43be72d35f874fec98f48f1a112c055c00eee5
|
|
|
|
Some CSITs may require multiple simulators. This adds a class that
will start all of the simulators. A tarball is generated from which
a docker image can be built.
Added simulators for Topics: appc and appc-lcm.
Fixed licenses in files in packages directory.
Fixed per review comments:
- add version to Dockerfile
Issue-ID: POLICY-2434
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: Id7aa9cb5a5874f7b4185273ab0d2c074198554ff
|
|
|
|
Issue-ID: POLICY-2389
Signed-off-by: Chris Ramstad <cramstad@ciena.com>
Change-Id: I7ce91e331a128aaadaab5ac837e19873a3388164
|
|
Unfortunately, some code uses "new Yaml(...)" to decode policies and
"Yaml" doesn't understand GSON annotations. As a result, this means
that the field names must match the serialized name, thus they must
continue to end with "UUID" instead of "Uuid".
Issue-ID: POLICY-1910
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I97c2f5c7e59b758b13a1dbcceed709b2e2e064a2
|
|
A few minor eclipse warnings and checkstyle warnings in policy models.
Issue-ID: POLICY-1910
Change-Id: I612c6514fa7d88cb8658a7ee1618869a525e48d7
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
|
|
|
|
Adding first cut at VFC Actor.
Issue-ID: POLICY-2386
Change-Id: I4749ed3f350f01d05026057ffe4149f3843d8475
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
Signed-off-by: Peyton Puckett <peyton.puckett@att.com>
|
|
Issue-ID: POLICY-2371
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I42b6980460c2075544eb7698479b386eff8ef7b5
|
|
Modified the Guard Actor to check the "disabled" flag during
start() instead of waiting until it gets down into startOperationAsync.
Issue-ID: POLICY-1625-guard-disabled
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: Id6ccf506b82504552d6e1f4637ad11e2e41e77e5
|
|
Rule notifications need the subrequest ID to be populated in the
OperationOutcome object, where possible.
Issue-ID: POLICY-2385
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: Ic44320f67ad8df7bcb3000cfa667f95427818e71
|
|
Actors only log the URI, without the host and port info. Modified
to log the full URL.
Also adjusted URLs for trailing "/".
Issue-ID: POLICY-1625
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I0c9a2d139258ef23bc363a873b56f5cb4680247d
|
|
Also added legacy ModifyConfig as an operation within APPC-LCM.
Added logging to topic registration keys.
Issue-ID: POLICY-2403
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: Ia54a573fd6218a8afe870184b9a3baebc05b766a
|
|
Redesigned the SO Operation classes; moved some code from the subclass
to the superclass so it could be reused by the VF Module Delete Operation.
JerseyClient does not support DELETE with a request body, so had to
implement a delete() method using java11 HttpClient.
Fix some issues found while testing with drools-apps.
Added "delete" operation to SO simulator.
Issue-ID: POLICY-2371
Signed-off-by: Jim Hahn <jrh3@att.com>
Change-Id: I269fe13cf90c295ec2bbac92bc5a59b3820ea265
|