Age | Commit message (Collapse) | Author | Files | Lines |
|
Change-Id: I0db6ab5efb0595cba86ce9cbaaafab88349818ff
Issue-ID: POLICY-3696
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
Change-Id: Ie39d324e02a302b882318d46735a727322ce409d
Issue-ID: POLICY-3587
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
plugins-events thru plugins-persistence-jpa-eclipselink
Fixed:
- use "var"
- disambiguate method parameter types
- rename parameter
Issue-ID: POLICY-3093
Change-Id: Ife5897015be495403e731754d5862b803a217c87
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
1) Remove the usage of inputField and outputField from Task definition.
2) Fix issues around populating events to task in case of State
Finalizer Logic.
Change-Id: Ief17f400729410b83c6b7c665980d443e0cf6f28
Issue-ID: POLICY-3336
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
This review addresses two main changes:
1) inputFields and outputFields are not tied to task definition anymore.
Instead inputEvent and outputEvents associated to a task is populated
as part of the policy state definition, as the state definition have
the information anyway.
- Clean up of the usage of inputFields and outputFields in task
definition will happen in a future review
- inputFields and outputFields defined in task definition in
policies until honolulu will not make the policy invalid as the
changes are done in backward compatible way.
2) Multiple output events can come out of a final state now.
- Define another policy state output with the relevant eventName in
the command file
- In the task logic, create a map to store the fields of the relevant
outputEvent, and then just call
"executor.addFieldsToOutput(<the_map_of_fields>)"
These 2 steps are enough to send multiple events to relevant
components as per the apex configuration.
Change-Id: Id88ca402704106404f529e595e1a76f6bf167876
Issue-ID: POLICY-3336
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
Issue-ID: POLICY-3105
Change-Id: I53abd7e9cf8c1425d115b0bb917db6389e9ded92
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Change-Id: I291f968367b1f864bb417a72fd0bdb22a07ba0f7
Issue-ID: POLICY-3127
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
Change-Id: I761ded255a97128f858d3a7ce95f2479b6208a4e
Issue-ID: POLICY-3088
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
Addressed the following issues:
- initialize mocks before use
- use parameterized queries
- Random() is not secure
- provide parameterized type for generics
- unused imports
- constructor visibility
- use compute() instead of containsKey()/put()
- make final fields static
- rename constants to all upper case
- no assert() in Thread.run() methods
- nested try
- nested if/else
- too many break/continue
- use try-with-resources
- repeatable annotations
- overlapping characters in reg ex
- hashcode is not sufficient in compareTo()
- need equals() with compareTo()
- make class an interface
- use parameterized test
- multiple calls in assert()
- log or re-throw
- use different type of lambda
- use parameterized logging
- use StringBuilder instead of concatenation
- use StandardCharsets.UTF_8
Issue-ID: POLICY-2906
Change-Id: I2cf8c885e3e22c2c6cbe6403a34906928afad022
Signed-off-by: Jim Hahn <jrh3@att.com>
|
|
Issue-ID: POLICY-2875
Change-Id: Ibb0e9a243bf92f0255cb5c145d31ccddb988fdf5
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
Change-Id: I5e2cd92ee3620dea4070d9fa3aeb5ffd65f9412c
Issue-ID: POLICY-2869
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
|
|
Apex documentation has now all been ported to use the ONAP recommended
rst format. This review removes the old asciidoc documents.
Issue-ID: POLICY-2824
Change-Id: I562bd344cb7d6ff36e7d54bdb8f95e3b656468f8
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Issue-ID: POLICY-2839
Change-Id: Ia7c847b3dafd28958d5df209af95cac19fa480a0
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
Change-Id: Idde5050b425eb1bebe312e80e795c21513367a1b
Issue-ID: POLICY-2789
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
Legacy format support is removed, and Tosca format support is added.
Change-Id: I3cfc181ccb5471a5d224c0162af18c1fa0fdbc70
Issue-ID: POLICY-2812
Signed-off-by: a.sreekumar <ajith.sreekumar@bell.ca>
|
|
|
|
Removal of remaining bugs in apex-pdp
Issue-ID: POLICY-2774
Change-Id: I370796e7c8a41c116155155c33176512e9e0cd91
Signed-off-by: waynedunican <wayne.dunican@est.tech>
|
|
Issue-ID: POLICY-2654
Change-Id: Ifd391ea9141d37cfd3ca4110d66c667962230ae1
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Issue-ID: POLICY-2799
Change-Id: Ifdc16e99093995665db61329f61455fc0f327386
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
|
|
|
|
Replaced try/catch blocks with assertj assertions in apex-pdp. Last
batch of changes
Issue-ID: POLICY-2451
Change-Id: I39bd02fdbd8389818dcf1b786189c1e344ddcea5
Signed-off-by: waynedunican <wayne.dunican@est.tech>
|
|
Removed try/catch blocks in apex-pdp with assertj assertions Part IV
Issue-ID: POLICY-2451
Change-Id: I4b6accb9f944195db9ccf8a5de165b169301ac7f
Signed-off-by: waynedunican <wayne.dunican@est.tech>
|
|
apex-pdp had an in-line docuemntation build system built on asciidoc. It
has proven to be very difficult to maintain and keep up to date and it
is incompatible with the ONAP official RST/ReadTheDocs docuementation
system.
In the Dublin release, we converted the asciidoc documentation to RST
format, and hae been updating the RST documentation since. The asciidoc
docuemntation system is now so out of date that it is being removed.
In this change, the "site" directory is removed from all modules in
apex-pdp and a profile called "apexSite" is removed from the pom.xml
file in all modules.
Issue-ID: POLICY-2686
Change-Id: I645f3ed101159d356cd90e619cef791067ad5eb5
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Issue-ID: POLICY-2695
Change-Id: I15b489de9ca2614e3a6c94b450a3d6397d17d4ff
Signed-off-by: Ram Krishna Verma <ram_krishna.verma@bell.ca>
|
|
This change is basically a rollback of https://gerrit.onap.org/r/c/policy/apex-pdp/+/104230, with minor refactoring.
The problem that 104230 tried to address was supposed to be caused by https://jira.onap.org/browse/POLICY-2463.
With this rollback, apex-pdp increases performance and reduces memory usage.
Issue-ID: POLICY-2572
Signed-off-by: huaxing <huaxing.jin@est.tech>
Change-Id: I3a52d0ead0c1e530d218f9e3aba8b3eb5558abaf
|
|
Issue-ID: POLICY-2387
Change-Id: I6d9222cfe8841a4a4532d95dbc6f8b3f3d15ac87
Signed-off-by: RossC <ross.carter@est.tech>
|
|
Issue-ID: POLICY-2188
Change-Id: I56bbf3aa3bd9e1ef433cc93ad2be0c54d141a2a5
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Issue-ID: POLICY-2514
Change-Id: I764cfe41ed8cf8ffebd4ff076d0b76b79d61e142
Signed-off-by: RossC <ross.carter@est.tech>
|
|
Issue-ID: POLICY-2514
Change-Id: I5b130b052150eb0e51759c751b83ab9d6f199759
Signed-off-by: RossC <ross.carter@est.tech>
|
|
Change-Id: Ic668ab20498bdc41edc78f52c7ce9d89a1385f19
Issue-ID: POLICY-2378
Signed-off-by: a.sreekumar <ajith.sreekumar@est.tech>
|
|
All the Kafka components need something near 10 seconds to come up
completely. This review tweaks the timing to allow the test Kafka server
to come up and to allow the consumers to connect to it.
Issue-ID: POLICY-2106
Change-Id: I6dd8ace0848bdc2549e658ef8908b4d85d5ea789
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Issue-ID: POLICY-2106
Change-Id: I7bd44f8dfe349d73a4c8c4f978a30b455c45b2c7
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Added tests to check threads start, execute, and stop correctly and
fixed JavescriptExecutor class to handle startup and shutdown correctly
without hanging.
Issue-ID: POLICY-2106
Change-Id: I9ab41023aae2ab1cbcaea53fdf5d48eccd90a2f1
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Change-Id: I59b93fccb0f3a41f8db753d1e8e2978bbb51e9a0
Issue-ID: POLICY-2378
Signed-off-by: a.sreekumar <ajith.sreekumar@est.tech>
|
|
This change ignores the tests that are hanging in Jenkins until the Apex
images are released. Once the images are released we'll fix these tests.
Issue-ID: POLICY-2106
Change-Id: I6a733896da044d3f24443ac202d6122a341db135
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
It is possible that the initiation and especially shutdown happens so
fast that the execution thread counts down the latch before the caller
calls await(), in which case the Javascript Rhino thread will hang. This
review adds a timeout on the await calls.
Issue-ID: POLICY-2106
Change-Id: Icfc6d4b478cca62522461c25d61ce9ee7328684a
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
When a policy is loaded, a separate thread is spawned for each
Javascript script executor. This allows us to precompile the Javascript
scripts and also to have a larger stack available for script execution.
Issue-ID: POLICY-2106
Change-Id: I97323aafb623ba537ac1889b3c9504b345b4f67e
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
All examples are converted to Rhino Javascript from Nashorn Javascript.
Issue-ID: POLICY-2106
Change-Id: Ia6fec5b57b1e6131799b4b069968e4766ce54c96
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Integation unit tests resultd in StackOverflowException errors in the
Graal Javascript interpreter. Following extensive testing and
trouobleshooting, it proved very difficult to fix these issues in Graal
as the stck overflow errors were occurring in native class instantiation
methods being invoked by Grall on the JVM.
The alternative Rhino Javascript engine is developed by mozilla, and was
incorporated into the Java 6 JVM and evolved into Nashorn in the Java 8
JVM. Oracle dropped Nashorn in Java 11. However, in parallel, Rhino
development has continued.
This review brings in the Rhino javascript engine into apex-pdp as the
replacement for Nashorn and instead of Graal. Graal seems to be pretty
unstable as yet so we may bring it in in future releases but for now
Rhino is a more stable and reliable alternative.
Issue-ID: POLICY-2106
Change-Id: I0edeff3b0bee404b38e3ebe22001a6e3375a44dc
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
use common version
Issue-ID: POLICY-2381
Change-Id: Id231aa8709b3c4515fcac80b82d53fb8ed45eaf0
Signed-off-by: Pamela Dragosh <pdragosh@research.att.com>
|
|
This is the main review to switch out the Nashorn Javascript interpreter
and switch in the Graal javascript interpreter for apex Javascript
tasks.
The Graal Javascript dialect differs slightly from Nashorn in the way it
accesses information passed over from the Java world so Javascript
tasks need to be converted.
Subsequent reviews will convert the JUnit tests marked @Ignore in
previous reviews to this one and will convert the Javascript tasks into
a format Graal can handle. Subsequent reviews will also convert the
examples over to Graal format.
Issue-ID: POLICY-2106
Change-Id: Ief582a36539e8a87724c17fb7e56864d5e471c07
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
A number of Java 11, checkstyle, and SONAR warnings have crept into the
Apex codebase over the last number of reviews. This change fixes those
issues.
Issue-ID: POLICY-1913
Change-Id: I2afd607e80f48323355380fb2fe5e048e18879f9
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
The new version of JRuby fixes this issue. Also amended unit test to
check for shutdown and immediate recreation of JRuby interpreter.
Issue-ID: POLICY-1276
Change-Id: I723e0396985d3163b483e52fdaceb4b4fab7274b
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
|
|
The Jython interpreter pulls in many dependencies that are flagged as
having security problems. We will disable Python support in apex-pdp
until Frankfurt, users can compile from source if they wish to re-enable
python support.
Issue-ID: POLICY-1509
Change-Id: I84215ba8d1009cb7af99cd98ba19008fdff5c378
Signed-off-by: liamfallon <liam.fallon@est.tech>
|
|
Also added defaultbranch to the .gitreview file to make it easier
to change it when apex-pdp is branched in the future.
Issue-ID: POLICY-2045
Signed-off-by: jrh3 <jrh3@att.com>
Change-Id: Ib2570fd395aae38862299f2193b07ab06e32ace5
|
|
Issue-ID: POLICY-1965
Signed-off-by: jrh3 <jrh3@att.com>
Change-Id: Icd437f10c4d991030652ffa1abf48b7b5738145c
|
|
Added integration test that sets properties in a dummy plugin and amends
them in tasks in a policy. Variosu tests added to check combinations of
where properties are set in plugins or in tasks or both.
Implementaiton changed to:
- Always pass in a Properies object, the properties object coming into
the policy cannot be null because the task/TSL/SFL may wish to set it
- Fix a bug where the properties were not passed from the ApexEvent to
the engine event in the ApexEventUnmarshaller class
Issue-ID: POLICY-1743
Change-Id: I6aa152b28d46cf3cc6fa56a1a95b76a8e55f5a49
Signed-off-by: liamfallon <liam.fallon@est.tech>
|