Modification:

- introduce an anotation called ForeignKey. This can be use to ensure
    that the uuid reference in another entity is valid and cascading
    effect are maintained.
- modify some entities and added some new entities like Image, Person
    and Software. Also, corresponding repositories and services.
- redesign the base service and create a kind of base repository

Target: master

Motivation:

We need to store media data for some of the entities and neo4j is
not design to store images, video, etc.. Hence, to avoid using
multiple dbs, we sort for a solution that can store the metadata
and the media data.

Modification:

- remove neo4j dependency and add mongo
- change the entities details and re-design it to suit the new db
- adjust the api design
- delete all neo4j relationship pojos
- adjust the tests according to the new definitions

Result:

Cerebrum now uses mongo db.

Target: master

Motivation:

Meeting up with the new API design requires that the market-user
and organisation entity are realign.

Modification:

- factor out some part of market-user and create a new class (person)
- make organisation enity to extend audit-metadata and some properties
    were added.
- use the lombok project to reduce the boilerplate code in both
    the market-user and organisation entities.
- add 'person' to the list of PrefixEnum in the CerebrumEntityUuidGenerator

Result:

Market-user and Organisation entity are now re-worked

Target: master

Motivation:

Some reported vulnerabilities of the dependencies by OWASP make
the pipeline broken.

Modification:

- upgrade spring boot starter parent version from 2.3.5.RELEASE
    to 2.3.10.RELEASE
- upgrade hibernate-validator from 6.0.19.Final to 6.2.0.Final
- updgrade neo4j-java-driver-spring-boot-starter from 4.1.1.0
    to 4.2.4.0
- update list of OWASP check and supress some false positive
    alarms

Result:

Build pipeline succeeds.

Target: master

Motivation:

To gain insight on cerebrum-app, the spring framework provide
the facility to expose the application metrics to monitoring
service like Prometheus on Kubernetes.

Modification:

- enable graceful shutdown
- add spring boot actuator and prometheus registry micrometer
    dependencies
- expose the actuator endpoint that will provides: health,
   info, metric etc.
- avoid 404 error on request to /favicon.ico

Result:

Cerebrum now makes it metrics available and can be shutdown
gracefully.

Target: master

Motivation:

To fullfill the service metadata description defined by the
cloud management board, new properties need to be added to
the market-service entity.

Modification:

- add lombok dependency to the cerebrum project to avoid some
    boilterplate codes
- add some few new properties to the market-service entity
- enable database auditing to keep track of when an entity was
    created and when it was last modified.
- adjust the MarketServiceRepository to fit the new market-service
    properties definition

Result:

Market-service entity contain some new properties and some few name
change.

Target: master

Motivation:

Automated ci/cd is desirable process we will like to adhere to in 
our pipeline. Since we already have a local kubernetes cluster running 
on the production machine, we need to ensure that when a new tag
is release, the cluster is notified.

Modification:

- add a new pipeline stage called `docker`. This will be use for 
the building and deploy the image into the docker hub.
- deploy stage will now be responsible for trigger automated update 
of the image in the cluster.

Result:

Getting closer to fully automated ci/cd

Target: master

Result: avoid `pseudo-terminal will not be allocated because stdin is not a terminal` error.

Motivation:

Since we are moving to dockerize deployment, some variables,
like the neo4j hostname needs to be set different base on
the environment set up.

Also, we need to seperated the images that will be generated
when a normal commit is made and when a release is trigger.

Modification:

- expose the neo4j host name as configurable variable
- seperate docker image generated from a commit and a tag
    release
- add production_deployment to ci/cd deploy stage. This
   will be use to depoly to the production system. However,
   it just include a test to see if the ssh will work.

Result:

Two different docker images will be build, one for stable
tagged release and the other for development. These images
will be push to different docker repositories.

Target: master

Motivation:

A new uuid is generated for an entity that will be save on the db..
This is use as a primary way of uniquely identify a particular node.
To ensure the uniqueness of the uuid, two factor were considered
which are:
- uuid version 1
- secure random number
However the secured random number is currently generated causing the
thread to block in a VM (for details explaination see
https://tersesystems.com/blog/2015/12/17/the-right-way-to-use-securerandom/).
Hence the behaviour of the system become unpredicate and most likely
will fail if the request will required the uuid to be generated.

Modification:

- change how the secured random number is generated
- minor typo fix inside the MarketUserRepository

Result:

The generation of cerebrum custom uuid now work seemly on both local
machine and VMs. Hence, all requests are now process as expected.

Target: master

Motivation:

Keep cerebrum up-to-date.

Modification:

- change the spring-boot version to 2.3.5
- upgrade the springdoc-openapi-ui to version 1.4.8 to
    avoid the classpath problem
- use the latest spring boot version of the neo4j driver
- unity aai jwt header token is at+jwt however, the latest
    springframework security does not support this. Hence.
    exclude the current version of nimbus-jose-jwt and
    add version 7.9 that support at+jwt header.
- inside the application.yml and application-dev.yml use the
    properties setting for the neo4j instead of the spring
    neo4j properties
- enable more neo4j logs info
- use the proper variable name for the sceret value inside
    the docker file.

Result:

No visible changes to the end user

change to debug mode

`tar` -> `jar`

Motivation:

When the test stage of the ci/cd is trigger, it currently fail. 
The main reason is that it took more than the stipulated time of 
one hour to run.

Modification:

skip unit/integration tests

Result:

no testing

Motivation:

Maven test are unncessarily ran at least twice and some
stage that were declared that were not use. This cause
some unexpected behaviour in the ci/cd build.

Also, the ci/cd fail because of some false postive in the
dependency vulnerabilities scan

Modification:

- Define only neccessary stages and use them accordingly.
- add some few dependecies supression

Result:

- Keep the ci/cd simple and avoid running multiple tests
- when `mvn org.owasp:dependency-check-maven:check` is run
Vulnerabilities found equal 0

Target: master
Review: https://gitlab.hzdr.de/hifis-technical-platform/helmholtz-cerebrum/-/merge_requests/39

Motivation:

All authenticated user should be store into cerebrum
data-storage (neo4j). At the momement this is not the
case.

Modification:

- ensure that when the autheticated user hit /api/v0/users/whoami
    and the user is unknown to cerebrum, a new user will
    be created.
- add findByEmailAndSub method in the user repository to search
    the database if user with the supplied email and sub exist
- add a createEnity method inside CerebrumService interface that
    are implemented inside CerebrumServiceBase abstract class.
    This will be use when the uri location is not needed.

Result:

Unkown user that have been authenticated will be
store by cerebrum.

Target: master
Review: https://gitlab.hzdr.de/hifis-technical-platform/helmholtz-cerebrum/-/merge_requests/38