Documentation of RAN PM
Added documentation for ICS.
Signed-off-by: PatrikBuhr <patrik.buhr@est.tech>
Issue-ID: NONRTRIC-868
Change-Id: I9a1597c0deab511d9d652acdf43b4280699a9a0c
diff --git a/api/ics-api.json b/api/ics-api.json
index cd3b61b..346ef46 100644
--- a/api/ics-api.json
+++ b/api/ics-api.json
@@ -240,7 +240,7 @@
"required": ["info_job_identity"],
"properties": {
"owner": {
- "description": "The owner of the job",
+ "description": "The owner of the job. This is a string that indentifies the job owner, which could be an application, a POD or something else.",
"type": "string"
},
"last_updated": {
@@ -252,7 +252,7 @@
"type": "string"
},
"target_uri": {
- "description": "URI for the target of the produced Information",
+ "description": "URI for the target of the produced Information. Note, this is deprecated and will be removed. The information on how the data is delivered is type specific and should be defined in the type specific info_job_data.",
"type": "string"
},
"info_job_data": {
diff --git a/api/ics-api.yaml b/api/ics-api.yaml
index 4a8d3bd..6762bc5 100644
--- a/api/ics-api.yaml
+++ b/api/ics-api.yaml
@@ -1562,7 +1562,8 @@
properties:
owner:
type: string
- description: The owner of the job
+ description: The owner of the job. This is a string that indentifies the
+ job owner, which could be an application, a POD or something else.
last_updated:
type: string
description: The time when the job was last updated or created (ISO-8601)
@@ -1571,7 +1572,10 @@
description: Identity of the Information Job
target_uri:
type: string
- description: URI for the target of the produced Information
+ description: URI for the target of the produced Information. Note, this
+ is deprecated and will be removed. The information on how the data is
+ delivered is type specific and should be defined in the type specific
+ info_job_data.
info_job_data:
type: object
description: Json for the job data
diff --git a/docs/overview.rst b/docs/overview.rst
index 4bdb50f..27e3767 100644
--- a/docs/overview.rst
+++ b/docs/overview.rst
@@ -1,6 +1,6 @@
.. This work is licensed under a Creative Commons Attribution 4.0 International License.
.. SPDX-License-Identifier: CC-BY-4.0
-.. Copyright (C) 2021 Nordix
+.. Copyright (C) 2021-2023 Nordix
Information Coordination Service
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
@@ -31,22 +31,75 @@
Information Jobs and types are stored persistently by ICS in a local database. This can be either using Amazon S3 - Cloud Object Storage or file system.
To restrict which data that can be consumed by by whom there is support for finegrained access control. When data subscriptions/jobs are modified or read, an access check can be performed.
-ICS can be configured to call an external authorizer.
+ICS can be configured to call an external authorizer.
This can be for instance Open Policy Agent (OPA) which can grant or deny accesses based on an access token (JWT) used by the calling data consumer.
In addition to this the information type, accesstype (read/write) and all type specific parameters can be used by access rules.
The URL to the authorization component is defined in the application.yaml file and the call invoked to by ICS is described in API documentation.
+
+****************
+Key Requirements
+****************
+
+* Multiple producers, posibly from different vendors, should be able to
+ produce the same type of data.
+
+* The system should allow for arbitrary installation, upgrade, scaling
+ up, scaling down, and restarting of software, in any order.
+
+* ICS should decouple data producers from data consumers. Consumers
+ should not be aware of producers, their endpoints, or the number of
+ producers. Upgrades, scaling, and other changes to producers should not
+ affect consumers. That also means that a producer can be replaced by
+ another one.
+
+* Data consumers should be able to start subscriptions at any time,
+ including before a producer is installed, or during a producer's
+ upgrade, restart, or scaling up/down.
+
+* The system should impose as few restrictions as possible on how
+ software is implemented. Applications may, for example, be implemented
+ as multiple identical executing entities for load sharing.
+
+* The lifecycle of a data type in the system should not be controlled
+ by a single producer. There may be data types and subscriptions for
+ data types that have no producers, such as during an upgrade scenario.
+
+* A data type specification should contain all the information needed
+ to implement both producers and consumers. The data type specification
+ forms the API between these. It must define the ID of the data type,
+ all possible parameters used at subscription creation (such as filter
+ constructions), and all details about the delivered format and the
+ delivery mechanism. This is comparable to a Managed Object Model.
+
+* ICS should mediate data type versioning and choose compatible data
+ producers. For example, if there are two producers, one producing
+ dataType version 1.5.0 and another producing version 1.4.0, and a
+ consumer subscribes to data version 1.1.0, both these producers should
+ be involved (since version 1.4.0 and version 1.5.0 are backwards
+ compatible with version 1.1.0). The system should be able to handle
+ these versioning scenarios automatically.
+
+* A consumer shall be able to retrieve its jobs after a restart. Therefore it must be possible
+ to group the jobs based on a label "owner" which is define by the consumer. This must be unique,
+ which suggests that it should be based on UUIDS, but this is up to the consumer.
+ This "owner" can be a POD, an application etc. ICS should not restrict that.
+
*********************
Summary of principles
*********************
* ICS provides APIs for control of data subscriptions, but is not involved in the delivery of data. This means that any delivery protocol can be used.
* Data for one Information type can be produced by many producers from different vendors.
-* Data filtering is done at the producer.
+* Data filtering is done by the producer. ICS does not restrict how data selection/filtering is done.
* A Data Consumer can create a data subscription (Information Job) regardless of the status of the data producers. The producers can come and go without any need for the Data Consumer to take any action.
A subscription indicates the need for a type of data and the system should do its best to fulfill this.
-
+* ICS is by design not aware of any subscribeable data types.
+* When a consumer creates a subscription/jonb, ICS shall choose the information type version with the lowest available
+ compatible version. All producers that has registered a type that is compatible with the chosen version are included.
+ Example, if a consumer creates a job with type version 1.1.0, the chosen type may 1.2.0 and a producer
+ supporting version 1.9.0 will be included (but not a producer that supports version 2.0.0).
*******
Example
@@ -67,9 +120,17 @@
If a new producer is started, it will register itself and will get notified of all jobs of its supported types.
-
**************
Implementation
**************
Implemented as a Java Spring Boot application.
+
+*************
+Configuration
+*************
+
+The component is configured by the usual spring boot application.yaml file.
+
+An example application.yaml configuration file: ":download:`link <../config/application.yaml>`"
+
diff --git a/docs/requirements-docs.txt b/docs/requirements-docs.txt
index f266a10..3ecb51e 100644
--- a/docs/requirements-docs.txt
+++ b/docs/requirements-docs.txt
@@ -9,4 +9,5 @@
sphinxcontrib-swaggerdoc
sphinx_bootstrap_theme
sphinxcontrib-redoc
-lfdocs-conf
\ No newline at end of file
+lfdocs-conf
+urllib3~=1.26.15
diff --git a/src/main/java/org/oransc/ics/controllers/r1producer/ProducerJobInfo.java b/src/main/java/org/oransc/ics/controllers/r1producer/ProducerJobInfo.java
index bbaa438..6dc9717 100644
--- a/src/main/java/org/oransc/ics/controllers/r1producer/ProducerJobInfo.java
+++ b/src/main/java/org/oransc/ics/controllers/r1producer/ProducerJobInfo.java
@@ -27,9 +27,7 @@
import org.oransc.ics.repository.InfoJob;
-@Schema(
- name = "producer_info_job_request",
- description = "The body of the Information Producer callbacks for Information Job creation and deletion")
+@Schema(name = "producer_info_job_request", description = "The body of the Information Producer callbacks for Information Job creation and deletion")
public class ProducerJobInfo {
@Schema(name = "info_job_identity", description = "Identity of the Information Job", required = true)
@@ -47,12 +45,12 @@
@JsonProperty("info_job_data")
public Object jobData;
- @Schema(name = "target_uri", description = "URI for the target of the produced Information")
+ @Schema(name = "target_uri", description = "URI for the target of the produced Information. Note, this is deprecated and will be removed. The information on how the data is delivered is type specific and should be defined in the type specific info_job_data.")
@SerializedName("target_uri")
@JsonProperty("target_uri")
public String targetUri = "";
- @Schema(name = "owner", description = "The owner of the job")
+ @Schema(name = "owner", description = "The owner of the job. This is a string that indentifies the job owner, which could be an application, a POD or something else.")
@SerializedName("owner")
@JsonProperty("owner")
public String owner = "";
@@ -63,7 +61,7 @@
public String lastUpdated = "";
public ProducerJobInfo(Object jobData, String id, String typeId, String targetUri, String owner,
- String lastUpdated) {
+ String lastUpdated) {
this.id = id;
this.jobData = jobData;
this.typeId = typeId;
@@ -74,7 +72,7 @@
public ProducerJobInfo(InfoJob job) {
this(job.getJobData(), job.getId(), job.getType().getId(), job.getTargetUrl(), job.getOwner(),
- job.getLastUpdated());
+ job.getLastUpdated());
}
public ProducerJobInfo() {
diff --git a/tox.ini b/tox.ini
index 859612e..6bb5116 100644
--- a/tox.ini
+++ b/tox.ini
@@ -16,7 +16,7 @@
# documentation only
[tox]
-minversion = 2.0
+minversion = 4.0
envlist =
docs,
docs-linkcheck,
@@ -28,6 +28,8 @@
commands =
sphinx-build -W -b html -n -d {envtmpdir}/docs/doctrees ./docs/ {toxinidir}/docs/_build/html
+allowlist_externals = echo
+
[testenv:docs-linkcheck]
basepython = python3