Background
inQuiry BB has multiple components like assessment-service and Flink jobs spread across multiple GitHub repositories. The objective is to decouple the inQuiry specific code from KP (Knowledge-Platform) Git Repos and bring it under the inQuiry BB space in GitHub.
Current Code Structure:
Repo: github.com/project-sunbird/knowledge-platform
/knowledge-platform/assessment-api (assessment-service specific code)
/assessment-api/ assessment-service (play application)
/assessment-api/ assessment-actors (actor implementation)
/assessment-api/ qs-hierarchy-manager (hierarchy-manager)
/knowledge-platform/ontology-engine (assessment-service uses entire ontology-engine)
/ontology-engine/graph-common
/ontology-engine/graph-core_2.11
/ontology-engine/graph-dac-api
/ontology-engine/graph-engine_2.11
/ontology-engine/parseq
/knowledge-platform/platform-core (assessment-service uses entire platform-core directly or via ontology-engine)
/platform-core/actor-core
/ontology-engine/schema-validator
/ontology-engine/platform-telemetry
/ontology-engine/platform-common
/ontology-engine/platform-common
/ontology-engine/cassandra-connector
/ontology-engine/kafka-client
/knowledge-platform/platform-modules (assessment-service uses only import-manager module)
/platform-modules/import-manager
/platform-modules/mimetype-manager
/platform-modules/url-manager
Repo: github.com/project-sunbird/knowledge-platform-jobs
/knowledge-platform-jobs/jobs-core (common module for all flink jobs)
/knowledge-platform-jobs/auto-creator-v2 (inQuiry BB job)
/knowledge-platform-jobs/publish-pipeline
/publish-pipeline/publish-core (common module for other publish job)
/publish-pipeline/questionset-publish (inQuiry BB job)
Dependency with Knowlg BB:
Currently, all common modules are part of knowlg BB and are tightly coupled with knowlg/inQuiry services/jobs.
Open Questions:
Is separate infra monitoring (e.g: monitoring service, grafana dashboard) will be installed for inQuiry?
How load testing will be performed for inQuiry components? Do we get common load test infra?
Currently, all assessment APIs are onboarded with
ContentRole (e.g: Content Create or Content Update). Should we change the API roles to Question/QuestionSet Role (e.g: Question Create)?
Note: this will have an impact on SunbirdEd/Diksha/etc)What will happen to service telemetry (e.g: info/error logs) dashboard such as Graylog? Do we need it? If yes, what all services/components need to be installed?
Proposed inQuiry BB Changes:
Proposed Git Repos:
Sunbird-inQuiry/quml-service (
for assessment-service)
/quml-service/assessment-service
/quml-service/assessment-actors
/quml-service/qs-hierarchy-manager
/quml-service/build (build script)
/quml-service/schema (question & questionset schema)
/quml-service/definition (sunbird supported primary category & its definition script for quml)
/quml-service/tests (API Functional Test Script/Code)Sunbird-inQuiry/quml-jobs (
for backend jobs)
/quml-jobs/questionset-publish (for Question/QuestionSet Publish)
/quml-jobs/auto-creator-v2 (for import of Question/QuestionSet)Sunbird-inQuiry/quml-editor (
for questionset editor)Sunbird-inQuiry/quml-player (
for quml player)Sunbird-inQuiry/devops (
devops script for provisining, deployements)Sunbird-inQuiry/devops-cred or devops-private (private repo)
Sunbird-inQuiry/quml-portal (
dummy portal for experiencing inQuiry feature)
quml-service:
configuration change:
Cassandra's keyspace configuration will be updated as per the new keyspace format.
New Keyspace Format: [env_name]_[bb_name]_[keyspace_name] . e.g: dev_inquiry_hierarchy_store
Kafka topic configuration will be updated as per the new format. New format: [env_name]_[bb_name]_[topic_name]
if any other dependent services (e.g: taxonomy-service) will be deployed to inQuiry infra, that will also have similar inQuiry specific configuration changes.
schema/script change:
Cassandra database script will be modified to have a building block name in keyspaces.
Kafka topic creation script will be modified to have a building block name.
code changes:
all common modules listed in the
Current Code Structuresection will be pushed to either a common repo or different repo for each module - Need to decide by considering maven central movement for each component?For resolving common module dependency, we have the below options:
option 1:
Check out all dependent repos along with the main repo and do a sequential build (dependent module first. e.g: platform-core then graph-engine) and then build the main repo.
A custom Jenkins build script with the parameterised branch will be written for the same. so that inQuiry can point to a specific branch/tag for the dependent component.
We need to have a similar build script for local checkout and development.
option 2:
Add all dependent modules as submodules.
Modify the Jenkins build script to checkout the main repo along with all submodules.
For all submodules, a specific branch (e.g: 4.9.0) will be configured.
For service level code movement, we have the below option:
option 1:
only assessment-service related code will be committed manually in the proposed structure because we need to take the code out from
knowledge-platform/assessment-apifolder.
option 2:
the entire knowledge-platform repo can be moved and then we can delete the unused code and modify the current service folder structure.
the graph-engine module needs to be enhanced to either use a DB connection or API call for fetching category definition. the default option would be DB call unless it is configured for API call. - for now, we can deploy taxonomy-service under inQuiry BB and use the DB call. this enhancement can be taken up in the next release.
Note:
1. inQuiry will build & deploy specific release versions of other services (e.g: taxonomy-service, search-service, etc) and upgrade these services as and when required.
quml-jobs:
Two Jobs (questionset-publish & auto-creator-v2) will be moved to the inQuiry Github repo.
For the above two jobs, two dependent components
jobs-coreandpublish-coreare required.These two common components should be present in the separate repo and should be pulled during the build of inQuiry specific jobs.
Three other common jobs (search-indexer, audit-event-generator, audit-history-indexer) can be build and deployed under inQuiry infra from Knowlg BB.
the inQuiry will maintain only the build & deployment script and BB Specific configuration for other jobs. the codebase will remain under Knowlg BB.
quml-editor:
Option 1:
inQuiry will copy the collection code and make changes on top of it.
quml-editor code will not have any collection & content related changes.
Option 2:
A base editor needs to be designed & implement
Both quml-editor & collection editor will use this base editor.
quml-player:
quml-player code is already independent. So just we need to move from project-sunbird to inQuiry Github repo.
Databases/Streams/Tools:
Neo4j:
inQuiry will have its own neo4j graph database.
No data migration will be done.
A Post-installation script will be created to create necessary objects in graph DB.
Cassandra:
inQuiry will share the Cassandra DB with Knowlg BB but will have its own keyspaces having building block names.
No data from SunbirdEd will be migrated.
Redis:
inQuiry will share the Redis DB with Knowlg BB but will have its own DB number specific to the building block.
No data from SunbirdEd will be migrated.
Elasticsearch:
inQuiry BB will share the ES DB with Knowlg BB but will have its own DB index specific to the building block.
search indexer and other jobs will be configured to write data to inQuiry specific ES index.
search-service will be configured to inQuiry specific ES index to consume the data.
No data from SunbirdEd will be migrated.
Kafka:
inQuiry will share the Kafka server with Knowlg BB but will have its own topic name specific to building block.
inQuiry will not install the data product which performs Kafka topic backup.
Jenkins:
inQuiry will have its own Jenkins Server.
K8s Cluster:
inQuiry will use a shared K8s Cluster but will have a custom namespace.
0 Comments