Q: Where do I find documentation that helps me understand the big picture?
A: The following resources will help to understand the big picture:
- README - Introduction
- More information describing the big picture including the two main flows IRS Iterative and IRS Recursive and involved components and services.
Further information can be found in the IRS Architecture Documentation, for example:
- Architecture Documentation - Business Context
- Architecture Documentation - Structure
- Architecture Documentation - Building block view
- Architecture Documentation - Runtime view
Q: Where do I find the main documentation?
A: See Item Relationship Service Documentation.
Q: Is there a glossary?
A: Yes, see Item Relationship Service Documentation.
Q: Where do I get help?
A: Please read the documentation and tests first. If that does not answer your questions, ask the developers (see CONTACT).
Q: Are there any known issues and limitation?
A: See Known Issues and Limitations.
Q: Where can I report a vulnerability?
A: See SECURITY.
Q: Where can I report a bug?
A: Normal bugs may be reported as public github issues. For reporting vulnerabilities see SECURITY please.
Q: Can I autoscale irs across multiple pods? What do I have to do?
A: at the current state it is technically possible to use Kubernetes autoscaling by setting "autoscaling.enabled=true"
in values.yaml.
This will spin up additional IRS Pods once the target load is hit.
However, due to IRS being a stateful application, it is not very useful and would require additional implementation
efforts to balance IRS workload across multiple instances.
Q: Where do I find information concerning compatibility of dependencies?
A: See COMPATIBILITY-MATRIX.
Q: How can I contribute?
A: See CONTRIBUTING.
Q: How do I create a release?
A: See CONTRIBUTING.
Q: How do I format the code?
A: See CONTRIBUTING.
Q: Where can I find more information concerning the type of tests and used technologies and libraries?
A: See README - Tests.
Q: Where do I find test data?
A: Testdata is located under local/testing/testdata. For more information see the test data README.
Q: Where do I find information concerning monitoring?
See README - Monitoring.
Q: How to reset / purge the minio database?
A: To reset the MinIO database, simply uninstall the item-relationship-service helm chart and install it again by running
helm uninstall your-irs-app-name
helm install -f your-values.yaml your-irs-app-name irs/item-relationship-service
Q: How can a job that has hung up (no reaction) be canceled?
A: To cancel a running job, call the Job cancellation endpoint PUT /irs/jobs/<job-id>. This will stop all further
processes from executing and sets the Job state to "CANCELED".
Q: Which resources are recommended for which use case? (cores/RAM)
A: The item-relationship-service helm chart has default resources set to 1.5Gi memory with a minimum of 250m
and a limit of 750m CPU.
Depending on your use-case, this might not be sufficient when requesting a lot of jobs in parallel or jobs which go very
deep in the chain.
Based on our testing, a CPU limit of 1.5 is capable of handling >40 Jobs with a depth of ~3 simultaneously without any
issues.
In case you run into OutOfMemory Exceptions, increase the memory request and limit.