Stability tests are relatively limited for the moment

• perform resources snapshot
• looping one use case (no // processing)
• re-perform resources snapshot
• compare

stability test is usually executed after RC1

but it does not see possible race condition due to // testing detected even with low traffic taht could create objects that will never be cleaned and then lead to memory issues.

it would be nice to start working on a realistic traffic model based on the feedback of the Service provider operating ONAP and try to apply it to CI chains.

It shall be possible to run weekly CI chains to study the stability of intermediate version and report possible issues as early as possible

Metrics shall be define to qualify ONAP stability

It also means that logs must be properly managed (dependency to another Requirement)

A lab would be needed to work on this task.