Context:
Currently, the zeebe records dont contain any information in regards to the tenants. This information will be added to the exported records and we have to adjust the zeebe record import logic so that these changes are imported in Optimize as well.

When importing, we should make sure the tenant id has been updated for:

• Process Definitions
• Process Instances
• Incidents inside Process Instances
• FlowNode instances inside process instances

Additionally, importers need to be able to handle the scenario of "old" exported zeebe data in the zeebe record indices that do not yet contain tenant information. In this case, the tenantId should default to <default>

ATs:

• tenantId from zeebe data is imported to Optimize, incl:
  • process definitions
  • process instances
  • incidents
  • flownodes
• importer defaults missing tenant data to the <default> tenant

Testing Notes:

Two scenarios need tested in CCSM

1. "Normal" use case:
   1. deploy data in zeebe on different tenants and confirm tenant info is imported correctly to Optimize. This can be confirmed by inspecting the tenantId field in ES for the process definition and process instance documents. For process instances, also confirm the correct tenantId has been applied to incidents and flownodes within instances
2. Presence of "old" exported zeebe records;
   1. Stop Optimize
   2. Run Zeebe version that does not yet have multi tenancy (pre 8.3)
   3. Deploy data without tenant info in zeebe
   4. Stop Zeebe
   5. Run Zeebe version that does support multi tenancy (post 8.3)
   6. Deploy data with tenant info in zeebe
   7. Start Optimize
   8. Let Optimize import records and confirm that records imported from step 3 all have "<default>" tenant and records from step 6 all have the correct tenantIds