We are looking at the possibility of moving archived cases between 2 systems that don’t have the same applications. We found articles on moving it between systems with the same application and tested it as working by copying the ZIP files and running the pzPerformReIndexing activity against the classes of the archived cases. But when the target system is not running the same application it does not work.
The question is what class related rules need to be present in the target system, and in the target application ruleset stack, in order for the indexing/search of those copied archived cases to work?
I am not 100% sure, but I think at a minimum for archived cases to be searchable in a different application, the target system must still have the rule/data model needed to materialize index data.
That means the target needs:
1) The case class hierarchy (concrete + work pool classes)
The original Work-<Type> class (or mapped equivalent)
Its inheritance chain (Work-, Assign-, etc.)
If missing → indexing fails silently or produces incomplete results
Because indexing uses class metadata + properties to populate search indexes. If the class doesn’t exist, the engine can’t resolve fields.
2) The properties referenced by indexing
Standard properties (pyID, pyStatusWork, pxCreateDateTime, etc.)
Any custom properties used in:
Declare Index rules
Search (Elastic) mappings
Report definitions used by search
If properties don’t exist:
They won’t be indexed
Search returns incomplete or empty results
3) Declare Index rules (critical)
These are often the missing piece in cross-app scenarios.
Declare Index rules define:
What gets indexed
Into which index class (e.g., Index-Work*)
If they are not present → pzPerformReIndexing has nothing to execute
This is usually why it works when it’s the same app but fails when it is a different app.
4) Index classes
Typically:
Index-Work*
Index-* for custom indexes
These must exist with the same schema (properties)
5) Search model / Elasticsearch mappings
If using modern Pega search:
The search index definitions (DSS / mappings) must support the class
Reindexing will attempt to push documents—but without mappings, results may not appear correctly
6) Supporting rules the engine implicitly relies on
These are often overlooked:
Data transforms / activities used during archival restore
Class group / work pool configuration
History / attachment class references (if included in indexing)
@sloan Thank you for your reply is there any way to get the exact list of things that would be required? For example I am not sure which declare index would be required, not which search index definitions. Is there any documentation available?