Replication Guide
Replication data stores
Managing specifications using requirements engineering tools is based on the concept of identifiers for all information objects in the tool’s database. The uniqueness and permanence of these identifiers is key to the requirements engineering process.
A fundamental aspect of Windchill Requirements Connector is the ability to store information about the data store and the objects within that data store with respect to state and change information. This so-called replication data can be stored either in a local data store (local replication), or in a centralized data store (global replication). As a convenience, Windchill Requirements Connector offers complete implementations of a local replication store, and a centralized replication store.
Replication content
Requirements are normally identified by a system attribute referred to as the requirements identifier, which once assigned cannot be modified for the entire lifetime of the requirement. To ensure that requirements can be replicated to other data sources, each requirement must be assigned an identifier that is independent of the underlying tool that maintains the requirement: the so-called replica identifier. The replica identifier must be universally unique (UUID).
The replication store is a database that maintains the mappings between an object identifier originating in the requirements database, and the replica identifier that is generated the first time an object is exchanged.
Examples of mapped objects stored in the replication store are as follows:
• Requirements
• Requirements hierarchy containers
• Links between requirements
• Link containers
Figure 110. ID Mapping
 | Please note that the copying of a HOME ID for information purposes has to be done as a separate attribute (which means the original Object Identifier, e.g. using DOORS, shall not be used. The Object Identifier has to be exported in a separate attribute which could be called e.g Object ID by using an alias name. |
Choosing replication stores
The choice of replication store should be carefully considered before a productive deployment of the Windchill Requirements Connector within an organization.
Local replication store
The local replication store is stored in a file that is local to the exchange project. When opening the first roundtrip configuration within a project, the local replication store is locked. Therefore only one user can work with that project at a time. Each project will use a different replication store. The consequence is that if a requirement is exchanged by different projects, it will be assigned different replication identifiers.
Local replication stores should be used by organizations that do not distribute the requirement specifications further to third parties, for example, end-tier supplier.
Distributed (global) replication store
Using a distributed replication store ensures that a requirement is assigned a single, unique replication identifier, irrespective of where the requirement appears in exchange configurations. The same replication identifier will then appear at all locations (worldwide) with the same identifier. There is no limitation on concurrent usage of the replication store. The replication data for all projects is available in all projects, so there is no limitation on linking between documents residing in different projects. In contrast to the local replication store, Windchill Requirements Connector needs a global replication configuration to access the global replication store. That configuration is created with the PTC Requirements Connector Administration Tool once to configure host, port, database name, user and password for the database. All users of Windchill Requirements Connector will create entries in the global replication database with the configured user.
 | Distributed replication stores should be used by organizations that need requirements to be tracked across their entire lifecycle irrespective of exchange configurations, for example, OEM or tier 1 supplier. |
Questions to help you with the decision
• Do you need a central backup of the replication data?
• Do you need to export or import the same set of documents in parallel?
• Do you need to use a document created within one project in other projects?
• Do you need to export the same document in different projects with the same identity?
When you answer one of those questions with Yes you will have to setup a global replication.
Do you need a central backup of the replication data?
When working with local replication every user must make sure that the projects he creates are part of a backup strategy (e.g. are saved on a central network drive). The global replication guarantees that all replication data is stored in a single instance of a central database that can be part of a regular backup strategy.
Do you need to export or import the same set of documents parallel?
The local replication store can only be opened by one Windchill Requirements Connector instance at a time. This means that two persons cannot work on a single Connector project in parallel. If you need multiple persons to work concurrently on one Connector project, you have to use global replication.
For a big exchange process multiple roundtrip configurations of the same Connector project have to be used for export. To save time, you split up the roundtrip configurations with a colleague and you both export some of the data.
Do you need to use a document created in one project in other projects?
When working with local replication the Connector projects cannot access the replication data of documents within the tool that were created in other Connector projects. This means in particular that during an import PTC Requirements Connector will not be able to create links to documents that were not created with the current Connector project. If you need to use central documents that can be the target of links in multiple Connector projects, you have to use global replication.
You receive a document that is the target of links from multiple other Connector projects (for example a central company wide glossary).
Do you need to export the same document in different projects with the same identity?
When working with local replication, a document gets a new identity in each Connector roject. Each exported identity will be created as a new document in the tool (requirements engineering tool) on the receiver's end. For example each imported identity corresponds to a new document Windchill RV&S database. If you need to use one identity across multiple Connector projects, you have to use global replication.
You manage your company wide glossary in a central document that is the link target of other documents that are used in different Connector projects. You want your exchange partner to be able to import this glossary only once (see question above).
Parent topic