Recommendations when planning and implementing ELM DOORS Next 7.0.x upgrade project

A major change in DOORS Next between V6.0.x and V7.0.x is the way data is stored and queried. In our labs we have seen significant improvement in scale and reliability on V7. We have also observed that there are database queries that need further optimization (see support flash 6391660).

These queries are sensitive to customers' data shape and volume which can impact DOORS Next upgrade time as well as runtime performance. They may occur when using Oracle, IBM Db2 or Microsoft SQL Server.

Recommended Path

Therefore we recommend the following:

• Contact IBM Support by opening a case and asking for the latest information on fixes and open defects, if any, related to DOORS Next upgrades to V7.
• Apply the latest iFixes for your ELM applications.
• With production-like data, ideally a recent production clone, test "day in the life", business critical scenarios that are part of a typical project lifecycle for your users. Be sure to include the scenarios outlined below. This should be done as part of a production readiness assessment before proceeding with a production upgrade.
  1. Perform single-user testing of these scenarios.
  2. Perform multi-user testing, for example, with five users exercising the system at the same time.
  3. If any functional issues or performance regressions occur while running the tests, we recommend gathering the data described in Troubleshooting DOORS Next 6.x to 7.x Upgrades (in particular the database metrics and ISADC output) and submit a case to IBM Support along with the gathered data.

Scenarios to include in day in the life testing

The following requirements scenarios should be included in your 'day in the life' testing. We have tested these scenarios ourselves with the data shapes and volumes in our repositories. However, since the scenarios have proven to be sensitive to organizations' particular data shape and scale, it is important that you validate your usage patterns.

1. Open your largest modules (under 10000 per guidelines) with your typical views - from a stream, baseline, or changeset context. As applicable, include views with filters, columns with custom attributes (especially enumerations), traceability links. Repeat similarly from the ‘All Artifacts’ tab, if used.
2. Export these views to your common output formats 
3. Open the history of the largest modules 
4. Create and deliver a changeset to a stream 
5. Deliver the changeset from stream to stream 
6. Create your typical links to requirements artifacts (e.g. requirement to requirement, test artifact to requirement, work item to requirement) using your typical mechanisms (e.g. link by attribute, links panel, links explorer, import, etc.).
7. View links to requirements (e.g. requirement to requirement, link by attribute, test artifact to requirement, work item to requirement) as you typically would view them, e.g. hover over links, use links explorer, etc. 
8. Generate a traceability report document from the large module view (both within DOORS Next and from PUB, as applicable). This scenario can be inclusive of your typical customer deliverables that are created using RM publishing service (as opposed to data in LQE).

Related topics: Interactive Upgrade Guide, Understanding DOORS Next Sizings in 6X, Troubleshooting DOORS Next 6.x to 7.x