Jira thoughts
Ticket types
The idea here is to reduce the number of ticket templates we have to maintain. The plan is something like
Epics - Project level tickets
Bugs - Something went wrong
Feature - Feature level tickets
Task - A catch-all for anything not covered in other tickets
Subtasks - for breaking down features, bugs and tasks should we need to
Support - support tickets
Current types | Keep or Remove | Replacement Type |
|---|---|---|
Epic | Keep |
|
Bug | Keep |
|
Change | Remove | New feature |
Data Correction | Keep |
|
Discussion | Remove | None |
Emailed request | ?? | ?? |
Improvement | Remove | Feature |
Incident | Remove | Bug? |
New Feature | Keep |
|
Pending Discussion | Remove | Task |
Problem | Remove | Bug |
Service Request | Remove | Task |
Story | Remove | New Feature |
Support | Keep |
|
[System] Change | Remove | Support |
[System] Incident | Remove | Support |
[System] Post-incident review | Remove | Support |
[System] Problem | Remove | Bug |
[System] Service request | Remove | Support |
[System] Service request with approvals | Remove | Support |
Task | Keep |
|
Sub-task | Keep |
|
Project vs product vs operational vs support backlogs
Projects are defined as short-lived feature driven backlogs. You would expect the features to be large enough to warrant their own project. They should comprise one or more epics with ten or more tasks that may be broken down into many subtasks. Once the feature described by the project is complete, these should be exported to the systems shares and deleted.
Product backlogs are defined as long-running maintenance and improvement backlogs. Small features, bugs and general maintenance tickets would live here.
Operational backlogs are defined as more general purpose backlogs for activities that are not specific to a single product, but might have cross over with one or many products.
Support backlogs are attached to helpdesks and are holding places for contact from users. Tickets here may be dealt with in place or funnelled off to other backlogs
Jira plan
It has been suggested that it might be best to start all new Jira projects, migrating new tickets over and removing the old. This means we could clearly identify tickets that have been reviewed, and we can remove all the old tickets in case they contain PID. Below is what I am thinking
Jira Name | Type | Old Jira projects | Work streams |
|---|---|---|---|
UKKA Helpdesk | Support | NA | UKRDC public help |
UKKA IT Help Centre | Support | NA | UKKA IT help |
Data Corrections | Operational | UKRDC data corrections, Workitems | All systems |
Operational Infrastructure | Operational | Sysadmin, Infrastructure (old), NBT migration | All systems |
Software Infrastructure | Operational | Development tasks | All systems |
Renal Registry | Product | Renal Registry (old), AKI Registry | RR database, validation, steve, editor, AKI registry, consented cohort, NHSBT |
UKRDC | Product | UKRDC-TNG, UKRDC Apps, UKRDC (old), PKB | UKRDC database, API and UI, patient checker, PKB, MRC |
Radar | Product | Radar (old) | Radar database, API and UI, International Radar |
CUPID | Project |
| CUPID |
Real time stats | Project |
| Real time stats |
Dev Servers | Project |
| Dev App, Mirth, DB |
Radar 3 | Project |
| Radar 3 |