Sample from bugzilla here

This is put on top as it influences the usefulness of all the other featueres. Pretty much every graph described below needs to allow for filtering by project, where project is something that is loosely defined on a case-by-case basis, using these fields:

• we currently have keywords, product, component, status_whiteboard
• needs flags, needs target_milestone, needs teams

Sample from bugzilla here

This ties in closely with the feature above. In principle, the same set of field is used to constrain data for a release (especially target_milestone and flags).

Sample metrics types here

This will need flags to allow for better selection of releases. Also, DWH data is incomplete (only bugs #560K to about #585K are included).

• opened/found per week (last 7 weeks): [1]

• closed/fixed over time (last 7 weeks): [2]

Samples from bugzilla here.

• Current open bugs by prio for products Firefox and Core: [3]

• Currently open bugs by severity, for products Firefox and Core: [4]

Sample from bugzilla here

• Current open bugs by component, for products Firefox and Core: [5]

• Same, but only the 10 components with the most bugs: [6]

• Ideally we want to support faceting by combination of component x status so we can have nice bar charts with multiple colors per column (one column per component, one segment/color per status). This is not currently in SOLR, but there is a patch ([7]) that we could use. Can also be done currently with one SOLR query per status.

Sample metrics types here

• For the new DWH, this needs team-data to be useful, especially for per-team-dashboards.
• Get open bugs by assignee (other than nobody), just filtering by product, components etc.: [8]

Sample metrics types here

needs flags

Sample metric types here

This can all be filtered by product, component, keyword etc.

• Average age of the currently open bugs, by component: [9]

• Average sum of OPEN time (including in re-opened state) per bug [10]

• Average time a bug is OPEN (or re-opened) before being closed: [11]

Sample metric types here

This can all be filtered by product, component, keyword etc.

• Currently open bugs, how many are still open for the first time vs. re-opened at least once: [12]

Samples from bugzilla here

This can all be filtered by product, component, keyword etc.

• Of the given bugs, normal bugs vs. regressions [13]

Sample metrics types here

This can all be filtered by product, component, keyword etc.

• How long did bugs stay UNCONFIRMED (on average) before being changed, not counting bugs that still are unconfirmed [14]

• How old are UNCONFIRMED bugs (avg) that still are UNCONFIRMED (hard to do in a single query with the above, but it really is something different). Interestingly, this number is much higher. So if your bug has not been confirmed after 5 days, there is an increasingly high chance that it never will be (this is also due to the inclomplete data set that I have here!). [15]

The metrics dashboard should automatically deliver this with all dashboard items.