🤝 This analysis is the result of a partnership between Bitergia and the FreeBSD Community during October 2024, as an effort to make a characterization of the backlog from FreeBSD’s Bugzilla instance.
⚙️ We produced a series of Dashboards to help with this characterization. Learn more details in the Characterization Dashboards section.
📈 Here you will find the slide deck summarizing the results from the characterization. Although the whole history has been analyzed, we agreed to focus on the open bugs from the last two years and the last year, starting with the bugs from the Base System product. This analysis can be extended to other periods and products by selecting the appropriate options.
📁 Index
- Slides: Characterization of the Bugzilla Backlog.
- Characterization Dashboards.
- How to import the dashboards from OpenSearch Dashboards.
- Bugzilla Dataset.
- Useful Links.
Click here to access the PDF version of the slide deck.
All these dashboards are available under the dashboards folder from this repository.
The instructions to import them into OpenSearch Dashboards are available in this section.
The dashboards contain instructions on how to use them and interpret the different visualizations. In most of the widgets, you will see a little "information" button next to the title. If you place your mouse over this button, you will see a helping text.
This dashboard offers an overview of the Bugs submitted for all the Products in the selected period, detailing the amount of activity coming from each Product.
- Overall numbers (Bugs, Submitters, Assignees, Products, Components).
- Number of Bugs, Submitters and Assignees per Product.
- Top Bug Submitters.
- Author name, Author organization, number of Bugs, number of Products and number of Collaborators.
- Evolution of Bugs Submitted by their Current Status.
- Distribution of Bugs Submitted by Product and Severity (%).
This dashboard focuses on efficiency in closing bugs in the selected time frame. The Evolution of the Backlog shows how the backlog has changed during the selected time frame, while the BMI (Backlog Management Index) measures efficiency in closing issues. The Bugzilla index does not contain the closing date of the Bugs, so the values are approximated using the date of the last comment in "Closed" bugs.
- Overall numbers of bugs by status (Submitted, Closed, and Open).
- Evolution of the Bug Backlog over time.
- BMI(Backlog Management Index) over time (%).
This dashboard provides a breakdown of the open Bugs submitted during the selected period, focusing on "Unattended" and "Abandoned" Bugs. A Bug will appear as "Unattended" when it does not receive any comment from a user different from the author who created it. A bug will appear as "Abandoned" when there were other comments, but the last comment was made more than 3 months ago.
- Overall numbers: Breakdown of the number of Bugs by their attention times.
- "Submitted", "Attended and Recently Commented", "Unattended", and "Abandoned".
- Time evolution: When were the Unattended and Abandoned Bugs created?
- Open Bugs by Time to First Attention.
- Open Bugs by Time since Last Comment.
Dashboard focused on the Open Bugs needing more attention, submitted during the selected period. A Bug will appear as "Unattended" when it does not receive any comment from a user different from the author who created it. A bug will appear as "Abandoned" when there are other comments, but the last comment was made more than 3 months ago. We recommend to exclude the bugs with severity "Affects Only Me" to get more relevant results.
- Overall numbers: Breakdown of the number of Bugs by their attention times.
- "Total", "Attended and Recently Commented", "Unattended", and "Abandoned".
- List of Unattended Bugs.
- List of Bugs by the Time since the Last Comment.
This dashboard offers an enhanced version of the bug list from Bugzilla, showing the list of Bugs submitted during the selected period. The results are limited to 2,000, in descending order by their creation date. The "Filters" widget on the left can be used to narrow down the search. Another way to reduce the number of results is to select a smaller time frame.
- Overall numbers (Bugs, Submitters, Assignees, Products, Components).
- Table of Bugs Submitted.
- ID, URL, Product, Summary, Keyword, Author, Assignee, Status, Resolution.
- Days to 1st attention, Creation date, Date of the last change, Date of the last comment.
bugzilla, because the index pattern and the dashboards point to this alias.
To import the Dashboards provided by Bitergia (including the related visualizations and
the index pattern), go to the “Dashboards Management” section, click on “Saved Objects”
and then “Import.” Then, select the .ndjson files under the dashboards directory and
import them. We recommend to choose the option "Check for existing objects" and then
select "Request action on conflict".
You can import each dashboard individually, or import all the Characterization
Dashboards at once by selecting the file all-characterization-dashboards.ndjson.
Now, the dashboards should be available under the “Dashboards” section.
In this section we describe the structure of the index containing the Bugzilla data. This index contains one document per Bugzilla ticket.
The documents in this index are about the bugs reported in a Bugzilla instance (one document per bug). BAP can extract information about the reporters, assignees, and resolvers of these issues. It can also analyze the status, the number of changes, the number of comments, and other metadata associated with each bug, such as the Product and Component on which the bug was reported to.
The information for this data set is provided through indices in OpenSearch. Those are used to calculate the metrics and charts in the dashboard. In case you need information about them, it is available below.
| Index Pattern | Content |
|---|---|
| bugzilla | Bugs of the Bugzilla instance |