How to manage issues that affect several projects

[Patrick Dubois]

Hello,

We need advice to setup an efficient way to manage the relationships between our libraries and projects. Let me give you an example. Say we have two libraries, Lib A and Lib B, each with its own JIRA project. Each lib has several Versions, each with different enhancement and bugs associated with it. Let's called these different versions A1 and A2 for Lib A and B1 and B2 for lib B.

Now we have project Gizmo which will use these two libraries. Gizmo has its own roadmap, with Version G1 and Version G2 planned. Version 1 will use A1 and B1 while version 2 will use the latest-and-greatest A2 and B2.

We'd like to be able to do a few things:

• Explicitly configure JIRA so that Version G1 of Gizmo is linked to Versions A1 and B1 of the libraries. In particular, that would mean that any library bug associated with A1 or B1 would be visible when looking at the Version G1 of the Gizmo roadmap. Any subsequent bug associated with A1 should also automatically show up in Gizmo G1 roadmap. Same thing with the Change Log Report.
• As a bonus feature: it would be great if we could selectively ignore some bugs from a library in the Project Gizmo context. For example, there's a bug in module foo of lib A1 but we don't care about that module, so let's ignore that bug somehow in Project Gizmo.
• When a new bug in a library is found, we need to proactively evaluate which projects are potentially affected and how. We need a way to log the results of that investigation for each project and track thre resolution progress.

Now from my limited experience with Jira, I see a few ways to do this:

• Use the Structure plug-in. Not sure to what extent it would work with our use cases. I'd like comments.
• Use the Multi Project Picker. Doesn't look like it supports the granularity (associate with versions) that we want. Also doesn't seem popular so extended support is questionable.
• Cloning issues. For each library issue that is created, we could also create a clone in each project that uses that library and create a link between them. That could work but seems quite labor intensive. Also the clones don't get automatically updated when the parent issue changes. This issue is raised in JRA-7948 (from 2005!?!).

Any suggestions to handle our use cases?

Thanks.

[FernandoAdorno]

A long time has passed since 2015. Does Jira now have something to sort this problem?

I have projects on Jira to handle Products evolution, but also others to handle Deliveries agreed with customers, and I simply would like to have a field on a user story which would allow be to "blood-link" them, instead of using the weak n-n link, which are not really visual and easy to handle.

Do we have a solution for this at the moment?
Tks

[Robert Cox]

Would like to see a solution for this too, if anyone has been able to find one.

[Patrick Dubois]

Hi John,

Thanks for your answer, we are actually using Agile so your assumption is correct. I see how your solution could work, but it's based on selective filters. I'd prefer a "deeper" relationship within JIRA itself if at all possible. Otherwise the relationship between project Gizmo and Lib A and B is not immediately obvious to an outsider. The relationship is also not visible outside that particular Agile board.

So far I'm leaning towards creating clone of issues between projects, perhaps using a plugin to facilitate this. The drawback is that creating the proper clones would be manual and someone creating a bug in Lib A for example could forget to clone the bug to project Gizmo.

I might have simplified our use case too much. In fact, our end product is a measurement instrument which includes software and hardware. The SW is composed of instrument-specific SW as well as libraries, which each have their own JIRA project. Same for the HW, it's partly composed on a common HW platform which has its own JIRA project with associated issues and instrument-specific HW (actually there's only one project for instrument specific stuff, mixing both HW and SW together). I'm trying to see how to reproduce in JIRA the links that exist in the real world between all these components.

I'm basically trying to break the silos that tend to exist between projects. John might work on instrument A while bob works on instrument B, but both instrument share half the SW and half the HW. They therefore probably also share half the bugs. John needs to know when Bob finds a bug in a common component that might affect his instrument.

I'm actually quite puzzled that JIRA is not better at handling this kind of cross-project links. Maybe I'm trying to use JIRA the wrong way?

[TJ Baker]

Hi Patrick, Have you found a solution to this yet? I found your inquiry looking for a solution to a very similar problem. In our case we have 'products' that are used for many different 'client projects'. We develop the products for distribution as well, and so have need to track product specific issues in the Product Project directly, but each different Client Project will also introduce issues that we will develop solutions for and add to the product. "Linked Issues" are okay for occasional usage, however, sometimes we may have 100+ issues we have to create for a Client Project, and each Client Project could actually involve several Products, so creating a clone and linking the issues in the Product Project seems overkill. Not to mention our developers would probably want to kill us if they had to manually create hundreds of duplicate issues and link them to the product's issues over and over again. Looking for some solution to this to no avail so far.

[TJ Baker]

I posted a followup question as while we have a very similar case I don't think it's exactly the same: https://answers.atlassian.com/questions/12167794/how-to-manage-cross-project-issue-dependencies

[John Paz]

What a fun problem to work on!

My solution assumes that you are using JIRA Agile (Greenhopper), and instructs on how to create an Agile board to manage all of this.

1. Create a new Agile board.
2. Configure the board so the JQL query looks something like this:
   
   project in (Gizmo, Lib A, Lib B) AND version in (A1,B1,G1,A2,B2,G2) ORDER BY Rank ASC

   This will give you all issue types, which may not be what you're looking for. But it gives you three projects and two versions to view from one Agile board. 
    
3. Change the card colors to queries, create three card colors and make each query point to a project:
   
   project = Gizmo / project = Lib A / project = Lib B

4. For selectively ignoring bugs, create Quick Filters so you can toggle their view on and off. The query would look something like this:
   
   project = Gizmo AND issuetype = Bug

   Name it "View/Hide Bugs" so others will know what the switch does.  

Let me know how it works out for you! Good luck!