What are useful integrations between TCM, Defect Tracker, Source Control and Project Management software?

Question

My company is looking for a new defect tracking tool. One of the issues I have with the old tool and with the current direction things are going in the search for a new tool is that we have standalone tools that do not play nicely together for our TCM, Defect Tracker, Source Control and Project Management. I know that it is nice to have integration between these systems and those integrations can save time and increase productivity as well as provide useful insight into trends, etc., however I was having trouble articulating what exactly those useful integrations are or could be. The tool we are currently looking at for Defect Tracking is JIRA, which I have read good things about, but it doesn't have any built in integration with TFS, TeamCity or AgileZen which are the other tools we are using. My concern is that if we have a bunch of tools that are not integrated or require additional work to integrate then we are creating more work for ourselves down the road. Is there need for concern? Should I just be happy with standalone tools with no integration, or should I commit time and energy to finding tools that work well together? Any links to data or documents on the subject would be useful to me.

A few good reasons for having integration between these systems that I came up with are:

1. Associate a test case failure with a bug, and vice-verse. This allows you to easily see what tests/scenarios are blocked by a particular bug and what bugs are currently blocking a specific scenario.
2. Associate a check-in and subsequently a build number with the resolution of 1 or more bugs. Currently, there is a manual process to a) make a check-in, b) resolve one or more bugs, c) include the change-list and build number in each resolved bug.
3. Get detailed commit graphs and statistics associated to bug resolutions.
4. Report on defects fixed within a build, or between a range of builds.

Any other ideas or additional integrations that are or could be useful?

Answer 1

In my company there were around 63 integrations methods I needed wanted - I'll write down some important ones here:

1. Support Request to Feature
2. Email to Bug/Feature/Enhancement
3. Commit(s) to Code Review
4. Feature Request to Requirement
5. Forum Thread to Feature Request

There are many of them - but you gotta begin with whats most important for your organization. We've arranged all our prospected integrations in order of priority for the same purpose.

As to defect manager suggestion: Switch to JIRA without any hesitation - its affordable, universally adapted and everything you can think of, or need is likely to be available as a JIRA plugin (well not if you have 63 integration links like us - but still it covers a lot).

We're using Redmine right now and very soon we'll be shifting to JIRA. We ourselves have gone through the process of this debate and found out JIRA was the ultimate tool we needed. And if you have a small team under 10, don't even think about it - switch to it right away for $10 only.

One serious suggestion though: Before switching, carefully establish costs of JIRA plugins - some of these can get very expensive.

Answer 2

I would focus on the specific integration features that you need, and then come up with specific reasons as to why this info is useful and how will it increase productivity or reduce costs.

Personally I find that many integration features to be bells and whistles. I know this doesn't exactly answer your question but I felt it important highlight. You could come up with many great integration ideas but most likely it will be very expensive. Integration generally requires an expensive tool that does everything or a team in your company to put it together, then manage it.

Answer 3

Well, on the one hand the benefits of a wholly integrated SDLC tooling solution are many. It's a beautiful world when all your tools work together.

However, to achieve this you MUST have a full time employee whose primary responsibility is to engineer that integration. That means you're in for at least $75k per year to enjoy the pleasures of tool integration. I don't know what your budget looks like, so you'll have to decide if it's worth it.

You can, optionally, hire consultancies to put together the integration. It'll cost more in the short run, but IF you think that your SDLC won't change much in the future then it might be worth it. In that case, you'll still need an engineer dedicated to maintaining the integration, but maybe only as a 50% time allocation.

Alternately, you can switch to an all-in-one tooling solution (I believe ClearCase and TFS both qualify). This costs a bunch (on a per-seat-per-year basis), but they come fully integrated out-of-the-box. It can save you a lot of headaches.

Lastly, you can choose to live with the current lack of integration. It may be painful, but the other options all cost real money.