How to do critical path analysis in JIRA ?

[Manoj Dadhich]

I would like to identify the task falling in critical path based on the time lines of the stories.

Please help with the step.

[Kristján Geir Mathiesen · Accepted solution]

Hi @Manoj Dadhich and welcome to the Community!

One way to do this, using just Jira out-of-the-box, is to flag those issues that are on the critical path. Then you can either or both:

a) find them with a JQL search like:  Flagged = Impediment 

b) on a Kanban or Scrum board, create a quick filter (Flagged = Impediment) and thereby see them all by using the quick filter.

HTH,
KGM 

[Sidan Gao]

Hi Kristján, I have same request coming up and your answer is very helpful. Just want to check by "is to flag those issues that are on the critical path", are you saying we need to manually flag the issues on the critical path? Or is there a way to setup in Jira to flag them automatically?

Thanks!!

[dstevens]

If I could find the Critical Path tasks... I wouldn't need to flag them manually.  How to find the critical path tasks?

[Alex Radnaev]

Hi all, we had the same problem, so we built the Jira app "Critical Path" to find the critical path tasks automatically and label them. Check it out! https://marketplace.atlassian.com/apps/1238303/critical-path?hosting=cloud&tab=overview

[Alex David]

Hey, on finding the critical path tasks rather than flagging them by hand, the short answer is that Jira won't work it out for you outside of Plans, and it's worth knowing why.

A longest path needs a dependency graph plus a weight on each item. Jira already has the graph, it's sitting in your issue links. What a plain instance doesn't have is anything that derives dates from those links. Start and due dates are values somebody typed in, and nothing recalculates them when a blocker slips.

Kristjan is right that manual flagging is where most plain instances land. Automation gets you a little further, a rule can walk one hop of links and tag everything that blocks something else, and you can chain rules for a couple of hops. What you can't get natively is a recursive, duration-weighted longest path.

Where that leaves you:

- Plans (formerly Advanced Roadmaps) is the closest native thing. It's on Jira Cloud Premium and Enterprise, and it ships with Jira Software Data Center. Its auto-scheduler does derive dates from estimates, capacity, rank and the dependency chain, and it writes them back to the date fields you map under Plan settings, Scheduling. There's no field or JQL clause called critical path, but the Dependencies report is the native thing that surfaces the driving chains.

- Build it against the API. Pull links in bulk rather than one call per issue, so a JQL search with fields=issuelinks (the endpoint is /rest/api/3/search/jql now, the old one is retired) or POST /rest/api/3/issue/bulkfetch at 100 keys a call. Three things will bite you. Every link comes back on both endpoints, so dedupe and normalise direction or your edge count doubles. Jira lets you create circular chains, so an unguarded walk never terminates and you need cycle detection. And link type names are admin configurable, so filter on the names actually in your instance rather than assuming Blocks. issuelinks gives you keys only, so estimates are a second pass. All doable if you're comfortable with the API, just budget more than an afternoon.

One thing that isn't a route: a JQL filter. linkedIssues() is native, so issue in linkedIssues("ABC-123", "is blocked by") does work, but it takes a hard-coded key and it only goes one hop. There's no walking a chain or counting links across a whole filter, so chain depth and "what blocks the most downstream work" stay out of reach in JQL.

Full disclosure, I work on Dependency Graph for Jira at ChefStackz. It picks out the longest blocking chain automatically from your Blocks links, and it's a Forge app, so Jira Cloud only: https://marketplace.atlassian.com/apps/2907488740/dependency-graph-for-jira

Last thing, whichever route you take. A longest blocking chain is a hop count, not a schedule. Story points aren't durations without a velocity assumption, and Original Estimate isn't a calendar span without capacity, so use the chain to decide where to look and do the date maths separately.

Alex

[Mariano Tallarico _Appricity_]

Hello everyone. I don't know if this is related to BigPicture, or if it is something native to Jira. In my case, we use BigPicture, and we have the same problem. We have the need to work, dynamically, with those tickets that (depending on the context of the project) belong to the Critical path.

I was recently in contact with Big Picture Support, asking them about it, and they told me that at the moment, in BigPicture, "the critical path is just a visual highlight of the tasks in Gantt and there is no way to identify and filter the tasks that are on a critical path according to Jira fields/tags."

That said, they have a pending ticket to cover this functionality: https://appfire.atlassian.net/browse/BP-6068 (use incognito mode to access it).

I hope I have been able to contribute to the cause, if this is still pending. In my case, I will continue looking for alternatives until they make the improvement.

Greetings community.