How to manage parallel streams/releases for a single customer

[Scott Silverthorne]

Hey all,

I'm fairly new to JIRA and am trying to set up an agile project but have hit a stumbling block.

I have a single client with multiple, parallel streams of development. What I mean by this is, for said client I have a maintenance stream that runs and releases every months and also alongside this I have a major development stream that delivers major upgrades every quarter. These are 2 separate streams and remain separate because we need to ability to be able to release quick fixes etc in the maintenance stream.
I run these with 2 separate scrum teams, however all defects or changes are essential from the same backlog. We groom the backlog to say what can go jnto maint and what should go into the major stream.

So my question, what would be the best way to set this up? A project for each? But then i would have 2 backlogs or can I have a single project with parallel versions/sprint running. But I need to be able to report on these as separate streams.

An added complication. Anything fixed in the maint stream needs to be cloned/updated into the major release stream. So I need to still be able to do this.

Hope all this makes sense.

Thanks
Scott

[Troy Spetz]

We use JIRA 7.10 server version.

If all the work is for a single product, I would use a single project.

We use JIRA native FixVersion field and work with 2 releases in parallel (the 'trunk' which is the next release to be announced and the 'branch' which is the subsequent release).

All changes to the trunk release are automatically pushed to the branch release. Or, using your terminology, everything in the current maintenance release would be automatically pushed to the next major release.

This is done by our software tools -- not as as result of cloning JIRA issues. Cloning a JIRA issue implies the work has to be done twice.

[Scott Silverthorne]

Hi Troy,

Thanks for getting back to me. 

What you describe is the exact same this as my situation and in fact, yes we use the terminology 'trunk' and 'branch' too.

So a few questions if i may....

When you are managing the work for the trunk and the branch in parallel, how do you set up your sprints? do you have parallel sprints enabled or do you have both int he same sprint but logged to a different version?

When you are talking about your software tools, are you referring to something in JIRA or do you mean that cloning is all done with you source control tools etc? if this is the case are you not using the cloning feature in JIRA at all.

Thanks
Scott

[Troy Spetz]

Hello Scott,

To answer your questions:

1. In the same sprint, we can have work on trunk (i.e. the release soon to be completed) and branch (i.e. the release which is just starting up).

We keep the work separated, in the same sprint, by using the FixVersion field (e.g. trunk work is FixVersion = 5.0.0 whereas branch work is FixVersion = 5.1.0).

Yes, doing this means that sometimes (rarely) our Developers commit their changes to the wrong FixVersion ... but this is quickly spotted (and wrists slapped ;-)

We try to minimise the amount of time we have a branch release open. As soon as the release on trunk is done, it is announced to our customers (for server upgrades) and the branch is merged back to trunk.

2. I only log the work in JIRA once (i.e. no cloning of issues to get a code change on both trunk and branch releases). The actual code merge (from trunk to branch) is done by our code integration tools.

Hope that clarifies better how we work with JIRA ...

Regards,

Troy.