How to loop through the lookup results and link each issue to some other, per criteria

[Inna S]

Hi, 

we are doing a migration of the large data set and need to link some issues. 

Unfortunately the trigger 'issue created' used to link every newly created issue to an existing one matching a criteria, hit the throttling limit and can't be used.

I was thinking of making a different automation, that would

- look up all the issues matching a criteria (e.g. no link of specific type is present or a field is empty)

- for every issue in the lookup result, create a link between the 'current' issue in the loop and the one with the one field value matching field value in the 'current' issue.

I'm having trouble iterating over the lookup results: added Log step to print the issue key but this doe not work as expected.

I then can't seem to find the way to link two issues when one of them is 

- not the trigger issue

- not a recently created issue

Is there a way to link just two arbitrary issues or full the system into treating the lookup result as a triggerIssue of a kind?

Thank you!

[Trudy Claspill · Accepted solution]

Hello @Inna S 

You can use a smart value as the issue to which to link, per the help text in that action

How are you going to trigger the rule?

How many issues do you expect to be retrieved in your when you look up issues?

There are a few ways you can get a list of issues to operate on.

1. You could use a Schedule trigger and put a JQL into the trigger. The actions in the rest of the rule will be performed against each issue returned by the trigger. I'm not sure what the limit is on the number of issues a Schedule trigger JQL can return. And you don't actually have to make use of the Schedule. As soon as you create the rule you can manually trigger it with the Run Now button rather than having it run on an actual schedule.

2. The Lookup Issue Action. Note that action is limited to retrieving 100 issues. You can iterate through the results set using an Advanced Branch option:

3. For Branch > Related Issues > JQL is another option for retrieving a set of issues. Within the branch you would then be operating against each issue retrieved by the JQL.

As to getting the issue that you want to link to, I would need more details about how you would retrieve that issue by JQL relative to the current issue against which you are operating.

[Inna S]

Hi @Trudy Claspill ,

I can run it scheduled or I can trigger it manually.

In one case, the throttling happened after processing 800 issues, leaving me with under 200 to take care of. 

So I expect to have some hundreds to process every time. Running several rounds of 100 to cover them all is ok with me.

I'll try your suggestions and report back.

Thanks a lot!

[Inna S]

Hi @Trudy Claspill , I've managed to craft the rule following your guidance.

Works great!

Thank you very much, these capabilities are not obvious at all.

Inna

[Trudy Claspill]

I'm glad I could help!

If your problem has been solved, please consider marking the Answer as Accepted to help other users find posts with validated solutions.

[Tim Martin]

Hi Inna,

Sometimes, if it's a one off exercise (with 100s-1000s of issues to update), I'll just export the data, use excel/powerbi/some other data manipulation tool, and then upload a CSV file back into Jira.

Otherwise, for automation, I would recommend using something like:

- Scheduled Trigger (run JQL, untick the "Only include issues..."): find all issues with missing link or whatever your scope is

- Branch - related issues - JQL: find other issue with same customfield value

- Link or Edit action - link the branched issue with the trigger issue

You may still run into service limits (Automation service limits | Cloud automation Cloud | Atlassian Support) but you can simply rerun as needed until you've linked everything.

[Inna S]

Hi @Tim Martin , 

our parallel team is working with Excel manipulations and report having hard times with it so I would not want to go there.

I'll try to follow your suggestion.

[Inna S]

Hi @Tim Martin ,

I've managed to create the rule that works. Starting with the scheduled trigger and going on.

Thank you.

[Mia Paulin]

@Inna S  I would create a Confluence Space, link the Jira Projects(scrum, Kanban, otherwise, etc) as pages or a blog with links to these different Jira Project or just the epics with the needed issues(and yes, you can use the whole epic key- I say Epic because you have a greater picture of what's needed), add a bitbucket integration link to the Space so you can have a DevOps functionality where the developer working on the project can send thing to an from the bitbucket repository and made changes when needed, either through the Atlassian portal or on their at home computers.  Also, one of the things I would do is get  database engine with Big Data functionality and the necessary software(Apache Spark(but something for SQL or JQL, not necessarily python)or something like that), that can query large quantities of Data in a small amount of time.  After that, it's just a matter of data wrangling- normalizing, cleansing of data, feature engineering and simple queries when you have that data in a format that the query can do.

[Inna S]

Hi @Mia Paulin , either I was unclear with my task at hand or you were anwering completely different question.

[Mia Paulin]

@Inna S  I misunderstood the question.  I saw large dataset and loop and my mind went in a totally different direction.  If you know the name of the issue, you can use the search.  Will be back to answer your question in a bit.  Have some stuff to work on.  I apologize.