Custom JQL Pre-computation API - Need a Delete API/Disable Pre-computation

[David S]

Hi,

I am trying to write a custom jql. But the precomputation functionality is bit complex to implement for my usecase. I can't store all precomputation ids and the parameters and check every one of them every time when there is a change in data to update with new computation. For example, some of my custom JQLs can be impacted by actions like update in custom fields or linking issues. For that I may have to update hundreds of precomputations. Also it is hard for tracking all the parameters for the precomputations. 

As a solution, I expect a way to delete precomputation via REST API or to disable pre-computation via atlassian-connect.json config. It seems, with precomputation, I have to make hundreds of requests to JIRA every minute.

[Mercy]

Jira Cloud’s custom JQL functions use precomputation to cache results for performance reasons, and that behavior is controlled entirely by Jira’s internal query engine. There isn’t a public REST endpoint or `atlassian-connect.json` flag to delete or disable precomputed results. The platform automatically invalidates cached results when issue data changes, but if your function depends on complex conditions such as issue links or custom field updates, you’ll need to manage the cache lifecycle yourself within the JQL function module by implementing the `invalidatePrecomputation` logic in your app’s backend.

[Sunny Ape]

Hello @Mercy 

Can you please provide a more detailed description of how to "manage the cache lifecycle within the JQL function module"?

1. What and where is that 'JQL function module'?
2. How exactly is "implementing the "invalidatePrecomputation` logic" done within that 'function module'

Can you also provide links to your sources of information on this topic.

Thanks in advance

[Sunny Ape]

@Mercy 

I spend hours trawling through all of Atlassian's development frameworks, libraries and APIs for Jira Cloud trying to find what you described... but found nothing. I found no mechanism, anywhere, that disables or prevents a JQL function from performing what the OP describes as 'pre-computation', or to "manage the [JQL] cache lifecycle" as you stated, whether or not that JQL function was a custom one or not.

Where are you getting your information from?

[David S]

I am not sure how reiterating whatever I have stated as problem as solution helps. I don't want JIRA to cache my custom JQL response and the answer from Atlassian is, "We will cache it, deal with it!". I can implement this "handling" on my side. But it will make me call precomputation and other JIRA REST APIs thousands of times every second. That is going to make things bad for everyone. I am not sure who or what (I mean some LLM) designed this!

[David S]

invalidatePrecomputation

Please give sample implementation or point me to a documentation for this @Mercy if you are not a Hallucinating LLM.