Linking issue to JIRA server failing

[dachakra]

Environment

SaaS (https://sentry.io/)

Steps to Reproduce

This only happens with certain projects. The Jira integration is definitely working since many work fine and can create issues.
I noticed that some of the Jira projects in the configuration section were blank. This led me to believe that some of the projects may have been removed or Sentry no longer had access to them, so I ensured that the required projects had access by re-requesting access for the service account used to access those projects, and I deleted the blank entries. However this did not fix the problem.
I then tried linking to an existing Jira ticket and this worked fine. This leads me to believe that the cached last used project might be the problem. The Sentry Issue pulls the last project used and fails because it no longer exists

Expected Result

Linking issue to Jira ticket.

Actual Result

This bug was reported a while ago in this issue https://sentry.sentry.io/issues/4130565830/?project=1 and a more recent event can be seen in this issue: https://sentry.sentry.io/issues/4267026714/events/e6c5ccc91a694b45824d96d56ee260c0/?project=1

Product Area

Issues

Link

No response

DSN

No response

Version

No response

[getsantry]

Assigning to @getsentry/support for routing ⏲️

[getsantry]

Routing to @getsentry/product-owners-settings-integrations for triage ⏲️

[armenzg]

@dachakra FYI If you need to share any private data, there's a Private Data field for our board.

This is the code I believe is related:

sentry/src/sentry/integrations/jira_server/integration.py

Lines 705 to 712 in 805821d

	project_id = params.get("project", defaults.get("project"))
	jira_projects = self.get_projects()
	if not project_id:
	project_id = jira_projects[0]["id"]
	client = self.get_client()
	issue_type_choices = client.get_issue_types(project_id)

[armenzg]

@dachakra could you please paste a link to a customer's issue where this happens?
(Under "Private Data" and let us know when you do)

The event you point me to shows a 400 error, could you please confirm that the jira_project_id is correct?
https://<jira_deployment>/rest/api/2/issue/createmeta/<jira_project_id>/issuetypes

If you can't get me a URL to an issue where this happens, could you please get me the API payload and response? (You can use private data if PII is contained)
In my case this is what I hit the Sentry API with:

{components: ""
description: "Sentry Issue: [SENTRY-10QN|https://sentry.sentry REDACTED"
issuetype: "10004"
priority: "4"
project: "10374"
reporter: "REDACTED"
title: "Test ticket feel free to delete"

Can you please confirm that they're using this prompt to create the Jira links?

I do not know how the UI would not send the project ID if they selected it.
The code would fall back to the first project if no value is passed:

        jira_projects = self.get_projects()

        if not project_id:
            project_id = jira_projects[0]["id"]

[dachakra]

I created a shadow JIRA ticket here that contains the details: https://getsentry.atlassian.net/browse/ISSUE-1673

[riot-sschott]

Hey, I was just wondering if there was any update on this issue?

[armenzg]

Hi @riot-sschott , a fix was deployed last night. Is this now working for you?

[riot-sschott]

Hmm unfortunately no. [image: image.png] Request URL: Request Method: GET Status Code: 500 Remote Address: Referrer Policy: strict-origin-when-cross-origin {"detail":"Internal Error","errorId":"8b0fafc688d14cdca69c1b86cf8ce974"} It's not part of a blue/green deployment or anything like that?

…

On Thu, Sep 7, 2023 at 9:32 PM Armen Zambrano G. ***@***.***> wrote: Hi @riot-sschott <https://github.com/riot-sschott> , a fix was deployed last night. Is this now working for you? — Reply to this email directly, view it on GitHub <#53013 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A37GCH4KI4XLBENU4LVEWPDXZGWDXANCNFSM6AAAAAA2NPNZAM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[ceorourke]

@lobsterkatie I have not, I have been trying when it pops up when I'm on triage rotation, and Steve tried a couple times too with no luck.

[dachakra]

Hey @riot-sschott, is your team integrating Sentry to Atlassian Data Center via the Jira Server integration? Or are you using Jira Server and experiencing this issue? CC @brianthi

[riot-sschott]

We are using Jira Server.

…

On Sat, Dec 9, 2023 at 3:07 AM dachakra ***@***.***> wrote: Hey @riot-sschott <https://github.com/riot-sschott>, is your team integrating Sentry to Atlassian Data Center via the Jira Server integration? Or are you using Jira Server and experiencing this issue? CC @brianthi <https://github.com/brianthi> — Reply to this email directly, view it on GitHub <#53013 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A37GCH3GS5OVCKXVSL5NFULYIM3LVAVCNFSM6AAAAAA2NPNZAOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNBXGQ2TIMJTHA> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[rachrwang]

@riot-sschott - do you plan on making any changes to your integration based on Atlassian stopping support for Server products? https://www.atlassian.com/migration/assess/journey-to-cloud

[riot-sschott]

We are in the process of migrating to cloud. I can wait till this has happened and test again if you like.

…

On Tue, Dec 12, 2023 at 7:51 AM Rachel Wang ***@***.***> wrote: @riot-sschott <https://github.com/riot-sschott> - do you plan on making any changes to your integration based on Atlassian stopping support for Server products? https://www.atlassian.com/migration/assess/journey-to-cloud — Reply to this email directly, view it on GitHub <#53013 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A37GCH5G6DL644QZDSWYMT3YI5W6TAVCNFSM6AAAAAA2NPNZAOVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQNJQHA3DQNZWHE> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[rachrwang]

@riot-sschott - that would be great, since we've had a difficult time reproducing this problem on our side with multiple devs looking into it, and we haven't seen any of the problems reported for our cloud integration. Thank you for the quick follow-up!

[riot-sschott]

@rachrwang We have completed the migration and are no longer experiencing the issue. Thanks for your assistance.