Issue: We have PR analysis set up with Azure DevOps. Since our upgrade to v2025.4.4 we are routinely having issues displayed under the Overview tab of the PR branch with “No Issues. Hooray!” displayed under the Issues tab. We’ve repeatedly tried deleting the Elasticsearch data and have ensured that a CI analysis has been performed on the master branch prior to attempting a PR analysis. Deleting es8 data sometimes works, but only briefly and not always. For example, I had just deleted es8 for one PR analysis which did fix it to have the very next analysis for a different project show 109 issues under Overview and nothing under Issues. Deleting es8 a second time didn’t fix it.
Is that different than deleting the es8 folder and restarting SQ? I’ve done that a few times as called out in the original post. I just did this process again and I still have a project showing 109 issues on Overview and no issues on Issues.
Sorry I wasn’t more focused on my first pass on this. This still sounds like a bad ES index. The question is why your indices are repeatedly being corrupted. This is not normal.
I think you need to look operationally at the index files and what outside forces might be impacting them.
The issues data wasn’t there after ElasticSearch reindex but it did show up after a forced update via issues/reindex API. I’ll keep an eye on it and will report back if that resolves future missing issue problems. It wouldn’t be ideal, but at least it will unblock our flow.
I don’t know what happened in our upgrade to v2025.4.4 but our issue details are just not reporting any more. I’ve tried clearing ES index and restarting. I’ve made sure that each project has a baseline run. I’ve tried forcing and issues reindex for specific pojects via the API. The summary will report issues and those issues are reported out to our Azure DevOps pull requests, but they are not displayed in the issue details. This means we cannot dismiss them. I’ve had to resort to force-completing many PRs which defeats the entire purpose of SQ PR decoration.
I welcome more ideas on how I might resurrect our instance and get it back to being a functional and vital part of our lifecycle.
This is not normal behavior. I still think you need to look at what external processes are running on your host that might be interfering with your ES indices. For instance, do you have something “helpful” like Windows Defender running on the box?