Forum Discussion

darias's avatar
darias
Occasional Contributor
2 months ago

Execution steps sporadically loading very slowly

Hello, dear SmartBear Community!

We have a sporadic issue with Zephyr Squad for Jira Cloud.
When it happens, we see a spinner instead of the test steps in the Test Execution (see attachment).
The test steps in the corresponding Jira ticket are opened quickly at the same time, so the issue only affects Test Executions.

The browser DevTools shows a specific slow endpoint - the one that loads execution details: https://prod-play.zephyr4jiracloud.com/connect/public/rest/api/1.0/execution/8cfb1aa5-...

This does not happen consistently, but when it does, it usually lasts for a long time and affects all our QA team members.
I.e. today none of the executions could be loaded for 3 hours (!).
This makes the work of the QA engineers so much slower than it could be as most of the time during test execution is spent on waiting for Zephyr Squad.

We filed a support ticket at the beginning of January. The response from the support team that we received 6 days ago (mid-February) was that a development ticket had been raised. There was no further update yet.
We still continue to see the issue quite often. I would say that it has become worse since our support ticket in January as now we have days like today - when even page refreshes do not help.

Does anyone else experience this issue? 
If you experienced and solved it, we would be grateful if you could share your solution.

P.S. Dear SmartBear team, could you please help? We have been your loyal customers for 4 years now, and we used to recommend your products to our colleagues.

5 Replies

  • We routinely have the 'spinner of death' when trying to open or during execution.  We sent a .har file and were told to do regular re-indexing and clearing cache (Setting --> Apps.  Within the left pane under Zephyr Squad click 'General Configuration'.  Within that page scroll down to re-indexing:  Project MetaData (all project or specific project), then Project Execution (all projects or specific project.  We do the specific one that is reporting the spinning/slowness.  Then scroll down to clear cache and permission cache).  This only resolves for a short time.  Have put in multiple tickets, but same suggestion to re-index, which cures for a 1/2 day, day, or a week or two, random.  No root cause at this point, but does get the Testers moving again.  Word of warning: Have all testers log out of Zephyr before starting, the longest it's taken is 26 - 27 min, shortest 9 min for the re-index of project executions.  A couple times I think someone got back in and had missing test cases or folders, had to re-index again, which resolved the issue. 

    • darias's avatar
      darias
      Occasional Contributor

      Thank you WendellG13!

      We have tried re-indexing once after a recommendation from the SmartBear support team, didn't make much of a difference.
      We will experiment with this one more time now.

  • Very slow and cache clearing/re-indexing is not a long-term solution, like you said. Today I can't even get to the General Config page to run these two jobs. Testers are waiting minutes for every test to load. So frustrating and not helping us meet our deadlines. Two questions for other users 1. How long ago did this start for you? and 2. Has anyone found a better solution to this issue? 

    • darias's avatar
      darias
      Occasional Contributor

      Hello AmyEB!

      For our team it started around the Dec 2024 - Jan 2025.

      Before that, there was also sporadic slowness in Zephyr, but usually it affected all parts of it - Test Details in the Jira Test tickets, and Executions, and also the problems were never consistent.

      Now, this specific issue with Executions is pretty consistent and isolated, as the Test Details in the actual Jira ticket seem to be loading much better.

    • WendellG13's avatar
      WendellG13
      New Contributor

      We first reported months ago, same as darias post, Dec 2024 it started in earnest.  Before I was able to stop testers, re-index anywhere from 10-20 min, now I've switched to doing each morning prior to test days.  Yesterday was 60 min, today 71 min.   Latest response was cloud version is being distributed to new users and could request to migrate, but looks like working on new clients first, then will look at migrating legacy customers.