Editing SLA configuration on the SLA settings page can cause the configured SLAs to disappear temporarily.

XMLWordPrintable

    • Type: Bug
    • Resolution: Unresolved
    • Priority: Low
    • Component/s: SLA
    • None
    • 2
    • Minor

      Issue Summary

      The SLA settings page in Jira Service Management Cloud can temporarily disappear or show an empty-state view after SLA-related admin actions because the server renders the page with empty timeMetrics while other config data such as calendarRefs is still present.
      This appears to be the same recurring root-cause family already investigated in support, where a transient backend/ERS failure in the jsd-rolex time metric fetch path is converted into an empty result and then treated as valid configuration state.
      Steps to Reproduce

      Step 1
      Open a Jira Service Management project’s SLA settings page:

      •  /jira/servicedesk/projects/<PROJECT_KEY>/settings/sla
      Step 2
      Perform SLA-related admin actions such as:

      •  editing an existing SLA,
      •  adding multiple goals to an SLA,
      •  saving SLA configuration changes,
      •  or triggering SLA recalculation around the same time.
      Step 3
      Refresh the SLA settings page after the save/action completes or fails.
      Step 4
      Observe that the SLA settings page may temporarily:

      •  show no SLA metrics,
      •  appear as if SLA configuration has disappeared,
      •  or render an empty/add-SLA style state for roughly 15 minutes to 1 hour.

      Expected Results

      The SLA settings page should continue to render the project’s configured SLA metrics correctly after SLA administration actions.
      If a transient backend dependency failure occurs while fetching time metrics, the page should surface a recoverable error state rather than rendering as if no SLA metrics exist.

      Actual Results

      The SLA settings page can be server-rendered with:

      •  timeMetrics: []
      •  while calendarRefs are still present
      This causes the page to behave as if no SLA metrics are configured, even though the customer’s SLA configuration still exists.

      Workaround

      Use the workaround identified in the previous investigation and support guidance:

      •  wait for the cached bad state to expire and retry later,
      •  avoid repeated rapid SLA admin changes while the page is in the bad state,
      •  if possible, retry once the temporary disappearance resolves,
      •  and reduce/optimize heavy SLA configurations where applicable.
      For the broader configuration pattern, support has also recommended simplifying expensive SLA JQL/goal setups where possible, although this does not address the underlying product defect.

       

              Assignee:
              Unassigned
              Reporter:
              Xinhe Yang
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

                Created:
                Updated: