1. Docs
  2. Administration
  3. Integrations
  4. How the numbers work

How the integration numbers work

Each number below shows its formula and a worked example, so you can check what you see on screen or in a message.

7 min read Updated CustomersAdmins

In short

On this page you can

  • Understand why a card says Ready while its page says Disabled.
  • Read the KPI counters of Jira, Azure DevOps, Xray, Zephyr, Confluence and Slack.
  • Recompute the pass rate in pipeline messages and webhook payloads.
  • Check what the daily summary counts and for which day.
  • Predict the CI exit code and the APM time window of a run.

How to use it

  1. Find the number you see in the headings below.
  2. Read its formula and where the inputs come from.
  3. Compare with the worked example.
  4. If your figure still differs, follow the linked troubleshooting entry.

Status chips#

Catalogue chip: Active or ReadyIntegrations catalogue, each card
Active = the integration is saved for this project AND switched on
         Email:   at least one mailbox with Active on
         Webhook: at least one webhook with Active on
Ready  = everything else (not set up, or set up but switched off)
Always Ready: BrowserStack, Active Directory (LDAP), Datadog, Dynatrace,
              GitHub Actions, GitLab CI, Jenkins

Example: Slack is saved but you turned Integration Active off. The Slack card shows Ready and the Slack page shows Disabled. Turn it on and both show Active. The catalogue cannot tell “switched off” from “never set up”; the page can.

The Active filter’s number counts the cards that show Active. Datadog, Dynatrace and the CI cards are never counted, even with a connection, because they have no catalogue status.

Page chip: Not configured, Disabled, ActiveTop of each integration page
Not configured = no settings saved for this project
Disabled       = settings saved AND the integration switch is off
Active         = settings saved AND the switch is on

Example: you open Zephyr Scale in a project where nobody set it up: Not configured and the Setup button. After saving with Integration active off: Disabled.

Last testLine under the page title
Testing…   = a test is running now
OK         = the last test in this browser session succeeded
Failed     = it failed; the message appears in the connection panel
Not tested = no test since the page was opened
Auto-test on open (integration on): Jira, Xray, Zephyr Scale, Azure DevOps
Email: read from the last saved test (OK when IMAP worked)

Example: you open the Slack page: Not tested, because a test posts a real message. Click Send Test Message: Testing…, then OK · 08/11/2026 10:32. Reload the page and it is Not tested again.

“n results”Catalogue header
n = cards whose name or description contains the search text (ignoring case)
    AND that are in the chosen category (All = every card; Active = cards showing Active)

Example: category Notification holds Slack, Microsoft Teams, Email (IMAP/SMTP) and WhatsApp Business: “4 results”. Type whats: “1 results”.

Jira and Azure DevOps counters#

Jira: Linked Tasks, Successful Syncs, Last SyncJira page, KPI row
Linked Tasks     = links between a task and a Jira issue in this project
Successful Syncs = links whose last sync succeeded
Last Sync        = time of the last full sync: Send to Jira, Get from Jira or a timed sync
                   (single tasks sent on save do not change it)

Example: 40 tasks were sent; 2 failed because their Jira project lost a required field. A link is created only when the issue is created, so Linked Tasks shows 38. If reading one of them back from Jira later fails, for example because the issue was deleted, that link is marked as an error and Successful Syncs shows 37.

Azure DevOps: Linked tasks, Sync problems, Runs in Test PlansAzure DevOps page, KPI row
Linked tasks       = links between a task and a work item
Sync problems      = links whose last sync ended in error + links whose work item was not found
Runs in Test Plans = pipeline runs written to Test Plans (all time)

Example: 25 linked tasks; someone deleted 2 work items in Azure DevOps and 1 update was rejected. Sync problems = 2 + 1 = 3.

Xray and Zephyr counters#

Xray: Mapped Tests, Successful Syncs, FailedXray page, KPI row
Mapped Tests     = links between a test case and an Xray Test
Successful Syncs = links whose status is SUCCESS
Failed           = links whose status is FAILED
Last Sync        = time and status (SUCCESS, FAILED or PENDING) of the last sync

Example: 60 mapped tests, 58 SUCCESS, 2 FAILED. Successful Syncs shows 58 with “Failed: 2” below it.

Zephyr Scale countersZephyr page, KPI row
Test cases from Zephyr     = scenarios imported from Zephyr
Test cases sent to Zephyr  = scenarios SmartestQA created in Zephyr
Requirements from Jira     = main requirements + sub-requirements created from linked issues
Runs written to Zephyr     = runs with status PUBLISHED among the last 20 runs (so at most 20)

Example: 120 imported cases link to 4 epics and 30 issues: Requirements from Jira = 4 + 30 = 34. After 25 nightly runs, Runs written to Zephyr shows 20, because the table keeps the last 20.

Confluence counters#

Total Mappings, Test Cases, Test Reports, RequirementsConfluence page, KPI row
Total Mappings = distinct Confluence pages linked to this project
Test Cases     = distinct documentation pages
Test Reports   = distinct report pages
Requirements   = distinct requirement pages

Example: one report page covers a run of 12 test cases. SmartestQA stores 12 links but the page counts once: Test Reports = 1.

Slack, Teams and WhatsApp counters#

Total Notifications and Notifications a / 7Slack, Teams and WhatsApp pages
Total Notifications = messages this channel delivered (all events + daily summaries; test messages not counted)
Last notification   = time of the last delivered message
Notifications a / 7 = switched-on events among the six event switches + 1 if the daily summary is on

Example: Pipeline Completed, Pipeline Failed, Critical Task and the daily summary are on: 4 / 7.

Rate and Duration in “Pipeline Completed”Slack and Teams message
Rate     = Passed ÷ Total × 100, one decimal (0 when Total = 0)
Duration = run end − run start, written as 1h 2m 3s, 4m 5s or 6s

Example: Total 48, Passed 48, Failed 0, started 07:21:15, ended 07:42:09: “Rate: 100.0% | Duration: 20m 54s”. A run with any failed test sends Pipeline Failed instead, which has no rate.

Daily summary#

Daily Test SummarySlack, Teams and WhatsApp, once a day
Day          = yesterday, 00:00:00 to 23:59:59 by the server's clock
Tests run    = test case executions that started that day (one per test case per run)
Passed       = executions with status Completed
Failed       = executions with status Failed
Rate         = Passed ÷ Tests run × 100, one decimal
Open Tasks   = today's tasks grouped by status, Closed left out
Critical     = tasks with priority Critical (any status)
Sent when    = the hour of Summary Time = the server's hour AND nothing was sent to this channel yet today

Example: yesterday 3 pipelines ran 16 test cases each: Tests run = 48. 45 Completed and 3 Failed: Rate = 45 ÷ 48 × 100 = 93.8%. Executions still running, or stopped, count in Tests run but in neither Passed nor Failed, so Passed + Failed can be less than Tests run.

Critical counts every task with priority Critical, closed ones included, so it can be higher than the critical tasks in the Open Tasks list.

Webhook passRate#

data.passRatePIPELINE_COMPLETED webhook payload
passRate = round(passedTests × 1000 ÷ totalTests) ÷ 10     (0.0 when totalTests = 0)

Example: 47 of 48 passed would give 979.17 → 979 → 97.9. In practice a completed event has every test passed, so it is 100.0.

Retry timingWebhook deliveries
attempt 1 at t, attempt 2 at t + 1 s + answer time, attempt 3 + 5 s, attempt 4 + 30 s
retry only on: network error, HTTP 5xx, 408, 429

Example: your endpoint answers 503 three times, then 200. The log shows one delivery with 4 attempts and a duration of a little over 36 seconds.

CI state and exit code#

Exit code of the CI stepCI job result
0 = state PASSED (every test case passed)
1 = run finished in any other state: FAILED, DISCONNECTED
2 = not finished within SMARTEST_TIMEOUT_MIN, or trigger / status errors
DISCONNECTED = a test case RUNNING or NEW for more than 2 hours

Example: timeout 60 minutes, the run needs 75. At minute 60 the step prints “timed out after 60 min; the run continues in SmartestQA.” and exits with 2. The run still finishes and appears in Report.

APM time window#

Window read for a load test reportServer metrics (APM) panel
from = run start − 60 s
to   = max(run end, run start + 60 s) + 60 s
at most 24 hours (longer windows keep the last 24 hours)
saved with the report when "to" is more than 5 minutes in the past

Example: a load test ran 14:00:00–14:20:00. SmartestQA reads 13:59:00–14:21:00. At 14:26 the series are saved and the panel says “saved with the report”.

Didn’t find your answer?Write to support with your project name and what you see on screen.

info@smartestqa.ai
© 2026 Selenity Technology · SmartestQAsmartestqa.aiThis page as MarkdownChecked against smarttest-fe predev 2a8c79cc (IntegrationsPage, IntegrationPageShell, utils/statusTone), smarttest-be 633c2a03 (IntegrationStatusController, PipelineNotificationService, DailySummaryScheduler, WebhookEventAspect, CiRunRules, ApmService)