Skip to content

Synthetic journeys

A synthetic journey is a real user journey that Identiguard replays in a real browser, on a schedule, from fixed Identiguard addresses. It tells a client whether their customers can actually sign in, without waiting for a customer to complain.

Logs show how many sign-ins succeeded. They cannot show the sign-ins that never started because the page did not load. A journey tests the whole path, from the sign-in page to the application.

Action Admin Member Viewer
Open Journeys, read status, run history, screenshots and journey incidents Yes Yes Yes
Set up monitoring for a tenant Yes No No
Record a journey, Edit steps Yes No No
Run now, change the schedule, Pause, Resume, Delete Yes No No
Reissue probe credentials, Reset MFA, Remove journey monitoring Yes No No

A member or viewer sees the schedule as text, for example “Every 5 minutes · Running”. Under Not set up yet they see the badge An admin can set this up. The record and edit pages open for admins only.

If the feature is not switched on for the workspace, Journeys is not in the sidebar and the page address shows “Not Found”.

Type What it does Shortest schedule Password and address
Login Signs in as the probe user and checks the application loads Every 5 minutes Types the probe password, which never changes. Uses the same probe address every run.
Password reset Requests a reset, opens the emailed link and sets a new password Every 5 minutes Sets a new password each run, and Identiguard keeps it for the next run. The reset email goes to the probe mailbox.
Sign-up Registers a new account and follows the verification email Hourly Registers a fresh address each run, because an address can sign up only once. Every run leaves an account behind in Auth0.
Custom Any other journey you can walk through in a browser Every 5 minutes Uses the probe password wherever you typed it. Uses the same probe address every run.

The record wizard calls the third type Sign-up. The badge on the journey row calls it Registration. They are the same type.

  1. An admin sets up monitoring for the tenant and creates a dedicated probe user in Auth0.
  2. The admin records the journey once in a remote browser, reviews the steps and saves it.
  3. The admin chooses Run a test. Nothing runs automatically.
  4. When the test passes, the admin chooses Start running. Until then the journey stays under Awaiting first pass.
  5. Identiguard replays the journey on the schedule, from the EU, the US or both.

Journeys can run every 5, 10, 15, 20 or 30 minutes, hourly, daily or weekly. Sign-up journeys can run hourly, daily or weekly only.

See Record a journey for the steps, and Manage journeys for the day-to-day tasks.

Open Journeys in the sidebar. The page title is Journeys, with the line “Real sign-in journeys we replay on a schedule, from fixed Identiguard addresses.”

From top to bottom the page shows the Live journey map, a health banner, a totals strip, one card for each monitored tenant, and the list Not set up yet.

What you see Meaning
“No journeys configured yet” with “Add an Auth0 tenant to start monitoring journeys.” The workspace has no Auth0 tenant. Add an Auth0 tenant opens the Tenants page.
Not set up yet, “These Auth0 tenants are not monitored yet.” Auth0 tenants without journey monitoring. Each has a Set up monitoring button (admins).
“No journeys recorded yet.” The tenant is set up but has no journeys.

A Set up monitoring button also appears in the page header when exactly one tenant needs setup.

The map draws one lane for each running journey. Paused journeys are not drawn.

  • Summary sentence. For example “5 journeys running across 2 tenants from EU and US: 3 passing, 1 failing, 1 couldn’t finish. 1 in flight now.” With nothing running it reads “No journeys running yet.”
  • Legend. Passed, Failed, Couldn’t finish, Running.
  • Lanes. The map shows at most 4 lanes. After that it reads “+N more running journeys, listed below”.
  • Each lane shows the journey name, a region pill (EU or US), a track of step dots, the tenant pill, and figures: the last duration and the 30-day availability, for example “99.9% 30d”. Before figures exist it shows Passed, Failed, Couldn’t finish or No run yet.
  • Step dots. A lane shows up to 8 dots. A longer journey folds its middle steps into one +N dot.
  • A failed lane draws a larger dot at the step where the run stopped, dashes the rest of the track, and prints “Stopped at (step): (reason)”.
  • A run in progress. A small dot travels along the lane, and the figures start with Queued or Running.
  • Click a lane (or press Enter on it) to jump to that journey’s row.
  • Tooltips. Hover a dot to see the step name and its duration. With a lane focused, the Left, Right, Home and End keys move between steps, and Escape hides the tooltip. Before the first run the tooltip reads “Steps appear after the first run”.
  • Empty map. “Switch a journey on and it appears here.”

The banner shows the worst news first.

Banner Colour Meaning
“N journeys failing: (journey) on (tenant) and N more” Red At least one journey has a real failure. The journey name is a link to its row.
“N journeys couldn’t finish: (journey) on (tenant). Our script, not your outage.” Amber The probe could not complete the journey. This is a monitoring problem, not a customer outage.
“No journeys running” Neutral Nothing is switched on.
“N journeys running · waiting for the first check” Neutral Journeys are on but have no result yet.
“All journeys passing · last check 4 min ago” Green Every running journey passed its last run.

The totals strip shows Running, Paused, Failing and Couldn’t finish (with the hint “Our script, not your outage”). Failing and Couldn’t finish count running journeys only.

Each monitored Auth0 tenant has one card.

  • Title and status pill. The tenant name and one pill. See Statuses.
  • Sub-line. “N journeys · MFA” followed by a badge for each factor the probe holds: TOTP, Email code, Passkey or security key, Recovery code.
  • Record a journey (admins) opens the record wizard for this tenant.
  • Sections. Journeys are grouped under Running N, Paused N and Awaiting first pass N · can be scheduled after one passing run.
  • Incidents (N) at the bottom. See Journey incidents.

Admins see a menu on the tenant card, labelled “More options for (tenant)”.

Menu item What it does
Reissue probe credentials Creates a new probe password at once, with no confirmation, and shows it once in the dialog New probe credentials. The email address does not change. The old password stops being used immediately, so journeys show Probe credentials rejected until you set the new password in Auth0.
Reset MFA Makes Identiguard forget every MFA factor the probe enrolled. The password is not changed. You must also reset MFA for the probe user in the Auth0 dashboard. Identiguard cannot do that for you.
Remove journey monitoring Removes monitoring from the tenant and deletes every recorded journey for it. The Auth0 accounts are not deleted. A message lists the addresses to delete in Auth0.

The steps for each are in Manage journeys.

Each journey is one row.

Element What it shows
Status icon and text The outcome of the newest run. See Statuses.
Relative time When the newest run happened, for example “4 min ago”.
Type badge Login, Password reset, Registration or Custom.
Down since (date and time) An outage is open for this journey.
Slow A slowdown is open and the journey is not down. The tooltip reads, for example, “Taking 9.1 s, usually 3.0 s, since (time)”.
History strip The last 30 runs. See History strip.
Availability The 30-day availability as a percentage, or “n/a” when there is no verdict yet.
p95 The 95th percentile duration, for example “p95 1.2 s”.
Region EU or US: the region of the newest run.
Schedule select (admins) How often the journey runs. A change saves at once with the message “Schedule updated”.
Run now (admins) Queues one run, with the message “Run queued”. It is disabled while a run is queued or running.
“More actions for (journey)” menu (admins) Edit steps, Pause or Resume, and Delete. Pause and Resume appear only after the journey has passed once.
Chevron, or a click on the row Shows or hides the recent runs.

You cannot rename a journey, change its type or change its start URL. To change those, delete the journey and record it again.

Three different places show a status. They use different words.

Journey row: the outcome of the newest run

Status Meaning
Passing The newest run completed every step.
Failing, or the specific failure label such as Timed out A real failure. A customer would probably see the same problem.
Couldn’t finish: (reason) The probe could not complete the journey. The tooltip reads “Our script, not your outage”.
Queued A run is waiting for a probe.
Running now A probe is replaying the journey.
Skipped The run was not attempted.
Not run yet The journey has no result.

Section: whether the journey is scheduled

Section Meaning
Running The journey is on and runs on its schedule.
Paused The journey has passed at least once and is switched off.
Awaiting first pass The journey is saved but has never passed. It cannot be scheduled until one run passes.

Tenant pill: the tenant as a whole

Pill Meaning
Failing At least one journey has a real failure.
Needs attention At least one journey could not finish or was skipped.
Nothing running No journey is switched on.
Waiting for first check Journeys are on but have no result yet.
Healthy Every running journey passed.

The strip shows 30 bars, with the oldest run on the left.

Bar colour Meaning
Green Passed
Red Failed
Grey Did not complete, or skipped
Pale No run

Hover a bar, or move focus to it, to open a popover. It shows the outcome (Passed, Failed, Did not complete or Skipped), the time, Region, Duration, Usual (the normal duration), Much slower than usual when it applies, and the failure reason with the step. View run opens that run in the run history, when it is among the 50 most recent runs. Escape closes the popover.

There is no separate page for a run. Click a journey row to open its recent runs under the row.

  • The list holds the last 50 runs, 5 to a page.
  • Each run shows an icon, the time (for example “14:05, 2 min ago”), the duration, and the result: “Signed in”, or the failure and the step, for example “Timed out at Submit”.
  • Failed runs are open by default. They show each step with its name, its duration, an HTTP status such as “HTTP 500” when one was recorded, and Stopped here on the step where the run ended.
  • View screenshot opens “Screenshot from (journey)”: the page the probe was on when the journey failed. Not every failure has a screenshot. When there is none, the dialog reads “No screenshot was taken for this step.”
  • With no runs the list reads “No runs recorded yet.”
  • While a run is queued or running, the card refreshes every 5 seconds.

Every run that does not pass belongs to one of three groups. The group decides the colour, whether the run counts against availability, and whether anyone is alerted.

These are red. They count against availability and can open a journey incident.

Label Usual meaning
Server error The application or Auth0 returned a server error.
Timed out A page or step did not respond in time.
Unreachable The probe could not reach the site.
Sign-in rejected Auth0 refused the probe user’s sign-in on the page.
Did not reach the expected page A check failed: the expected text or page address did not appear.
Reset email never arrived The password reset email did not reach the probe mailbox.
Reset link rejected The reset link did not work.
Multi-factor code or key rejected Auth0 refused the probe’s MFA code or key.
Multi-factor email code never arrived The MFA email code did not reach the probe mailbox.

Couldn’t finish: “Our script, not your outage”

Section titled “Couldn’t finish: “Our script, not your outage””

These are grey or amber. The probe could not complete the journey, so the run says nothing about what a customer would see. These runs do not count against availability, never open or close a journey incident, and are never sent to an alert channel.

Label Usual meaning
A recorded step no longer matches the page The sign-in page changed after the journey was recorded.
Recipe could not be read The saved steps are not valid.
Probe credentials rejected The password Identiguard holds for the probe user is wrong.
Blocked by bot detection Auth0 attack protection or a firewall blocked the probe.
Rate limited Auth0 or a firewall limited the probe’s requests.
Redirected to an unexpected host The journey went to a host that is not in Allowed hosts.
Multi-factor step failed The probe could not complete the MFA step.
No stored secret for the multi-factor challenge Auth0 asked for a factor that Identiguard does not hold.
Only unsupported multi-factor methods were offered Auth0 offered only factors the probe cannot use.
Multi-factor enrolment could not be completed The probe could not enrol its own factor.
Probe mailbox unavailable The Identiguard mailbox for the probe could not be read.
Reset email failed authentication checks The reset email failed the sender checks, so the probe did not trust it.
Browser crashed The probe’s browser stopped.
Probe did not start in time The probe was too slow to start.
No probe runner available No probe was available in the region.
Internal error An Identiguard error.

The run was never attempted. The reasons are Monitoring not enabled, Billing hold, Flow disabled, Backing off and Waiting for another run to finish with the probe credential.

A journey incident records one outage or one slowdown of one journey. Journey incidents are separate from the detector incidents on the Incidents page.

Open Incidents (N) at the bottom of the tenant card. The list is collapsed by default. It covers the last 30 days, 10 to a page.

Row type What it shows
Outage Journey name, Still down or Resolved, the time it opened, the downtime (“at least N min” or “at least N.N h”), “N failed checks”, and the reason
Slowdown Journey name, Still slow or Slowdown, “slow for about N min”, and the peak against the usual time, for example “peak 9.1 s (usual 3.0 s)”

The rules for an outage:

  • Opens after two consecutive real failures from the same region. There is one open incident for each journey.
  • Resolves when every region that has reported is passing again. If one region recovers while another still fails, the incident stays open.
  • Couldn’t finish runs neither open nor resolve an incident.
  • An open incident with no results for a long time (three times the schedule interval, and at least 24 hours) is closed as “monitoring stopped”.
  • Deleting the journey or removing monitoring closes the incident without a message.

Identiguard compares each passing run with the journey’s usual duration in that region.

Rule Value
Usual duration The median of the last 30 passing runs for the journey and region. Runs already marked slow are left out.
Minimum history 10 passing runs. Before that, nothing is judged.
A run is slow It takes more than twice the usual duration, and at least 3 seconds longer
A slowdown opens After 2 consecutive slow passing runs from one region
A slowdown clears After 2 consecutive recovered runs in every region that was slow. A recovered run is under 1.5 times the usual duration, or within 1.5 seconds of it.

Failed and unfinished runs are ignored. No slowdown opens while an outage is open. A slowdown never counts as downtime.

Only confirmed outages, their recovery, and slowdowns are sent. A single failed run sends nothing. Couldn’t finish runs never send anything.

What happened Severity Slack and Teams Email channel Other
Outage opens ERROR “[DOWN] (tenant) - (journey) journey failing”, with the time it started failing, the number of consecutive failed probes and the reason The same DOWN message. The email heading is “A customer journey is failing”, with the button View journey availability. Workspace admins are emailed directly, even when no channel is configured. A ServiceNow incident is raised when the connector is enabled.
Outage resolves ERROR “[RESOLVED] (tenant) - (journey) journey recovered”, with “Down for about N min” Nothing A work note is added to the ServiceNow incident.
Slowdown opens WARN “[SLOW] (tenant) - (journey) taking Xs (usual Ys)” Nothing Nothing
Slowdown clears WARN “[RESOLVED] (tenant) - (journey) back to its usual speed” Nothing Nothing

A channel receives a message only when both of these are true:

  • The message’s severity is at or above the channel’s Minimum severity.
  • The tenant is in the channel’s Tenant forwarding scope.

With the default settings, a Slack channel (minimum WARN) receives outages and slowdowns. A Teams channel (minimum ERROR) receives outages only. To send slowdowns to Teams, lower its minimum severity to WARN. The email channel never receives recoveries or slowdowns, and it shares the limit of 50 alert emails an hour.

See Alert channels to configure the channels.

If the workspace has the ServiceNow connector connected and enabled (Super plan), a journey outage raises a ServiceNow incident automatically. Its title is “[DOWN] (tenant) - (journey) journey failing”. When the journey recovers, Identiguard adds a work note. It does not close the ticket: someone must close it in ServiceNow. See Ticketing.

Automatic stop on rejected probe credentials

Section titled “Automatic stop on rejected probe credentials”

If the probe password that Identiguard holds no longer works in Auth0, runs end with Probe credentials rejected. This is a Couldn’t finish result, not a failure of the client’s service.

After 5 consecutive results of this kind, Identiguard switches off journey monitoring for the tenant automatically, so that it does not keep sending bad sign-ins. Real journey failures never count towards this. After that, Run now returns “Journey monitoring was disabled automatically after repeated failures. Resolve that first.”

To prevent this, fix the password as soon as you see Probe credentials rejected: use Reissue probe credentials and set the new password in Auth0. See Manage journeys.

The probe can complete these MFA methods: TOTP (an authenticator app code), email code, passkey or security key, and recovery code. Push notification and SMS are not in the supported list.

  • Identiguard keeps a factor only when it was enrolled through the recorder: Fill code for TOTP, or the recorder’s built-in passkey.
  • If you enrol the probe user with your own phone or key, Identiguard holds nothing and the probe cannot answer the challenge. Reset MFA for the probe user in Auth0 so that the probe can enrol its own factor.
  • The tenant card shows a badge for each factor the probe holds.
  • Reset MFA in the tenant menu makes Identiguard forget the factors. You must also reset MFA for the same user in the Auth0 dashboard.

Before monitoring starts, an admin reads the Authorisation wording and confirms that:

  • They are authorised to permit automated sign-in testing against the tenant.
  • Identiguard signs in repeatedly, on the chosen schedule, from its own probe addresses, and these sign-ins appear in the tenant’s logs as ordinary authentication events.
  • The probe uses a dedicated test account with no more access than a test account needs, and the client removes it if monitoring is switched off.
  • Auth0 publishes no exception for synthetic monitoring in its load-testing policy, and the client is responsible for any approval its agreement with Auth0 requires.

Raise this early with the client. It needs someone with authority over the tenant.

The edit page of an Auth0 tenant has a read-only card, Synthetic journey flows, with a Manage journeys button that opens the Journeys page. It shows one of these:

  • “Journey monitoring is not set up for this tenant. Set it up on the Journeys page.”
  • “Set up, but nothing recorded yet.”
  • “N journeys, N running”, with the badge “N awaiting a passing run” when it applies.

When the feature is not available to the workspace, the card reads “Synthetic journey monitoring is available on Max and Super plans.”

Messages you see while recording, reviewing and testing are in Record a journey. Messages from day-to-day tasks are in Manage journeys.

What you see Why What to do
Journeys is missing from the sidebar, or the page shows “Not Found” The feature is not switched on for the workspace, or the plan is below Max Ask a platform operator to switch on synthetic journeys. Check the plan.
“Synthetic journey monitoring is not available on this plan.” The same cause, met when you try an action The same fix
An admin can set this up You are a member or viewer, and the tenant is not set up Ask a workspace admin
“No journeys configured yet” The workspace has no Auth0 tenant Connect an Auth0 tenant. See Connect Auth0.
“Could not load the runs” with “The request did not come back.” The run history did not load Choose Retry
“No screenshot was taken for this step.” No screenshot was captured for that failure Use the step list to see where the run stopped
Availability shows “n/a”, or the history strip is missing There is no verdict yet, or the health data could not be read Wait for more runs. This is not an outage.
The message “The change did not save” A pause, resume, schedule change, run or delete failed Try again and read the detail in the message
What you see Why What to do
Sign-in rejected Auth0 refused the probe user on the sign-in page. This is a real failure. In Auth0, check the probe user exists, is not blocked and has the right password. If you are not sure of the password, reissue it.
Probe credentials rejected The password Identiguard holds is wrong. After 5 of these in a row, monitoring stops automatically. Reissue probe credentials, then set the new password in Auth0
Did not reach the expected page A check failed. The application did not load, or its text or address changed. Check the application. If the page was redesigned, use Edit steps and correct the check.
A recorded step no longer matches the page The sign-in page was redesigned. This is a gap in monitoring, not an outage. Edit steps, or record the journey again. See Manage journeys.
Redirected to an unexpected host The journey went to a host that is not allowed Add the host under Allowed hosts in Edit steps
Blocked by bot detection or Rate limited Auth0 attack protection or a firewall blocked the probe Ask the client to allow the probe user or the Identiguard probe addresses in Auth0 or the firewall. Ask the product team for the addresses.
Multi-factor code or key rejected or Multi-factor email code never arrived A real MFA failure Check MFA and email delivery in Auth0
No stored secret for the multi-factor challenge, Only unsupported multi-factor methods were offered or Multi-factor enrolment could not be completed The probe’s MFA state does not match Auth0 Reset MFA in Auth0 and in Identiguard. Set the MFA step’s Method to a supported factor. See Manage journeys.
Reset email never arrived or Reset link rejected A real failure of the password reset flow Check the Auth0 email provider and the reset settings
Probe mailbox unavailable or Reset email failed authentication checks An Identiguard mailbox problem, or the email failed the sender checks (SPF, DKIM) Check the sending domain’s email authentication. Escalate to the product team if that is correct.
Server error, Timed out or Unreachable A real outage of the application or Auth0 Investigate the service. Check the Auth0 status page.
Browser crashed, Probe did not start in time, No probe runner available or Internal error A problem with the Identiguard probes. Not the client’s outage. Escalate to the product team
Skipped The run was not attempted: a billing hold, the journey or monitoring is off, the tenant is backing off, or another reset or sign-up run is still using the probe credential Wait for the next run. Fix billing if that is the reason.
Slow, or Much slower than usual 2 consecutive passing runs took more than twice the usual time and at least 3 seconds longer Investigate performance. The badge clears after 2 normal runs.
Down since (date and time), or Still down An outage is open: 2 consecutive real failures from one region Investigate. It resolves automatically when the journey passes again.
What you see Why What to do
A journey failed once and no alert arrived Alerts need a confirmed outage: 2 consecutive real failures from the same region Nothing. This is expected.
A Couldn’t finish journey sent no alert These results never alert Watch the amber banner on the Journeys page
Teams did not receive a slowdown Slowdowns are WARN, and the Teams default minimum is ERROR Lower the channel’s Minimum severity to WARN
A channel received nothing for one tenant The tenant is not in the channel’s Tenant forwarding scope Add the tenant to the scope
Email recipients did not hear about the recovery or a slowdown The email channel receives the DOWN message only Use Slack or Teams for recoveries and slowdowns
The ServiceNow incident is still open after recovery Recovery adds a work note and leaves the ticket open Close the ticket in ServiceNow