Demo Sample data — this is an interactive demo Northwind Logistics is a made-up company. Nothing here is connected to a real account.

Start free 14-day trial — no cardCheck your own domain free
Northwind Logistics · Demo

Findings

Everything that needs attention, worst first. Open a finding to see what it means and how to fix it.

Showing 13 findings

CriticalTLS certificate expires in 6 daysapi-northwind-tlsKey Vault kv-nw-prod-eus (East US)Azure24 days ago

What it means

The certificate behind api.northwind.example expires in 6 days. It has no auto-renewal policy, so customers will see a browser warning and integrations will fail the TLS handshake when it lapses.

How to fix it

  1. Open Key Vault kv-nw-prod-eus → Certificates → api-northwind-tls.
  2. Choose "New version" and renew with the same issuer, or set an auto-renew lifetime action at 30 days before expiry.
  3. Rebind the new version on the App Gateway listener (it follows the versionless secret ID automatically).
  4. CloudMonitor confirms the new expiry on its next scan and closes this finding.
CriticalApp registration client secret expires in 3 daysnw-billing-syncEntra ID app registrationAzure87 days ago

What it means

The nightly billing sync signs in with this client secret. When it expires the job fails silently and invoices stop flowing to the ERP.

How to fix it

  1. Entra ID → App registrations → nw-billing-sync → Certificates & secrets → New client secret (12 months).
  2. Store the new value in Key Vault secret billing-sync-client-secret; the job reads it from there.
  3. Run the sync once by hand, then delete the old secret.
CriticalEndpoint down: 3 failed checks in a rowhttps://track.northwind.example/healthChecked from 3 regionsEndpointToday

What it means

The shipment tracking site returned HTTP 503 from every region for the last 3 checks (15 minutes). The last good response was 18 minutes ago.

How to fix it

  1. Check the app service plan for track-web: a deployment slot swap at 09:12 is the most recent change.
  2. Roll back the slot swap, or restart the app if the new build is known good.
  3. CloudMonitor closes the incident after 2 healthy checks and posts the outage length to Slack.
CriticalS3 bucket allows public readnw-invoice-exportsus-west-2 · account 1111-2222-3333AWS2 days ago

What it means

A bucket policy grants s3:GetObject to "*". The bucket holds CSV invoice exports. Block Public Access is off at the bucket level.

How to fix it

  1. S3 → nw-invoice-exports → Permissions → turn on Block Public Access (all four settings).
  2. Remove the "Principal": "*" statement from the bucket policy.
  3. If a partner needs the files, share them with pre-signed URLs instead.
CriticalGlobal admin without MFA[email protected]Entra ID · Global AdministratorMicrosoft 36541 days ago

What it means

One of 4 Global Administrators has no MFA method registered and is not covered by a Conditional Access policy. A phished password is full tenant takeover.

How to fix it

  1. Require MFA for all admin roles with a Conditional Access policy (or turn on security defaults).
  2. Ask the account owner to register Microsoft Authenticator or a FIDO2 key.
  3. Consider removing the role: 2 Global Administrators plus break-glass accounts is the usual target.
WarningVM idle for 30 days (avg CPU 1.2%)vm-erp-test-02rg-erp-test · West US 2 · D8s v5Azure30 days ago

What it means

This test VM averaged 1.2% CPU and no network traffic for 30 days. It costs about $561 a month while running.

How to fix it

  1. Confirm with the ERP team that the test environment is not in use.
  2. Stop (deallocate) the VM to stop compute charges, or snapshot the disk and delete it.
  3. Add an auto-shutdown schedule if the box is needed occasionally.
Warning3 managed disks unattached for 60+ daysdisk-old-sql-0, -1, -2rg-legacy-sql · East US · Premium SSD 1 TBAzure64 days ago

What it means

The disks were left behind when the old SQL VM was deleted. Premium SSD is billed whether or not anything uses it.

How to fix it

  1. Snapshot each disk to Standard storage if you might need the data.
  2. Delete the three disks.
Warning2 Elastic IPs not attached to a running instanceeipalloc-0a1b…, eipalloc-0c2d…us-east-1AWS19 days ago

What it means

Idle Elastic IPs are billed hourly. Both belonged to instances that were stopped in a cleanup.

How to fix it

  1. EC2 → Elastic IPs → select both → Release Elastic IP addresses.
Warning7 licensed accounts with no sign-in for 90+ days7 usersMicrosoft 365 Business PremiumMicrosoft 36512 days ago

What it means

Seven accounts still hold a paid licence and an enabled sign-in but nobody has used them in 90 days. Three belong to people who left.

How to fix it

  1. Review the list on the Microsoft 365 page and block sign-in for leavers.
  2. Convert leaver mailboxes to shared mailboxes, then remove the licence.
WarningService account key older than 365 days[email protected]Project nw-analyticsGCP33 days ago

What it means

A user-managed key was created 412 days ago and is still used daily. Long-lived keys are the most common way GCP projects leak.

How to fix it

  1. Create a new key, deploy it, and disable the old one for a week before deleting it.
  2. Better: switch the ETL job to Workload Identity Federation and delete the key entirely.
WarningCluster on a Kubernetes version leaving support in 45 daysaks-nw-prodAKS · East USKubernetes15 days ago

What it means

The control plane and both node pools run a minor version that drops out of AKS standard support in 45 days.

How to fix it

  1. Upgrade the control plane one minor version, then each node pool.
  2. Run the upgrade on aks-nw-staging first; check deprecated APIs with the AKS diagnostics blade.
InfoDMARC policy is "none" (monitor only)northwind.exampleDNS TXT _dmarcDomain120 days ago

What it means

SPF and DKIM pass, but DMARC only reports. Spoofed mail from your domain is still delivered.

How to fix it

  1. Read two weeks of DMARC reports to confirm every legitimate sender passes.
  2. Move to p=quarantine, then p=reject.
InfoSQL database backup retention below 14 dayssqldb-orderssql-nw-prod · East USAzure9 days ago

What it means

Point-in-time restore is set to 7 days. Most audits expect 14 to 35.

How to fix it

  1. SQL server → Backups → Retention policies → set point-in-time restore to 14 days or more.

See this for your own cloud

Connect Azure, AWS, Google Cloud or Microsoft 365 with read-only access. Your first findings arrive within the hour.