Important: The architecture and investigation model below describe the scenario represented by this case study. A Spamhaus listing should always be verified against the exact connecting IP, listing type, evidence and the network owner’s authoritative records before assigning a specific root cause.

1. The email delivery architecture

The first lesson is simple: trace the message all the way to the recipient. When Microsoft 365 Exchange Online hands outbound mail to a Netskope SMTP proxy or Email Connector, the tenant is not necessarily the final Internet-facing SMTP endpoint. Downstream Netskope egress infrastructure can become the source IP seen by the recipient mail server.

EMAIL FLOW / ACTUAL INTERNET-FACING PATH
User / ApplicationMessage origin
Microsoft 365 / Exchange OnlineOutbound mail service
Netskope SMTP ProxyEmail Connector · inspection · DLP
Netskope Egress IPInternet-facing upstream source
InternetSMTP delivery path
Recipient Mail ServerAccept / reject decision
DLP / Security Inspection — policy enforcement, threat protection and content scanning may operate within the security path.

The critical diagnostic point is the actual connecting IP. If the recipient sees a Netskope-owned or upstream egress address, that address—not the original Microsoft 365 service address—is the first candidate for reputation investigation.

2. Where the Spamhaus condition fits

A recipient mail server can apply reputation checks to the source IP that connects to it. If that source IP is listed by a reputation service, the recipient may reject the SMTP session or message. The exact behavior depends on the recipient’s policy and the listing involved.

What must be established

  • Exact public IP
  • Recipient SMTP response
  • Timestamp and message context
  • Listing name/type and evidence
  • IP ownership and responsibility

What can cause rejection

  • Spam / unsolicited activity signals
  • Compromised sender or customer systems
  • Poor reputation or complaint history
  • Shared infrastructure impact

What must not be assumed

  • M365 was the connecting IP
  • A listing automatically proves organizational abuse
  • SPF/DKIM/DMARC overrides recipient IP reputation
  • The listing alone proves the RCA

3. Why shared egress infrastructure matters

Security services may operate shared or pooled egress infrastructure. That creates a risk model different from a dedicated corporate SMTP IP: reputation can be influenced by activity associated with infrastructure used by multiple customers or systems.

This does not by itself establish the cause of a listing. It explains why IP ownership, allocation and upstream responsibility must be part of the RCA.

4. How to investigate the incident

01 — Capture the failure. Preserve the NDR, SMTP response code, enhanced status, recipient, message ID and timestamps.
02 — Trace the route. Confirm Exchange Online → Netskope connector → egress infrastructure → recipient.
03 — Identify the source IP. Use recipient-side evidence, SMTP transcripts, message traces and authoritative network records where available.
04 — Verify reputation. Check the exact IP and exact listing. Record the listing type and evidence rather than relying on screenshots or assumptions.
05 — Determine ownership. Establish whether the address belongs to the customer, Netskope, an upstream provider or another service boundary.
06 — Correlate controls. Review DLP, security inspection, connector logs and relevant sender activity for the same time window.
07 — Remediate. Remove the underlying abusive or compromised activity, correct configuration, engage the responsible network owner and follow the applicable delisting process.
08 — Validate. Send controlled test messages and confirm acceptance at the recipient side. Record the result.
INCIDENT INVESTIGATION FLOW
01Delivery failure
02NDR / SMTP response
03Actual connecting IP
04Trace complete mail path
05Verify reputation
06Remediate / delist
07Controlled retest

5. Remediation and delisting

The remediation owner depends on who controls the listed IP. If the address is part of provider-managed egress infrastructure, the service or network owner normally needs to investigate the contributing activity and coordinate the reputation remediation.

A strong remediation record should contain:

  • Exact listed IP or CIDR
  • Listing type and verification evidence
  • Observed SMTP rejection
  • Scope and duration of impact
  • Abuse or compromise investigation
  • Corrective action
  • Delisting request / provider engagement
  • Post-remediation delivery tests

6. What happens if this occurs again?

Business continuity should be designed before the next incident. A fallback route can reduce outage duration, but it should not simply become an uncontrolled bypass around security inspection.

Primary path

Microsoft 365 → Netskope Email Connector → Netskope egress → Internet → recipient.

Alternate path

Microsoft 365 → approved secure email gateway / MTA → alternate egress → Internet → recipient.

Security requirement

Preserve DLP, malware inspection, TLS, SPF, DKIM, DMARC, logging and compliance controls where required.

7. Fallback design checklist

  • Pre-approved alternate routing
  • Clear failover criteria and ownership
  • Real-time alerting for SMTP failures
  • Maintained DLP and security controls
  • SPF/DKIM/DMARC alignment for the alternate route
  • Documented rollback and recovery procedure
  • Continuous egress-IP reputation monitoring
  • Regular controlled testing of the fallback path

8. Key takeaways

Trace

Follow the message to the recipient. Do not stop at Exchange Online when downstream security infrastructure is part of the route.

Evidence

Identify the real egress IP and separate observed facts from assumptions.

Continuity

Build a tested fallback path without turning business continuity into a security bypass.

9. Conclusion

The most useful lesson from this type of incident is not simply “check the block list.” It is to understand the complete delivery architecture, identify the actual Internet-facing source, preserve evidence, establish ownership, remediate at the correct boundary and validate the result from the recipient’s perspective.

When the mail path crosses a third-party security service, that service becomes part of the operational dependency chain. Monitoring, clear ownership and a tested continuity plan therefore belong in the architecture—not only in the incident response document.

Related engineering knowledge

Engineering principle: Always identify the real Internet-facing SMTP egress point before assigning responsibility for an outbound email delivery failure.