NDR 5.5.4 Service Unavailable: Diagnosis & Resolution

Fix 5.5.4 "Service Temporarily Unavailable" errors. This error indicates transient connector issues or throttling. Includes decision tree for identifying root cause and safe relief procedures.

Symptom Definition

Senders receive NDR:

451 4.4.0 Service Temporarily Unavailable 550 5.5.4 Service Unavailable

Key characteristics:

  • This is a TEMPORARY error (may succeed on retry)
  • Usually affects multiple recipients/domains, not just one
  • Often correlates with high mail volume or connector overload
  • Intermittent rather than persistent
  • Different from 5.1.x (permanent recipient issues)

⚠️ Business Consequence: Why This Matters

  • Financial Impact: Transport queue backlog = delayed invoices/contracts ($3K–$8K per hour)
  • Compliance Exposure: Time-sensitive notifications missed (financial disclosures, legal deadlines)
  • Operational Risk: Mass communication failures impact business continuity
  • Service Degradation: Cascading delays affect entire organization's mail flow

Average diagnosis time: 8–12 minutes — prevents queue escalation.

Root Causes

  • Connector overloaded: Too many messages queued, connector resources exhausted
  • Throttling policy triggered: Sender/recipient rate limits exceeded
  • EOP processing backed up: Exchange Online Protection queue deep
  • Temporary network issue: Brief connectivity loss between on-premises and EOP
  • Transport service restarting: Mail server service restart in progress
  • Directory service unavailable: Azure AD or on-premises AD temporarily unreachable

Diagnostic Steps

Step 1: Check Connector Queue Depth (2 min)

Get-Queue | Select-Object Identity, MessageCount

If any queue shows >1000 messages → connector overloaded. Proceed to Relief 1.

If all queues <500 → go to Step 2.

Step 2: Check Throttling Policy (3 min)

Get-Mailbox -Identity sender@domain | Select-Object ThrottlingPolicy

If custom policy applied with low limits → adjust policy (see Relief 2).

Step 3: Check Connector Health (2 min)

In Exchange admin center → Connectors → Check status (green = healthy, orange = warning)

If orange/red → restart service (see Relief 3).

Step 4: Check Network Connectivity (2 min)

telnet outlook.office365.com 25

If timeout/connection refused → network issue. Contact network team.

Safe Relief Procedures

Relief 1: Clear Transport Queue

  1. Restart-Service MSExchangeTransport
  2. Wait 5 minutes and recheck queue depth
  3. If 5.5.4 errors stop → queue cleared successfully

Relief 2: Adjust Throttling Policy

  1. Set-ThrottlingPolicy -Identity "Policy" -RecipientRateLimit 900/minute
  2. Retest sending to confirm 5.5.4 resolves

Relief 3: Restart Transport Service

  1. Restart-Service MSExchangeTransport
  2. Wait 2 minutes and test mail delivery

When to Escalate

  • 5.5.4 persists after queue cleared and service restarted
  • All queues healthy but 5.5.4 continues (indicates EOP-side issue)
  • Network test fails (indicates upstream connectivity problem)

Related Guides

NDR 550 5.1.1 Throttling Detection Delayed Email