You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: learn-pr/wwl-sci/design-solutions-securing-server-client-endpoints/includes/2-specify-server-security-requirements.md
+26-9Lines changed: 26 additions & 9 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -9,6 +9,7 @@ When specifying server security requirements, organize them into categories that
9
9
-**Identity and access management**: Authentication methods, privileged access controls, service account management
@@ -17,12 +18,19 @@ For each category, specify requirements that are measurable and auditable. For e
17
18
18
19
### MCSB Endpoint Security controls
19
20
20
-
The Microsoft Cloud Security Benchmark (MCSB) provides a structured framework for endpoint security requirements. The Endpoint Security (ES) domain defines three controls specific to servers and other endpoints:
21
+
The Microsoft Cloud Security Benchmark (MCSB) provides a structured framework for endpoint security requirements. In MCSB v2, the Endpoint Security (ES) domain is organized into two pillars:
22
+
23
+
**Cloud endpoint threat protection** — Deploy comprehensive threat detection and response capabilities for servers, including anti-malware and behavioral analysis:
21
24
22
25
| Control | Requirement | Implementation guidance |
23
26
| --- | --- | --- |
24
27
| ES-1 | Use Endpoint Detection and Response (EDR) | Deploy EDR solutions like Microsoft Defender for Endpoint to detect, investigate, and respond to advanced threats on servers |
25
28
| ES-2 | Use modern anti-malware software | Require anti-malware solutions that provide real-time protection, behavior monitoring, and integration with cloud-based threat intelligence |
29
+
30
+
**Cloud endpoint security configuration** — Enforce security baselines and hardening standards across all servers:
31
+
32
+
| Control | Requirement | Implementation guidance |
33
+
| --- | --- | --- |
26
34
| ES-3 | Ensure anti-malware software and signatures are updated | Configure automatic updates for anti-malware definitions and engines; specify maximum acceptable age for signatures |
27
35
28
36
These controls apply across Windows and Linux servers in Azure, AWS, GCP, and on-premises environments. Microsoft Defender for Servers implements ES-1 and ES-2 through its integration with Defender for Endpoint. For detailed implementation guidance, see [Endpoint security controls in MCSB](/security/benchmark/azure/mcsb-endpoint-security).
@@ -59,7 +67,7 @@ Server security requirements vary based on where servers are deployed. Your spec
59
67
60
68
For Azure-hosted servers, use platform capabilities in your requirements:
61
69
62
-
-**Microsoft Defender for Servers**: Require either Plan 1 or Plan 2 based on protection needs. Plan 1 provides Defender for Endpoint integration for EDR capabilities. Plan 2 adds agentless scanning, file integrity monitoring, and just-in-time VM access.
70
+
-**Microsoft Defender for Servers**: Require either Plan 1 or Plan 2 based on protection needs. Plan 1 provides Defender for Endpoint integration for EDR capabilities. Plan 2 adds agentless scanning (vulnerabilities, secrets, and malware), file integrity monitoring, just-in-time VM access, compliance assessment, and a free data ingestion benefit. Defender for Servers no longer relies on the Log Analytics agent or Azure Monitor Agent — all security features are delivered through the Defender for Endpoint integration and agentless machine scanning.
63
71
-**Network security groups**: Require NSGs on all subnets with rules that follow least-privilege principles
64
72
-**Azure Bastion**: Require Azure Bastion for administrative access instead of exposing RDP/SSH ports
65
73
-**Managed identities**: Require managed identities for Azure resource access instead of stored credentials
@@ -69,7 +77,7 @@ For Azure-hosted servers, use platform capabilities in your requirements:
69
77
70
78
For servers in AWS or GCP, specify how to integrate with your security management:
71
79
72
-
-**Azure Arc onboarding**: Require all non-Azure servers to connect through Azure Arc for centralized management
80
+
-**Azure Arc onboarding**: Require all non-Azure servers to connect through Azure Arc for centralized management. The multicloud connector enabled by Azure Arc can auto-discover EC2 instances and GCP VMs and install the Connected Machine agent at scale.
73
81
-**Connector configuration**: Specify whether to connect at the organization/account level or individual project/account level
74
82
-**Defender for Servers coverage**: Extend Microsoft Defender for Servers protection to EC2 instances and Compute Engine VMs
75
83
-**Compliance standards**: Apply consistent standards across clouds—AWS Foundational Security Best Practices, GCP CIS benchmarks, and MCSB all map to common controls
@@ -78,7 +86,8 @@ For servers in AWS or GCP, specify how to integrate with your security managemen
-**Azure Arc deployment**: Require the Azure Connected Machine agent on all servers to enable cloud-based management
89
+
-**Azure Arc deployment**: Require the Azure Connected Machine agent on all servers to enable cloud-based management and full Plan 2 functionality
90
+
-**Direct Defender for Endpoint onboarding**: For servers where Azure Arc isn't feasible, installing the Defender for Endpoint agent directly provides Plan 1 functionality only (EDR and premium Defender Vulnerability Management features, but no agentless scanning, FIM, or JIT access)
82
91
-**Network connectivity**: Specify required endpoints for Azure Arc, Defender for Cloud, and any extensions
83
92
-**Proxy configuration**: Define proxy requirements for servers without direct internet access
84
93
-**Update management**: Integrate with Azure Update Manager for centralized patching across hybrid environments
@@ -101,7 +110,11 @@ Specify scanning frequency and remediation timelines:
101
110
102
111
**File integrity monitoring**
103
112
104
-
For servers processing sensitive data, require monitoring of critical system files and configuration files with alerts on unauthorized changes.
113
+
For servers processing sensitive data, require monitoring of critical system files and configuration files with alerts on unauthorized changes. File integrity monitoring is now delivered through the Defender for Endpoint integration in Defender for Servers Plan 2, replacing the previous Log Analytics agent-based approach.
114
+
115
+
**Backup and recovery**
116
+
117
+
Require immutable or offline backups for critical servers to protect against ransomware. Specify backup frequency, retention periods, and regular restoration testing. MCSB includes a dedicated Backup and Recovery (BR) domain with controls for backup configuration, protection, and validation.
105
118
106
119
**Security alerts**
107
120
@@ -127,11 +140,15 @@ When specifying Defender for Servers requirements, choose the appropriate plan b
|**Agentless scanning**| Not available | ✓ Software inventory, secrets, malware |
143
+
|**Defender for Endpoint (EDR)**| ✓ Included | ✓ Included |
144
+
|**Vulnerability assessment (agent-based)**| ✓ Included | ✓ Included |
145
+
|**Vulnerability assessment (agentless)**| Not available | ✓ Included |
146
+
|**Agentless malware and secrets scanning**| Not available | ✓ Software inventory, secrets, malware |
147
+
|**Premium Defender Vulnerability Management**| Not available | ✓ Security baselines, block vulnerable apps, certificate and hardware assessment |
133
148
|**Just-in-time VM access**| Not available | ✓ Reduces attack surface |
134
-
|**File integrity monitoring**| Not available | ✓ Detects unauthorized changes |
149
+
|**File integrity monitoring**| Not available | ✓ Via Defender for Endpoint integration |
150
+
|**Compliance assessment**| Not available | ✓ Regulatory standards assessment |
151
+
|**Free data ingestion**| Not available | ✓ 500 MB/day for specific data types |
135
152
136
153
For most organizations, specify Plan 1 as the baseline requirement for all servers, with Plan 2 required for servers that process sensitive data, are internet-facing, or are otherwise deemed critical.
0 commit comments