mobile wallpaper 1mobile wallpaper 2mobile wallpaper 3mobile wallpaper 4
327 字
1 分钟
SOC 安全运营中心技术详解
2024-10-24

一、SOC 发展历程#

1.1 SOC 演进#

flowchart LR A["日志采集"] --> B["规则匹配"] B --> C["关联分析"] C --> D["AI/ML 检测"] D --> E["SOAR 编排"] E --> F["XDR 扩展检测"] style A fill:#87CEEB style D fill:#FFD700 style F fill:#90EE90
阶段能力局限
传统 SIEM日志采集、规则检测误报高、分析慢
智能 SIEMML 检测、UEBA上下文不足
SOAR自动化响应剧本固定
XDR跨源关联、自动化集成复杂

二、SIEM 核心架构#

2.1 日志采集#

# 日志源配置
log_sources:
# 网络设备
firewalls:
- vendor: sangfor
protocol: syslog/UDP
port: 514
format: CEF
# 主机日志
endpoints:
- type: osquery
schedule: 300s
query: SELECT * FROM process_events
# 云日志
cloud:
- provider: aliyun
service: ActionTrail
format: JSON

2.2 日志处理流水线#

flowchart TB A["原始日志"] --> B["日志解析"] B --> C["归一化"] C --> D["关联存储"] D --> E["规则引擎"] E --> F["告警"] F --> G["事件聚合"] G --> H["研判"] style B fill:#90EE90 style H fill:#FFD700

三、规则引擎#

3.1 Sigma 规则#

# Sigma 规则示例
name: 暴力破解检测
logsource:
product: windows
service: security
definition: 4625 登录失败事件
detection:
selection:
EventID: 4625
LogonType: 3
timeframe: 5m
condition: |
count() by IpAddress > 10
actions:
- log
- alert

3.2 关联分析#

# 关联规则
correlation_rules:
- name: 横向移动检测
queries:
- query: |
event_name: process_create
parent_name: wmiprc.exe
- query: |
event_name: network_connect
destination_port: 445
interval: 10m
with_keys:
- hostname
condition: join_by_hostname
- name: 持久化检测
queries:
- reg_key_enable
- scheduled_task_create
condition: same_host_same_user
severity: HIGH

四、SOAR 编排#

4.1 剧本编排#

# SOAR 剧本
playbooks:
- name: 暴力破解响应
trigger:
rule: brute_force_alert
severity: HIGH
steps:
- action: threat_intel_lookup
input: source_ip
- condition: intel.is_malicious
true_path:
- block_ip_firewall
- isolate_endpoint
- create_ticket
false_path:
- enrich_ip
- send_notification

4.2 自动响应动作#

# 响应动作库
actions_library = {
"block_ip": {
"type": "network",
"target": ["firewall", "waf"],
"parameters": ["ip", "duration"],
"verification": "session_query"
},
"isolate_endpoint": {
"type": "edr",
"target": ["sangfor_edr", "tanium"],
"parameters": ["hostname", "isolation_type"]
},
"reset_password": {
"type": "iam",
"target": ["ldap", "ad"]
}
}

五、国内厂商方案#

5.1 深信服 SIP#

sangfor_sip:
# 采集能力
collectors:
- type: agent
endpoint: EDR Agent
- type: network
devices: NGFW/AC
- type: cloud
services:
- VMWare
- K8s Audit
- Aliyun ActionTrail
# AI 检测
ai_models:
- anomaly_detection
- behavior_baseline
- threat_intel
# 联动响应
integrations:
- block_ip: NGFW
- isolate: EDR
- ticket: ITSM

5.2 奇安信 NGSOC#

qianxin_ngsoc:
features:
- 时间线分析
- 攻击链映射
- 威胁情报
- 自动化响应
# 核心能力
core:
- 关联分析
- UEBA
- SOAR

5.3 厂商对比#

厂商产品核心优势
深信服SIPAI 检测、EDR 联动
奇安信NGSOC 态势感知与安全运营实战攻防、告警管理
阿里云云安全中心云原生集成
新华三CSAP 安全分析平台统一管控、交换机集成
绿盟科技智能安全运营平台 ISOP漏洞管理、风险评估

六、态势感知#

6.1 可视化矩阵#

dashboard_widgets:
- type: security_score
metrics:
- endpoint_security
- network_security
- vulnerability_index
- compliance_rate
- type: attack_timeline
events:
- source: siem
- source: edr
- source: ngsoc

6.2 风险评估模型#

class SecurityScoreModel:
def calculate(self, assets: List[Asset]) -> float:
weights = {
'vulnerability': 0.3,
'threat': 0.25,
'compliance': 0.25,
'exposure': 0.2
}
score = 0
for asset in assets:
score += (
asset.vulnerability * weights['vulnerability'] +
asset.threat * weights['threat'] +
# ...
)
return min(100, score)

七、XDR 扩展检测#

7.1 XDR 架构#

flowchart TB subgraph 数据源 N[网络探针] E[EDR 日志] C[云日志] end N --> U[统一数据湖] E --> U C --> U U --> D[检测引擎] D --> A[告警] A --> R[响应动作] style U fill:#90EE90 style D fill:#FFD700

7.2 跨域关联#

xdr_correlation:
# 网络行为
network_events:
- DNS 查询
- TLS 指纹
- HTTP 行为序列
# 端点行为
endpoint_events:
- 进程链
- 注册表修改
- WMI 执行
# 关联分析
rules:
- name: 无文件攻击检测
query: |
dns_query + process_create + registry_modify
model: supervised

九、总结#

SOC 的核心是海量日志的采集、分析、响应能力。国内厂商在 AI 检测、自动化响应方面已接近国际水平。

支持与分享

如果这篇文章对你有帮助,欢迎支持作者或分享给更多人

SOC 安全运营中心技术详解
https://blog.souloss.com/posts/cloud-security/soc-security-operations-center/
作者
Souloss
发布于
2024-10-24
许可协议
CC BY-NC-SA 4.0

部分信息可能已经过时