feat: K8S智能诊断平台 - 一键诊断+AI分析+Web仪表盘

- 后端: FastAPI + kubectl 9大检查项(Pod/Node/Event/Deploy/Service/PVC/资源/网络) - AI Agent: OpenAI兼容接口自动分析故障根因+多轮追问对话 - 前端: Vue3 + Element Plus 仪表盘, 健康评分, 问题列表, 资源监控 - 一键启动脚本 start.sh
2026-07-25 10:32:26 +08:00
commit d5d7e45ded
14 changed files with 2574 additions and 0 deletions
@@ -0,0 +1,80 @@
+"""AI Agent 分析模块 - 调用 OpenAI 兼容接口分析诊断结果"""
+import httpx
+from backend.config import AI_API_BASE, AI_API_KEY, AI_MODEL
+
+SYSTEM_PROMPT = """你是一位资深的 Kubernetes 运维专家和 SRE 工程师。你的任务是分析 K8S 集群诊断报告，找出根本原因并给出修复建议。
+
+分析要求：
+1. **问题定位**：明确指出每个问题的根本原因（Root Cause），不要只描述表面现象
+2. **影响评估**：说明每个问题对业务的潜在影响
+3. **修复方案**：给出具体的修复命令或操作步骤（kubectl 命令、YAML 修改等）
+4. **优先级排序**：按紧急程度排序，先处理影响最大的问题
+5. **预防建议**：给出避免类似问题再次发生的建议
+
+输出格式：使用清晰的 Markdown，包含标题、列表、代码块。语言使用中文。"""
+
+
+async def analyze_with_ai(report: str, question: str = "") -> dict:
+    """调用 AI 分析诊断报告"""
+    user_msg = f"以下是 K8S 集群诊断报告，请进行详细分析：\n\n{report}"
+    if question:
+        user_msg += f"\n\n用户额外问题：{question}"
+    
+    try:
+        async with httpx.AsyncClient(timeout=120) as client:
+            resp = await client.post(
+                f"{AI_API_BASE}/chat/completions",
+                headers={
+                    "Authorization": f"Bearer {AI_API_KEY}",
+                    "Content-Type": "application/json",
+                },
+                json={
+                    "model": AI_MODEL,
+                    "messages": [
+                        {"role": "system", "content": SYSTEM_PROMPT},
+                        {"role": "user", "content": user_msg},
+                    ],
+                    "temperature": 0.3,
+                    "max_tokens": 4096,
+                },
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            content = data["choices"][0]["message"]["content"]
+            return {"success": True, "analysis": content, "model": AI_MODEL}
+    except httpx.HTTPStatusError as e:
+        return {"success": False, "analysis": f"AI 接口返回错误: {e.response.status_code} - {e.response.text[:500]}", "model": AI_MODEL}
+    except httpx.ConnectError:
+        return {"success": False, "analysis": f"无法连接 AI 服务 ({AI_API_BASE})，请检查配置", "model": AI_MODEL}
+    except Exception as e:
+        return {"success": False, "analysis": f"AI 分析失败: {str(e)}", "model": AI_MODEL}
+
+
+async def chat_with_ai(messages: list[dict], context: str = "") -> dict:
+    """多轮对话 - 基于诊断上下文追问"""
+    msgs = [{"role": "system", "content": SYSTEM_PROMPT}]
+    if context:
+        msgs.append({"role": "system", "content": f"当前诊断上下文：\n{context}"})
+    msgs.extend(messages)
+    
+    try:
+        async with httpx.AsyncClient(timeout=120) as client:
+            resp = await client.post(
+                f"{AI_API_BASE}/chat/completions",
+                headers={
+                    "Authorization": f"Bearer {AI_API_KEY}",
+                    "Content-Type": "application/json",
+                },
+                json={
+                    "model": AI_MODEL,
+                    "messages": msgs,
+                    "temperature": 0.3,
+                    "max_tokens": 4096,
+                },
+            )
+            resp.raise_for_status()
+            data = resp.json()
+            content = data["choices"][0]["message"]["content"]
+            return {"success": True, "reply": content}
+    except Exception as e:
+        return {"success": False, "reply": f"AI 对话失败: {str(e)}"}