mirror of
https://github.com/infiniflow/ragflow.git
synced 2025-12-08 20:42:30 +08:00
Feat: add splitter (#10161)
### What problem does this PR solve? ### Type of change - [x] New Feature (non-breaking change which adds functionality) --------- Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: Lynn <lynn_inf@hotmail.com> Co-authored-by: chanx <1243304602@qq.com> Co-authored-by: balibabu <cike8899@users.noreply.github.com> Co-authored-by: 纷繁下的无奈 <zhileihuang@126.com> Co-authored-by: huangzl <huangzl@shinemo.com> Co-authored-by: writinwaters <93570324+writinwaters@users.noreply.github.com> Co-authored-by: Wilmer <33392318@qq.com> Co-authored-by: Adrian Weidig <adrianweidig@gmx.net> Co-authored-by: Zhichang Yu <yuzhichang@gmail.com> Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com> Co-authored-by: Yongteng Lei <yongtengrey@outlook.com> Co-authored-by: Liu An <asiro@qq.com> Co-authored-by: buua436 <66937541+buua436@users.noreply.github.com> Co-authored-by: BadwomanCraZY <511528396@qq.com> Co-authored-by: cucusenok <31804608+cucusenok@users.noreply.github.com> Co-authored-by: Russell Valentine <russ@coldstonelabs.org> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: Billy Bao <newyorkupperbay@gmail.com> Co-authored-by: Zhedong Cen <cenzhedong2@126.com> Co-authored-by: TensorNull <129579691+TensorNull@users.noreply.github.com> Co-authored-by: TensorNull <tensor.null@gmail.com>
This commit is contained in:
105
api/apps/HEALTHCHECK_TESTING.md
Normal file
105
api/apps/HEALTHCHECK_TESTING.md
Normal file
@ -0,0 +1,105 @@
|
||||
# 健康检查与 Kubernetes 探针简明说明
|
||||
|
||||
本文件说明:什么是 K8s 探针、如何用 `/v1/system/healthz` 做健康检查,以及下文用例中的关键词含义。
|
||||
|
||||
## 什么是 K8s 探针(Probe)
|
||||
- 探针是 K8s 用来“探测”容器是否健康/可对外服务的机制。
|
||||
- 常见三类:
|
||||
- livenessProbe:活性探针。失败时 K8s 会重启容器,用于“应用卡死/失去连接时自愈”。
|
||||
- readinessProbe:就绪探针。失败时 Endpoint 不会被加入 Service 负载均衡,用于“应用尚未准备好时不接流量”。
|
||||
- startupProbe:启动探针。给慢启动应用更长的初始化窗口,期间不执行 liveness/readiness。
|
||||
- 这些探针通常通过 HTTP GET 访问一个公开且轻量的健康端点(无需鉴权),以 HTTP 状态码判定结果:200=通过;5xx/超时=失败。
|
||||
|
||||
## 本项目健康端点
|
||||
- 已实现:`GET /v1/system/healthz`(无需认证)。
|
||||
- 语义:
|
||||
- 200:关键依赖正常。
|
||||
- 500:任一关键依赖异常(当前判定为 DB 或 Chat)。
|
||||
- 响应体:JSON,最小字段 `status, db, chat`;并包含 `redis, doc_engine, storage` 等可观测项。失败项会在 `_meta` 中包含 `error/elapsed`。
|
||||
- 示例(DB 故障):
|
||||
```json
|
||||
{"status":"nok","chat":"ok","db":"nok"}
|
||||
```
|
||||
|
||||
## 用例背景(Problem/use case)
|
||||
- 现状:Ragflow 跑在 K8s,数据库是 AWS RDS Postgres,凭证由 Secret Manager 管理并每 7 天轮换。轮换后应用连接失效,需要手动重启 Pod 才能重新建立连接。
|
||||
- 目标:通过 K8s 探针自动化检测并重启异常 Pod,减少人工操作。
|
||||
- 需求:一个“无需鉴权”的公共健康端点,能在依赖异常时返回非 200(如 500)且提供 JSON 详情。
|
||||
- 现已满足:`/v1/system/healthz` 正是为此设计。
|
||||
|
||||
## 关键术语解释(对应你提供的描述)
|
||||
- Ragflow instance:部署在 K8s 的 Ragflow 服务。
|
||||
- AWS RDS Postgres:托管的 PostgreSQL 数据库实例。
|
||||
- Secret Manager rotation:Secrets 定期轮换(每 7 天),会导致旧连接失效。
|
||||
- Probes(K8s 探针):liveness/readiness,用于自动重启或摘除不健康实例。
|
||||
- Public endpoint without API key:无需 Authorization 的 HTTP 路由,便于探针直接访问。
|
||||
- Dependencies statuses:依赖健康状态(db、chat、redis、doc_engine、storage 等)。
|
||||
- HTTP 500 with JSON:当依赖异常时返回 500,并附带 JSON 说明哪个子系统失败。
|
||||
|
||||
## 快速测试
|
||||
- 正常:
|
||||
```bash
|
||||
curl -i http://<host>/v1/system/healthz
|
||||
```
|
||||
- 制造 DB 故障(docker-compose 示例):
|
||||
```bash
|
||||
docker compose stop db && curl -i http://<host>/v1/system/healthz
|
||||
```
|
||||
(预期 500,JSON 中 `db:"nok"`)
|
||||
|
||||
## 更完整的测试清单
|
||||
### 1) 仅查看 HTTP 状态码
|
||||
```bash
|
||||
curl -s -o /dev/null -w "%{http_code}\n" http://<host>/v1/system/healthz
|
||||
```
|
||||
期望:`200` 或 `500`。
|
||||
|
||||
### 2) Windows PowerShell
|
||||
```powershell
|
||||
# 状态码
|
||||
(Invoke-WebRequest -Uri "http://<host>/v1/system/healthz" -Method GET -TimeoutSec 3 -ErrorAction SilentlyContinue).StatusCode
|
||||
# 完整响应
|
||||
Invoke-RestMethod -Uri "http://<host>/v1/system/healthz" -Method GET
|
||||
```
|
||||
|
||||
### 3) 通过 kubectl 端口转发本地测试
|
||||
```bash
|
||||
# 前端/网关暴露端口不同环境自行调整
|
||||
kubectl port-forward deploy/<your-deploy> 8080:80 -n <ns>
|
||||
curl -i http://127.0.0.1:8080/v1/system/healthz
|
||||
```
|
||||
|
||||
### 4) 制造常见失败场景
|
||||
- DB 失败(推荐):
|
||||
```bash
|
||||
docker compose stop db
|
||||
curl -i http://<host>/v1/system/healthz # 预期 500
|
||||
```
|
||||
- Chat 失败(可选):将 `CHAT_CFG` 的 `factory`/`base_url` 设为无效并重启后端,再请求应为 500,且 `chat:"nok"`。
|
||||
- Redis/存储/文档引擎:停用对应服务后再次请求,可在 JSON 中看到相应字段为 `"nok"`(不影响 200/500 判定)。
|
||||
|
||||
### 5) 浏览器验证
|
||||
- 直接打开 `http://<host>/v1/system/healthz`,在 DevTools Network 查看 200/500;页面正文就是 JSON。
|
||||
- 反向代理注意:若有自定义 500 错页,需对 `/healthz` 关闭错误页拦截(如 `proxy_intercept_errors off;`)。
|
||||
|
||||
## K8s 探针示例
|
||||
```yaml
|
||||
readinessProbe:
|
||||
httpGet:
|
||||
path: /v1/system/healthz
|
||||
port: 80
|
||||
initialDelaySeconds: 5
|
||||
periodSeconds: 10
|
||||
timeoutSeconds: 2
|
||||
failureThreshold: 1
|
||||
livenessProbe:
|
||||
httpGet:
|
||||
path: /v1/system/healthz
|
||||
port: 80
|
||||
initialDelaySeconds: 10
|
||||
periodSeconds: 10
|
||||
timeoutSeconds: 2
|
||||
failureThreshold: 3
|
||||
```
|
||||
|
||||
提示:如有反向代理(Nginx)自定义 500 错页,需对 `/healthz` 关闭错误页拦截,以便保留 JSON。
|
||||
@ -28,6 +28,7 @@ from api.db import CanvasCategory, FileType
|
||||
from api.db.services.canvas_service import CanvasTemplateService, UserCanvasService, API4ConversationService
|
||||
from api.db.services.document_service import DocumentService
|
||||
from api.db.services.file_service import FileService
|
||||
from api.db.services.task_service import queue_dataflow
|
||||
from api.db.services.user_service import TenantService
|
||||
from api.db.services.user_canvas_version import UserCanvasVersionService
|
||||
from api.settings import RetCode
|
||||
@ -48,14 +49,6 @@ def templates():
|
||||
return get_json_result(data=[c.to_dict() for c in CanvasTemplateService.query(canvas_category=CanvasCategory.Agent)])
|
||||
|
||||
|
||||
@manager.route('/list', methods=['GET']) # noqa: F821
|
||||
@login_required
|
||||
def canvas_list():
|
||||
return get_json_result(data=sorted([c.to_dict() for c in \
|
||||
UserCanvasService.query(user_id=current_user.id, canvas_category=CanvasCategory.Agent)], key=lambda x: x["update_time"]*-1)
|
||||
)
|
||||
|
||||
|
||||
@manager.route('/rm', methods=['POST']) # noqa: F821
|
||||
@validate_request("canvas_ids")
|
||||
@login_required
|
||||
@ -77,9 +70,10 @@ def save():
|
||||
if not isinstance(req["dsl"], str):
|
||||
req["dsl"] = json.dumps(req["dsl"], ensure_ascii=False)
|
||||
req["dsl"] = json.loads(req["dsl"])
|
||||
cate = req.get("canvas_category", CanvasCategory.Agent)
|
||||
if "id" not in req:
|
||||
req["user_id"] = current_user.id
|
||||
if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=CanvasCategory.Agent):
|
||||
if UserCanvasService.query(user_id=current_user.id, title=req["title"].strip(), canvas_category=cate):
|
||||
return get_data_error_result(message=f"{req['title'].strip()} already exists.")
|
||||
req["id"] = get_uuid()
|
||||
if not UserCanvasService.save(**req):
|
||||
@ -148,6 +142,14 @@ def run():
|
||||
if not isinstance(cvs.dsl, str):
|
||||
cvs.dsl = json.dumps(cvs.dsl, ensure_ascii=False)
|
||||
|
||||
if cvs.canvas_category == CanvasCategory.DataFlow:
|
||||
task_id = get_uuid()
|
||||
flow_id = get_uuid()
|
||||
ok, error_message = queue_dataflow(dsl=cvs.dsl, tenant_id=user_id, file=files[0], task_id=task_id, flow_id=flow_id, priority=0)
|
||||
if not ok:
|
||||
return server_error_response(error_message)
|
||||
return get_json_result(data={"task_id": task_id, "message_id": flow_id})
|
||||
|
||||
try:
|
||||
canvas = Canvas(cvs.dsl, current_user.id, req["id"])
|
||||
except Exception as e:
|
||||
@ -332,7 +334,7 @@ def test_db_connect():
|
||||
if req["db_type"] in ["mysql", "mariadb"]:
|
||||
db = MySQLDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
|
||||
password=req["password"])
|
||||
elif req["db_type"] == 'postgresql':
|
||||
elif req["db_type"] == 'postgres':
|
||||
db = PostgresqlDatabase(req["database"], user=req["username"], host=req["host"], port=req["port"],
|
||||
password=req["password"])
|
||||
elif req["db_type"] == 'mssql':
|
||||
@ -383,22 +385,31 @@ def getversion( version_id):
|
||||
return get_json_result(data=f"Error getting history file: {e}")
|
||||
|
||||
|
||||
@manager.route('/listteam', methods=['GET']) # noqa: F821
|
||||
@manager.route('/list', methods=['GET']) # noqa: F821
|
||||
@login_required
|
||||
def list_canvas():
|
||||
keywords = request.args.get("keywords", "")
|
||||
page_number = int(request.args.get("page", 1))
|
||||
items_per_page = int(request.args.get("page_size", 150))
|
||||
orderby = request.args.get("orderby", "create_time")
|
||||
desc = request.args.get("desc", True)
|
||||
try:
|
||||
canvas_category = request.args.get("canvas_category")
|
||||
if request.args.get("desc", "true").lower() == "false":
|
||||
desc = False
|
||||
else:
|
||||
desc = True
|
||||
owner_ids = request.args.get("owner_ids", [])
|
||||
if not owner_ids:
|
||||
tenants = TenantService.get_joined_tenants_by_user_id(current_user.id)
|
||||
tenants = [m["tenant_id"] for m in tenants]
|
||||
canvas, total = UserCanvasService.get_by_tenant_ids(
|
||||
[m["tenant_id"] for m in tenants], current_user.id, page_number,
|
||||
items_per_page, orderby, desc, keywords, canvas_category=CanvasCategory.Agent)
|
||||
return get_json_result(data={"canvas": canvas, "total": total})
|
||||
except Exception as e:
|
||||
return server_error_response(e)
|
||||
tenants, current_user.id, page_number,
|
||||
items_per_page, orderby, desc, keywords, canvas_category)
|
||||
else:
|
||||
tenants = owner_ids
|
||||
canvas, total = UserCanvasService.get_by_tenant_ids(
|
||||
tenants, current_user.id, 0,
|
||||
0, orderby, desc, keywords, canvas_category)
|
||||
return get_json_result(data={"canvas": canvas, "total": total})
|
||||
|
||||
|
||||
@manager.route('/setting', methods=['POST']) # noqa: F821
|
||||
|
||||
@ -182,6 +182,7 @@ def create():
|
||||
"id": get_uuid(),
|
||||
"kb_id": kb.id,
|
||||
"parser_id": kb.parser_id,
|
||||
"pipeline_id": kb.pipeline_id,
|
||||
"parser_config": kb.parser_config,
|
||||
"created_by": current_user.id,
|
||||
"type": FileType.VIRTUAL,
|
||||
@ -546,31 +547,22 @@ def get(doc_id):
|
||||
|
||||
@manager.route("/change_parser", methods=["POST"]) # noqa: F821
|
||||
@login_required
|
||||
@validate_request("doc_id", "parser_id")
|
||||
@validate_request("doc_id")
|
||||
def change_parser():
|
||||
req = request.json
|
||||
|
||||
if not DocumentService.accessible(req["doc_id"], current_user.id):
|
||||
return get_json_result(data=False, message="No authorization.", code=settings.RetCode.AUTHENTICATION_ERROR)
|
||||
try:
|
||||
e, doc = DocumentService.get_by_id(req["doc_id"])
|
||||
if not e:
|
||||
return get_data_error_result(message="Document not found!")
|
||||
if doc.parser_id.lower() == req["parser_id"].lower():
|
||||
if "parser_config" in req:
|
||||
if req["parser_config"] == doc.parser_config:
|
||||
return get_json_result(data=True)
|
||||
else:
|
||||
return get_json_result(data=True)
|
||||
|
||||
if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
|
||||
return get_data_error_result(message="Not supported yet!")
|
||||
e, doc = DocumentService.get_by_id(req["doc_id"])
|
||||
if not e:
|
||||
return get_data_error_result(message="Document not found!")
|
||||
|
||||
def reset_doc():
|
||||
nonlocal doc
|
||||
e = DocumentService.update_by_id(doc.id, {"parser_id": req["parser_id"], "progress": 0, "progress_msg": "", "run": TaskStatus.UNSTART.value})
|
||||
if not e:
|
||||
return get_data_error_result(message="Document not found!")
|
||||
if "parser_config" in req:
|
||||
DocumentService.update_parser_config(doc.id, req["parser_config"])
|
||||
if doc.token_num > 0:
|
||||
e = DocumentService.increment_chunk_num(doc.id, doc.kb_id, doc.token_num * -1, doc.chunk_num * -1, doc.process_duration * -1)
|
||||
if not e:
|
||||
@ -581,6 +573,26 @@ def change_parser():
|
||||
if settings.docStoreConn.indexExist(search.index_name(tenant_id), doc.kb_id):
|
||||
settings.docStoreConn.delete({"doc_id": doc.id}, search.index_name(tenant_id), doc.kb_id)
|
||||
|
||||
try:
|
||||
if req.get("pipeline_id"):
|
||||
if doc.pipeline_id == req["pipeline_id"]:
|
||||
return get_json_result(data=True)
|
||||
DocumentService.update_by_id(doc.id, {"pipeline_id": req["pipeline_id"]})
|
||||
reset_doc()
|
||||
return get_json_result(data=True)
|
||||
|
||||
if doc.parser_id.lower() == req["parser_id"].lower():
|
||||
if "parser_config" in req:
|
||||
if req["parser_config"] == doc.parser_config:
|
||||
return get_json_result(data=True)
|
||||
else:
|
||||
return get_json_result(data=True)
|
||||
|
||||
if (doc.type == FileType.VISUAL and req["parser_id"] != "picture") or (re.search(r"\.(ppt|pptx|pages)$", doc.name) and req["parser_id"] != "presentation"):
|
||||
return get_data_error_result(message="Not supported yet!")
|
||||
if "parser_config" in req:
|
||||
DocumentService.update_parser_config(doc.id, req["parser_config"])
|
||||
reset_doc()
|
||||
return get_json_result(data=True)
|
||||
except Exception as e:
|
||||
return server_error_response(e)
|
||||
|
||||
@ -64,7 +64,7 @@ def create():
|
||||
e, t = TenantService.get_by_id(current_user.id)
|
||||
if not e:
|
||||
return get_data_error_result(message="Tenant not found.")
|
||||
req["embd_id"] = t.embd_id
|
||||
#req["embd_id"] = t.embd_id
|
||||
if not KnowledgebaseService.save(**req):
|
||||
return get_data_error_result()
|
||||
return get_json_result(data={"kb_id": req["id"]})
|
||||
@ -379,3 +379,19 @@ def get_meta():
|
||||
code=settings.RetCode.AUTHENTICATION_ERROR
|
||||
)
|
||||
return get_json_result(data=DocumentService.get_meta_by_kbs(kb_ids))
|
||||
|
||||
|
||||
@manager.route("/basic_info", methods=["GET"]) # noqa: F821
|
||||
@login_required
|
||||
def get_basic_info():
|
||||
kb_id = request.args.get("kb_id", "")
|
||||
if not KnowledgebaseService.accessible(kb_id, current_user.id):
|
||||
return get_json_result(
|
||||
data=False,
|
||||
message='No authorization.',
|
||||
code=settings.RetCode.AUTHENTICATION_ERROR
|
||||
)
|
||||
|
||||
basic_info = DocumentService.knowledgebase_basic_info(kb_id)
|
||||
|
||||
return get_json_result(data=basic_info)
|
||||
|
||||
@ -414,7 +414,7 @@ def agents_completion_openai_compatibility(tenant_id, agent_id):
|
||||
tenant_id,
|
||||
agent_id,
|
||||
question,
|
||||
session_id=req.get("session_id", req.get("id", "") or req.get("metadata", {}).get("id", "")),
|
||||
session_id=req.pop("session_id", req.get("id", "")) or req.get("metadata", {}).get("id", ""),
|
||||
stream=True,
|
||||
**req,
|
||||
),
|
||||
@ -432,7 +432,7 @@ def agents_completion_openai_compatibility(tenant_id, agent_id):
|
||||
tenant_id,
|
||||
agent_id,
|
||||
question,
|
||||
session_id=req.get("session_id", req.get("id", "") or req.get("metadata", {}).get("id", "")),
|
||||
session_id=req.pop("session_id", req.get("id", "")) or req.get("metadata", {}).get("id", ""),
|
||||
stream=False,
|
||||
**req,
|
||||
)
|
||||
|
||||
@ -36,6 +36,8 @@ from rag.utils.storage_factory import STORAGE_IMPL, STORAGE_IMPL_TYPE
|
||||
from timeit import default_timer as timer
|
||||
|
||||
from rag.utils.redis_conn import REDIS_CONN
|
||||
from flask import jsonify
|
||||
from api.utils.health import run_health_checks
|
||||
|
||||
@manager.route("/version", methods=["GET"]) # noqa: F821
|
||||
@login_required
|
||||
@ -169,6 +171,12 @@ def status():
|
||||
return get_json_result(data=res)
|
||||
|
||||
|
||||
@manager.route("/healthz", methods=["GET"]) # noqa: F821
|
||||
def healthz():
|
||||
result, all_ok = run_health_checks()
|
||||
return jsonify(result), (200 if all_ok else 500)
|
||||
|
||||
|
||||
@manager.route("/new_token", methods=["POST"]) # noqa: F821
|
||||
@login_required
|
||||
def new_token():
|
||||
|
||||
Reference in New Issue
Block a user