[英]Running Node.js/Swagger application with pm2 in docker: Error opening Api.yaml
1. Summary : 1.摘要 :
I have created a simple application, which uses Node.js, Express and Swagger, by following this tutorial and using generator-express-no-stress . 通过遵循本教程并使用generator-express-no-stress ,我创建了一个使用Node.js,Express和Swagger的简单应用程序。 In addition the application should be executed with pm2 wihtin a Docker container.
另外,应使用pm2与Docker容器一起执行该应用程序。 Everything works fine until I start the container, which gives an error while trying to read the Api.yaml .
在启动容器之前,一切正常,在尝试读取Api.yaml时出现错误。
2. Steps: 2.步骤:
First, I create the project: 首先,我创建项目:
yo express-no-stress server
cd server
npm i
For the server I change localhost
to 0.0.0.0
( http.createServer(app).listen(port, '0.0.0.0', welcome(port));
), run the server with npm run dev
and do curl 0.0.0.0:3000/api/v1/spec
, which returns - as expected - the specification of the Swagger-file. 对于服务器,我将
localhost
更改为0.0.0.0
( http.createServer(app).listen(port, '0.0.0.0', welcome(port));
),使用npm run dev
运行服务器并执行curl 0.0.0.0:3000/api/v1/spec
,它按预期返回Swagger文件的规范。
Next, I add a configuration file for pm2 to the project, pm2.json : 接下来,我将pm2的配置文件添加到项目pm2.json :
{
"name": "server",
"script": "build/main.js",
"env": {
"APP_ID": "server",
"PORT": "3000",
"LOG_LEVEL": "debug",
"REQUEST_LIMIT": "100kb",
"SESSION_SECRET": "mySecret"
}
}
This time, I use pm2 to start the server: 这次,我使用pm2启动服务器:
$ pm2 start pm2.json
$ curl 0.0.0.0:3000/api/v1/spec
$ pm2 delete all
Again, the specification is returned. 同样,返回规范。
Next, I add a Dockerfile : 接下来,我添加一个Dockerfile :
FROM node:alpine
COPY build build/
COPY package.json .
COPY pm2.json .
RUN npm i pm2 -g \
&& npm i --production \
&& npm cache verify
CMD ["pm2-runtime", "start", "pm2.json"]
I build the image and run the container: 我建立映像并运行容器:
docker build -t server:0.0.1 .
docker run --rm --net=host server:0.0.1
3. Problem and question 3.问题与疑问
And here is my problem: 这是我的问题:
[2018-05-30 09:51:49] PM2 log: Launching in no daemon mode
[2018-05-30 09:51:49] PM2 log: Starting execution sequence in -fork mode- for app name:server id:0
[2018-05-30 09:51:49] PM2 log: App name:server id:0 online
{"level":30,"time":1527673910123,"msg":"up and running in development @: r2d2 on port: 3000}","pid":17,"hostname":"r2d2","name":"server","v":1}
Error: Error opening file "///server/common/swagger/Api.yaml"
ENOENT: no such file or directory, open '///server/common/swagger/Api.yaml'
at ReadFileContext.callback (/node_modules/json-schema-ref-parser/lib/resolvers/file.js:52:20)
at FSReqWrap.readFileAfterOpen [as oncomplete] (fs.js:352:13)
Error: ENOENT: no such file or directory, open '///server/common/swagger/Api.yaml'
I do not know, why the error appears and how to fix it. 我不知道为什么会出现错误以及如何解决。 I have no clue, why the path
///server/common/swagger/Api.yaml
looks a bit strange and I do not even know if this causes the error. 我不知道,为什么路径
///server/common/swagger/Api.yaml
看起来有些奇怪,我什至不知道这是否会导致错误。 Though, the container keeps running. 虽然,容器继续运行。 So: How can I fix it, or any idea where to look at?
所以:我该如何解决它,或者在哪里有任何想法?
4. Further steps 4.进一步的步骤
However, I have done the following steps to examine the error: 但是,我已经按照以下步骤检查了错误:
curl 0.0.0.0:3000/api/v1/spec
from my host machine returns: 我主机上的
curl 0.0.0.0:3000/api/v1/spec
返回:
<h1>500 Error</h1><pre>Error opening file "///server/common/swagger/Api.yaml"
ENOENT: no such file or directory, open '///server/common/swagger/Api.yaml'</pre>
Entering the container with docker exec -it aacfcc6c5744 sh
: 使用
docker exec -it aacfcc6c5744 sh
进入容器:
pm2 show server
┌───────────────────┬────────────────────────────────────┐
| status │ online │
│ name │ server │
│ restarts │ 0 │
│ uptime │ 3m │
│ script path │ /build/main.js │
│ script args │ N/A │
│ error log path │ /root/.pm2/logs/server-error-0.log │
│ out log path │ /root/.pm2/logs/server-out-0.log │
│ pid path │ /root/.pm2/pids/server-0.pid │
│ interpreter │ node │
│ interpreter args │ N/A │
│ script id │ 0 │
│ exec cwd │ / │
│ exec mode │ fork_mode │
│ node.js version │ 10.2.1 │
│ watch & reload │ ✘ │
│ unstable restarts │ 0 │
│ created at │ 2018-05-30T09:51:49.825Z |
└───────────────────┴────────────────────────────────────┘
And: 和:
cd build/ && tree
.
├── main.js
├── main.map
├── public
│ ├── api-explorer
│ │ ├── favicon-16x16.png
│ │ ├── favicon-32x32.png
│ │ ├── index.html
│ │ ├── oauth2-redirect.html
│ │ ├── swagger-ui-bundle.js
│ │ ├── swagger-ui-bundle.js.map
│ │ ├── swagger-ui-standalone-preset.js
│ │ ├── swagger-ui-standalone-preset.js.map
│ │ ├── swagger-ui.css
│ │ ├── swagger-ui.css.map
│ │ ├── swagger-ui.js
│ │ └── swagger-ui.js.map
│ └── index.html
└── server
└── common
└── swagger
└── Api.yaml
Sorry for the long question, but I wanted to make sure that all my steps are hopefully clear and reproducible. 很长的问题,很抱歉,但是我想确保我的所有步骤都希望清楚且可重复。 Let my know, if you need more information.
让我知道,如果您需要更多信息。
5. Edit : 5.编辑 :
The problem has something to do with pm2. 问题与pm2有关。 If I enter the container, change the port of the application to 3001 (since 3000 is already in use) and do
npm start
( cd build && node main
) it works and I can access the specification from my local machine. 如果我进入容器,将应用程序的端口更改为3001(因为已经在使用3000),并执行
npm start
( cd build && node main
),并且可以从本地计算机访问规范。
Problem is solved. 问题解决了。 I just needed to add the current working directory with
cwd
to the configuration file for pm2: 我只需要将带有
cwd
的当前工作目录添加到cwd
的配置文件中:
{
"name": "server",
"script": "build/main.js",
"cwd": "build/",
"env": {
"APP_ID": "server",
"PORT": "3000",
"LOG_LEVEL": "debug",
"REQUEST_LIMIT": "100kb",
"SESSION_SECRET": "mySecret"
}
}
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.