昨天克隆仓库后直接docker build，首先是用bolt类型的数据库，然后容器启动失败，日志报
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x8 pc=0xa75e18]

goroutine 1 [running]:
github.com/boltdb/bolt.(*Bucket).ForEach(0xc00032a2b8, 0xc00025f038)
/go/pkg/mod/github.com/boltdb/[email protected]/bucket.go:385 +0x18

然后换mongodb，这是我两个容器的启动参数

docker run --name mongo -d -p 27017:27017 mongo
docker run -d -p 80:8001
-v /var/run/docker.sock:/var/run/docker.sock
-e DB_TYPE=mongo
-e DB_ADDRESS=mongodb://192.168.3.70:27017/swirl
--name=swirl
swirl:test-v1

换mongo之后启动成功了，并且Web页面能进入，但是在设置登录名跟密码之后登录失败。


密码设置的123456，登录失败，swirl容器也查不到详细的日志。

您好，想向你请教关于在swirl中 创建表格问题。

我创建的表格显示是空白，如下：

这两个空格的仪表盘表达式参考的是您的这个配置：

我的设置如下（启动普罗米修斯后有重启整个服务）：

普罗米修斯的页面如下：

感谢您耐心的阅读，麻烦您看下这个问题该如何解决

作者你好
我们的企业用户需要自己运维docker，k8s有翻墙问题，对这个工具很感兴趣。
这是你们在生产上使用了吗？

I don't know where to configure service elastic scaling. Can you write a specific exampl?

我是直接采用 docker stack 启动的。

查看日志报错

https://docs.docker.com/compose/compose-file/#init

Hi, it's very cool project. Is it possible to implement that specific user can only see/modify his service/stacks only given permission for that user?

When creating a new user without admin role, information about the password is lost, making it impossible to login.

scanario:
Create a new user by api without admin role.

/api/user/save
{
    "type": "internal",
    "admin": false,
    "name": "Jon Doe",
    "loginName": "test",
    "password": "xxx",
    "passwordConfirm": "xxx",
    "email": "[email protected]",
    "roles": [
        "rolaxxx"
    ]
}

try logging in with your new credentials

api/user/sign-in
{"name":"test","password":"xxx"}

the response will be:

security: invalid principal or credential(1001)

[FIX] This is a fix for this case. The principle is the same as for creating a new admin account.
api/user.go:73

return func(c web.Context) error {
		args := &struct {
			Password string `json:"password"`
			*dao.User
		}{}
		err := c.Bind(args, true)
		if err == nil {
			ctx, cancel := misc.Context(defaultTimeout)
			defer cancel()

			user := args.User
			if user.ID == "" {
				user.Password = args.Password
				_, err = b.Create(ctx, user, c.User())
			} else {
				err = b.Update(ctx, user, c.User())
			}
		}
		return ajax(c, err)
}

Hi! I'm trying to build this locally to hack on / contribute to but I think there's a checksum error with the docker client dependency:

prologic@Jamess-MacBook
Sat Apr 13 14:06:42
~/Contributions/swirl
 (master) 0
$ go build
go: downloading github.com/docker/docker v0.0.0-20180415161252-ea579a619452
go: verifying github.com/docker/[email protected]: checksum mismatch
	downloaded: h1:KXTxrc9Q8FwG8x/lbSBYzBMKn3xWpwx9MBfP1Ko7FYg=
	go.sum:     h1:3DHHYhTKAP1ZgL0g+msknSuFuZBA0UW6BqcAA5qsWWc=

Can you look in to this? Perhaps consider upgrading this dependency (not trivial as the library has changed some API)

Hi,are you consider refactor this project to separate frontend and backend?login with jwt?
i can help write the ajax backend

Any opened Node detail page show tasks not related to chosen node.

服务已经移除了 但是任务还在 处于pending状态 而且点击服务是不存在的

现在节点有点多
需要进入非swirl节点的终端还是需要到链接指定机器上 比较麻烦

@cuigh 主要是用于直接操作swirl的rest接口直接重启服务或者上下线编排

I would love to run swirl on my raspberry pi's

I am a little confused by the instructions and have been unable to get this working. I am unclear if i have hit an issue, didn't read something i should have (like manually creating databases and users)?

I am trying to deploy in a swarm using mongodb, i am unclear if mongodb is in the docker image or not?
I have tried both of the following:

• assuming mongodb is in the image
• assuming i need separate mongodb

(note i have never used mongodb before)
for approach #1 above

panic: failed to load storage engine: no reachable servers

as such i believe there to be no mondodb server running in the container

for approach #2 above mycompose is further below, when swirl tries to connect to the mongodb it results in this if the db path ends in /swirl

panic: failed to load storage engine: server returned error on SASL authentication step: Authentication failed.

if it doesn't end in /swirl i get

panic: failed to load storage engine: server returned error on SASL authentication step: BSON field 'saslContinue.mechanism' is an unknown field.

compose file:

version: '3.8'

services:
  swirl:
    image: cuigh/swirl
    environment:
      DB_TYPE: mongo
      DB_ADDRESS: mongodb://mongoroot:password@mongodb:27017/swirl
      DOCKER_ENDPOINT: tcp://swirl_manager_agent:2375
      AGENTS: swirl_manager_agent,swirl_worker_agent
    ports:
      - 8001:8001
    deploy:
      replicas: 1
      placement:
        constraints: [ node.role == worker ]

  manager_agent:
    image: cuigh/socat
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
    deploy:
      mode: global
      placement:
        constraints: [ node.role == manager ]

  worker_agent:
    image: cuigh/socat
    volumes:
      - /var/run/docker.sock:/var/run/docker.sock
    deploy:
      mode: global
      placement:
        constraints: [ node.role == worker ]
        
  mongodb:
    image: mongo
    restart: always
    environment:
      MONGO_INITDB_ROOT_USERNAME: mongoroot
      MONGO_INITDB_ROOT_PASSWORD: password

  mongo-express:
    image: mongo-express
    restart: always
    ports:
      - 7081:8081
    environment:
      ME_CONFIG_MONGODB_ADMINUSERNAME: mongoroot
      ME_CONFIG_MONGODB_ADMINPASSWORD: password
      ME_CONFIG_MONGODB_URL: mongodb://mongoroot:password@mongodb:27017/      

note my mongo-express service connects with no issues to the database.

您好，我的swarm测试集群是1manager+2worker，
因为有--mount="type=bind,src=/var/run/docker.sock,dst=/var/run/docker.sock" ，我理解的是创建swirl服务必须在manager上执行。
想请教您一下 --constraint 'node.role == worker' 是怎么实现的？

就是指定规则在某几个机器上运行的那个参数

Hi,

I am getting the following message in my browser after I login to Swirl: "Cannot connect to the Docker daemon at unix:///var/run/docker.sock. Is the docker daemon running?"

Could provide my with advise or direction what I am missing or doing wrong?

Below is my docker-compose file section for starting Swirl and MongoDB:

` swirl:
image: cuigh/swirl
container_name: swirl
depends_on:
- mongo
ports:
- 8001:8001
environment:
# - DB_TYPE=bolt
- PGID=998
- DB_ADDRESS=/data/swirl
- DB_ADDRESS=mongo:27017/swirl # - DB_ADDRESS=/data/swirl
# - RESCRAMBLE=true
security_opt:
- apparmor:unconfined
- seccomp:unconfined
volumes:
- /var/run/docker.sock,dst=/var/run/docker.sock
- /var/lib/docker/:/var/lib/docker:ro
- /path/swirl/data=/data/swirl
networks:
- web
- net
deploy:
# privileged: true
mode: replicated
replicas: 1
update_config:
order: start-first # get new container running first before removing old, for less downtime
rollback_config:
order: start-first # get new container running first before removing old, for less downtime
placement:
constraints:
- node.role==manager

mongo:
image: mongo
volumes:
- /path/mongo/data:/data/db
networks:
- net
deploy:
mode: replicated
replicas: 1
update_config:
order: start-first # get new container running first before removing old, for less downtime
rollback_config:
order: start-first # get new container running first before removing old, for less downtime

placement:

constraints: [node.hostname == mongo]`

Thank you in advance.
Regards,
Angelo

服务过多的情况下排版比较奇怪
右侧的三个可以不用换行
左侧的 多的换行就行了 而不是全部换行

1. 日志的实时展示
2. 多服务日志展示
3. 服务内部命令执行

最近正好在学习Go相关的内容 有时间的话我自己也会试着写写看 提交PR

Whenever I want to edit the stack file (type input) this error pops up.
I'm using version v1.0.0rc2.

就是在服务日志页面 可以看到实时日志 而不是手动刷新
类似Jenkins的构建日志那样的

Hello,

I am doing a swarm deployment of Swirl. My swirl tasks get rejected with this message: invalid mount config for type "bind": bind source path does not exist: /data/swirl using the following compose.yml.

version: '3.3'
services:
  manager_agent:
    image: cuigh/socat:latest
    volumes:
     - /var/run/docker.sock:/var/run/docker.sock
    networks:
     - default
    logging:
      driver: json-file
    deploy:
      mode: global
      placement:
        constraints:
         - node.role == manager
  mongo-express:
    image: mongo-express:latest
    environment:
      ME_CONFIG_MONGODB_URL: mongodb://mongodb:27017/
    ports:
     - 7081:8081
    networks:
     - default
    logging:
      driver: json-file
  mongodb:
    image: mongo:latest
    networks:
     - default
    logging:
      driver: json-file
  swirl:
    image: cuigh/swirl:latest
    environment:
      AGENTS: swirl_manager_agent,swirl_worker_agent
      DB_ADDRESS: mongodb://mongodb:27017/swirl
      DB_TYPE: mongo
      DOCKER_ENDPOINT: tcp://swirl_manager_agent:2375
    ports:
     - 8001:8001
    networks:
     - default
    logging:
      driver: json-file
    deploy:
      placement:
        constraints:
         - node.role == worker
  worker_agent:
    image: cuigh/socat:latest
    volumes:
     - /var/run/docker.sock:/var/run/docker.sock
    networks:
     - default
    logging:
      driver: json-file
    deploy:
      mode: global
      placement:
        constraints:
         - node.role == worker
networks:
  default:
    driver: overlay

I also noticed there are no clear instructions on how to supply database credentials, am I going to be prompted for those in the UI once I get it running?

Thank you very much.

Query expression怎么写？

我觉得楼主直接定好模版就行了。

Login less than 1 minute, prompt 401

你好，不好意思还要麻烦一下，我按照readme中的所说的service label的方式配置了一个测试nginx，stack是这样的：

`version: '3.6'
services:

nginx:
image: nginx:latest
environment:
- TZ=Asia/Shanghai
deploy:
labels:
swirl.scale: swirl.scale=min=1,max=5,cpu=20:50
restart_policy:
condition: on-failure
resources:
limits:
cpus: "0.01"
memory: 200M
update_config:
parallelism: 1 # 每次更新1个副本
delay: 5s # 每次更新间隔
monitor: 10s # 单次更新多长时间后没有结束则判定更新失败
max_failure_ratio: 0.1 # 更新时能容忍的最大失败率
order: start-first # 更新顺序为新任务启动优先

ports:
- 18081:80`

但是在后面的压测中，一直没有看到这个service进行scale的动作，请问是哪里配置有问题吗？谢谢！

1、想要一些系统级的监控
2、能不能适配多个docker swarm集群管理

在HTTPS的情况下 WebSocket没有使用wss导致无法链接 希望修复一下

I notice there is a way to import dashboards. Can we provide/supply some samples or defaults baked in?

现在手动修改的方式太麻烦了
如果能通过Git更新就比较方便 管理也比较统一

Hello,

Did you think it's possible to extract the Scaler part as an autonomous application ??
The idea is brilliant.

Best regards

Some of us run reverse proxies with middleware that take care of authentication.
It would be nice if we could disable authentication or have a way to bypass the login.
I think LDAP would be kind of overkill for my setup to provide a somewhat unified login experience so it would be nice if I could disable the login functionality.

(maybe an option to disable the expiration of the session cookie so I can just pass it along as a request header to the host?)

I was wondering if anyone ever tried using swirl behind traefik, as I have not found a way to make it work with a base path other than / as the ui has all the references pointing to that.

This is the traefik config I'm trying to use:

    deploy:
      labels:
        - "traefik.enable=true"
        - "traefik.http.routers.swirl.entrypoints=https"
        - "traefik.http.routers.swirl.tls=true"
        - "traefik.http.services.swirl.loadbalancer.server.port=8001"
        - "traefik.http.routers.swirl.rule=PathPrefix(`/swirl`)"
        - "traefik.http.routers.swirl.middlewares=swirl-stripprefix"
        - "traefik.http.middlewares.swirl-stripprefix.stripprefix.prefixes=/swirl"
        - "traefik.http.middlewares.swirl-stripprefix.stripprefix.forceslash=true"