Пример запуска coreos кластера с контейнерами docker, приложением, базой данных и прокси сервером


PS. Исходники с Dockerfile, можно взять здесь:

https://github.com/sysadm-ru/Introduction_To_CoreOS

Они могу понадобиться, если захочется собрать собственные контейнеры или просто посмотреть примеры.


Для запуска примеров нужно:

1) Установить virtualbox
2) Установить vagrant


Vagrantfile и user-data

Скопировать файлы:

https://github.com/sysadm-ru/Native-Docker-Clustering


$ cd ~
$ git clone https://github.com/sysadm-ru/Native-Docker-Clustering
$ cd Native-Docker-Clustering


Сгенерировать ключ:

https://discovery.etcd.io/new?size=7

$ vi user-data

Заменить сгенерированным ключом.

discovery: https://discovery.etcd.io/89e341b6012e47d7e6654eea7b882418


$ vagrant box update


$ vagrant up


// Чтобы можно было по ssh ходить между узлами без пароля

$ ssh-add ~/.vagrant.d/insecure_private_key


$ vagrant status
Current machine states:

core-01                   running (virtualbox)
core-02                   running (virtualbox)
core-03                   running (virtualbox)
core-04                   running (virtualbox)
core-05                   running (virtualbox)
core-06                   running (virtualbox)
core-07                   running (virtualbox)


$ vagrant ssh core-01


$ fleetctl list-machines
MACHINE		IP		METADATA
047ef507...	10.0.11.5	-
104a924a...	10.0.12.5	-
2ccc7711...	10.0.14.5	-
3c89f9a9...	10.0.15.5	-
8df586c8...	10.0.16.5	-
b9048ab8...	10.0.13.5	-


Базы данных


[Unit]
Description=Announce RethinkDB %i service

[Service]
EnvironmentFile=/etc/environment
ExecStart=/bin/sh -c "while true; do etcdctl set /services/rethinkdb/rethinkdb-%i ${COREOS_PUBLIC_IPV4} --ttl 60; sleep 45; done"
ExecStop=/usr/bin/etcdctl rm /services/rethinkdb/rethinkdb-%i

[X-Fleet]
[email protected]*.service



[Unit]
Description=RethinkDB %i service
After=docker.service
[email protected]%i.service

[Service]
EnvironmentFile=/etc/environment
TimeoutStartSec=0
ExecStartPre=-/usr/bin/docker kill rethinkdb-%i
ExecStartPre=-/usr/bin/docker rm rethinkdb-%i
ExecStartPre=-/usr/bin/mkdir -p /home/core/docker-volumes/rethinkdb
ExecStartPre=/usr/bin/docker pull marley/coreos-rethinkdb:latest
ExecStart=/bin/sh -c '/usr/bin/docker run --name rethinkdb-%i   \
 -p ${COREOS_PUBLIC_IPV4}:8080:8080                        \
 -p ${COREOS_PUBLIC_IPV4}:28015:28015                      \
 -p ${COREOS_PUBLIC_IPV4}:29015:29015                      \
 marley/coreos-rethinkdb:latest rethinkdb --bind all \
 --canonical-address ${COREOS_PUBLIC_IPV4}                 \
 $(/usr/bin/etcdctl ls /services/rethinkdb |               \
     xargs -I {} /usr/bin/etcdctl get {} |                 \
     sed s/^/"--join "/ | sed s/$/":29015"/ |              \
     tr "\n" " ")'

ExecStop=/usr/bin/docker stop rethinkdb-%i

[X-Fleet]
[email protected]%i.service


Что возвращается:

$ cat /etc/environment
COREOS_PUBLIC_IPV4=10.0.11.5
COREOS_PRIVATE_IPV4=10.0.11.5


$ echo $(/usr/bin/etcdctl ls /services/rethinkdb |               \
>          xargs -I {} /usr/bin/etcdctl get {} |                 \
>          sed s/^/"--join "/ | sed s/$/":29015"/ |              \
>          tr "\n" " ")
--join 10.0.13.5:29015 --join 10.0.15.5:29015


$ fleetctl submit *

$ fleetctl list-unit-files
UNIT				HASH	DSTATE		STATE		TARGET
[email protected]	3f7611a	inactive	inactive	-
[email protected]		96c6e09	inactive	inactive	-



$ fleetctl list-units
UNIT				MACHINE			ACTIVE	SUB
[email protected]	09e7fca1.../10.0.13.5	active	running
[email protected]	16d2848f.../10.0.15.5	active	running
[email protected]		09e7fca1.../10.0.13.5	active	running
[email protected]		16d2848f.../10.0.15.5	active	running


$ curl 10.0.15.5:8080

Все ок. получил контент от сервера баз данных.


coreos cluster example


$ etcdctl ls --recursive

***

/services
/services/rethinkdb
/services/rethinkdb/rethinkdb-7
/services/rethinkdb/rethinkdb-6


Web Сервера

$ etcdctl get /services/rethinkdb/rethinkdb-6
10.0.13.5

$ etcdctl get /services/rethinkdb/rethinkdb-7
10.0.15.5


$ cd /tmp/
$ git clone --depth=1 https://github.com/sysadm-ru/Introduction_To_CoreOS
$ cd Introduction_To_CoreOS/Chapter5/todo-angular-express/


$ vi config.js

‘172.17.8.101’ меняю на ‘10.0.15.5’


10.0.15.5 - любой coreos хост с etcd, который предоставит информацию о подключении к базе. Разумеется, лучше потом какую-нибудь DNS запись для этого использовать.


$ docker build --rm -t marley/coreos-nodejs-web-app .


$ docker images
REPOSITORY                TAG                 IMAGE ID            CREATED             SIZE
marley/coreos-app         latest              54533dca8c65        8 minutes ago       736.1 MB
marley/coreos-rethinkdb   latest              ee254ccee514        8 weeks ago         181.8 MB
iojs                      2.2                 2a1868f3dfd8        20 months ago       703.8 MB


$ docker login

Ранее в веб интерфейсе создано репо.

$ docker push marley/coreos-nodejs-web-app


$ cd ~



[Unit]
Description=ToDo Service

Requires=docker.service
[email protected]%i.service
After=docker.service

[Service]
EnvironmentFile=/etc/environment
User=core

Restart=always
TimeoutStartSec=0
ExecStartPre=-/usr/bin/docker kill %p-%i
ExecStartPre=-/usr/bin/docker rm %p-%i
ExecStartPre=/usr/bin/docker pull marley/coreos-nodejs-web-app
ExecStart=/usr/bin/docker run --name %p-%i \
      -h %H \
      -p ${COREOS_PUBLIC_IPV4}:3000:3000 \
      -e INSTANCE=%p-%i \
      marley/coreos-nodejs-web-app
ExecStop=-/usr/bin/docker kill %p-%i
ExecStop=-/usr/bin/docker rm %p-%i

[X-Fleet]
[email protected]*.service



[Unit]
Description=ToDo Sidekick
[email protected]%i.service

After=docker.service
[email protected]%i.service
[email protected]%i.service

[Service]
EnvironmentFile=/etc/environment
User=core
Restart=always
TimeoutStartSec=0
ExecStart=/bin/bash -c '\
while true; do \
 port=$(docker inspect --format=\'\' todo-%i); \
 curl -sf ${COREOS_PUBLIC_IPV4}:$port/ > /dev/null 2>&1; \
 if [ $? -eq 0 ]; then \
   etcdctl set /services/todo/todo-%i ${COREOS_PUBLIC_IPV4}:$port --ttl 10; \
 else \
   etcdctl rm /services/todo/todo-%i; \
 fi; \
 sleep 5; \
 done'

ExecStop=/usr/bin/etcdctl rm /services/todo/todo-%i

[X-Fleet]
[email protected]%i.service


Следующая команда должна будет возвращать порт на котором работает вебсервер.

$ docker inspect --format="" todo-4
3000


$ fleetctl submit todo*


$ fleetctl list-unit-files
UNIT				HASH	DSTATE		STATE		TARGET
[email protected]	3f7611a	inactive	inactive	-
[email protected]	3f7611a	launched	launched	09e7fca1.../10.0.13.5
[email protected]	3f7611a	launched	launched	16d2848f.../10.0.15.5
[email protected]		96c6e09	inactive	inactive	-
[email protected]		96c6e09	launched	launched	09e7fca1.../10.0.13.5
[email protected]		96c6e09	launched	launched	16d2848f.../10.0.15.5
[email protected]		64bb9b6	inactive	inactive	-
[email protected]			3dc7e5b	inactive	inactive	-


$ fleetctl start [email protected]{3..5} [email protected]{3..5}


$ fleetctl list-units
UNIT				MACHINE			ACTIVE	SUB
[email protected]	09e7fca1.../10.0.13.5	active	running
[email protected]	16d2848f.../10.0.15.5	active	running
[email protected]		09e7fca1.../10.0.13.5	active	running
[email protected]		16d2848f.../10.0.15.5	active	running
[email protected]		72720a60.../10.0.17.5	active	running
[email protected]		56b7dcad.../10.0.14.5	active	running
[email protected]		b420d775.../10.0.11.5	active	running
[email protected]			72720a60.../10.0.17.5	active	running
[email protected]			56b7dcad.../10.0.14.5	active	running
[email protected]			b420d775.../10.0.11.5	active	running


$ curl 10.0.17.5:3000

Все ок. получил контент приложения от вебсервера.


$ etcdctl ls --recursive

***

/services
/services/rethinkdb
/services/rethinkdb/rethinkdb-6
/services/rethinkdb/rethinkdb-7
/services/todo
/services/todo/todo-3
/services/todo/todo-4
/services/todo/todo-5


coreos cluster example


На самом деле, с первого раза ничего не запустилось

Пришлось искать что это за виртуалка на которой располагается данный сервис.

Номер, виртуалки не совпадал.

// логи

$ fleetctl journal -f --lines=50 [email protected]
$ fleetctl journal -f --lines=50 [email protected]

Пришлось не только перестартовывать, но и удалять конфиги, удалять docker images руками.


$ fleetctl stop [email protected]{3..5} [email protected]{3..5}
$ fleetctl unload [email protected]{3..5} [email protected]{3..5}
$ fleetctl destroy [email protected]{3..5} [email protected]{3..5}
$ fleetctl destroy [email protected]
$ fleetctl destroy [email protected]

И далее повторять все с начала.


Proxy Nginx

$ cd /tmp/Introduction_To_CoreOS/Chapter5/nginx-proxy/


$ vi confd-watch

заменил

export HOST_IP=${HOST_IP:-172.17.8.101}

на

export HOST_IP=${HOST_IP:-10.0.15.5}

10.0.15.5 - любой coreos хост с etcd


$ docker build --rm -t marley/coreos-nginx .


$ docker images
REPOSITORY                TAG                 IMAGE ID            CREATED             SIZE
marley/coreos-nginx       latest              957faace8b64        11 seconds ago      159.3 MB
marley/coreos-app         latest              54533dca8c65        41 minutes ago      736.1 MB
marley/coreos-rethinkdb   latest              ee254ccee514        8 weeks ago         181.8 MB
nginx                     1.9.3               ea4b88a656c9        19 months ago       132.8 MB
iojs                      2.2                 2a1868f3dfd8        20 months ago       703.8 MB


$ docker login

Ранее в веб интерфейсе создано репо.

$ docker push marley/coreos-nginx


$ cd ~


$ vi nginx.service


[Unit]
Description=Nginx Proxy

Requires=docker.service
After=docker.service
After=etcd2.service
Requires=etcd2.service

[Service]
EnvironmentFile=/etc/environment
User=core

Restart=always
TimeoutStartSec=0
ExecStartPre=-/usr/bin/docker kill %p-%i
ExecStartPre=-/usr/bin/docker rm %p-%i
ExecStartPre=-/usr/bin/etcdctl mkdir /services/todo
ExecStartPre=-/usr/bin/docker pull marley/coreos-nginx
ExecStart=/usr/bin/docker run --name %p-%i \
      -h %H \
      -p ${COREOS_PUBLIC_IP}:80:80 \
      marley/coreos-nginx
ExecStop=-/usr/bin/docker kill %p-%i
ExecStop=-/usr/bin/docker rm %p-%i

[X-Fleet]
Global=true


$ fleetctl submit nginx.service


$ fleetctl list-unit-files
UNIT				HASH	DSTATE		STATE		TARGET
nginx.service			111d636	inactive	-		global
[email protected]	3f7611a	inactive	inactive	-
[email protected]	3f7611a	launched	launched	09e7fca1.../10.0.13.5
[email protected]	3f7611a	launched	launched	16d2848f.../10.0.15.5
[email protected]		96c6e09	inactive	inactive	-
[email protected]		96c6e09	launched	launched	09e7fca1.../10.0.13.5
[email protected]		96c6e09	launched	launched	16d2848f.../10.0.15.5
[email protected]		e8b8fa1	inactive	inactive	-
[email protected]		e8b8fa1	launched	launched	b420d775.../10.0.11.5
[email protected]		e8b8fa1	launched	launched	72720a60.../10.0.17.5
[email protected]		e8b8fa1	launched	launched	56b7dcad.../10.0.14.5
[email protected]			b6473ba	inactive	inactive	-
[email protected]			b6473ba	launched	launched	b420d775.../10.0.11.5
[email protected]			b6473ba	launched	launched	72720a60.../10.0.17.5
[email protected]			b6473ba	launched	launched	56b7dcad.../10.0.14.5


$ fleetctl start nginx.service


$ fleetctl list-units
UNIT				MACHINE			ACTIVE	SUB
nginx.service			09e7fca1.../10.0.13.5	active	running
nginx.service			16d2848f.../10.0.15.5	active	running
nginx.service			56b7dcad.../10.0.14.5	active	running
nginx.service			72720a60.../10.0.17.5	active	running
nginx.service			b420d775.../10.0.11.5	active	running
nginx.service			e5b75cfb.../10.0.12.5	active	running
nginx.service			f8083379.../10.0.16.5	active	running
[email protected]	09e7fca1.../10.0.13.5	active	running
[email protected]	16d2848f.../10.0.15.5	active	running
[email protected]		09e7fca1.../10.0.13.5	active	running
[email protected]		16d2848f.../10.0.15.5	active	running
[email protected]		b420d775.../10.0.11.5	active	running
[email protected]		72720a60.../10.0.17.5	active	running
[email protected]		56b7dcad.../10.0.14.5	active	running
[email protected]			b420d775.../10.0.11.5	active	running
[email protected]			72720a60.../10.0.17.5	active	running
[email protected]			56b7dcad.../10.0.14.5	active	running


$ curl 10.0.17.5:80

Ок. Контент от вебсервера через proxy


coreos cluster example