Confd 介绍

CONFD是有一套集中化配置管理系统，一方面提供统一的配置管理，另一方面提供配置变更的自动下发，及时生效。

简介

使用

部署

wget https://github.com/kelseyhightower/confd/releases/download/v0.16.0/confd-0.16.0-linux-amd64

chmod +x confd-0.16.0-linux-amd64
mv confd-0.16.0-linux-amd64 /usr/local/bin/confd

confd --help

confdir

  -confdir string
    	confd conf directory (default "/etc/confd")

• 创建配置文件夹

sudo mkdir -p /etc/confd/{conf.d,templates}

• conf.d: Template Resources，包含：prefix、src、dest、keys、owner、mode、check_cmd、reload_cmd
• templates: Template，参照https://golang.org/pkg/text/template/语法规范

• Template Resources

示例/etc/confd/conf.d/app-tengine.toml：

[template]
prefix = "/"
src = "tengine.tmpl"
dest = "/etc/tengine/conf.d/app_upstreams.conf"
keys = [
  "/apps/upstreams",
]
owner = "root"
mode = "0644"
check_cmd = "tengine -t -c /etc/tengine/tengine.conf"
reload_cmd = "echo reloading tengine && tengine -s reload"

参数说明：

• dest (string) - The target file.
• keys (array of strings) - An array of keys.
• src (string) - The relative path of a configuration template.
• gid (int, optional) - The gid that should own the file. Defaults to the effective gid.
• mode (string, optional) - The permission mode of the file.
• uid (int, optional) - The uid that should own the file. Defaults to the effective uid.
• reload_cmd (string, optional) - The command to reload config.
• check_cmd (string, optional) - The command to check config. Use {{.src}} to reference the rendered * source template.
• prefix (string, optional) - The string to prefix to keys.

• Template

示例/etc/confd/templates/tengine.tmpl：

#for prod env
{{- range gets "/apps/upstreams/*"}}
upstream {{base .Key}} {
  {{ range jsonArray (getv .Key)}} server {{.}} max_fails=1 fail_timeout=30s; {{end}}
}
{{end}}

语法参考：https://github.com/kelseyhightower/confd/blob/master/docs/templates.md

后端存储

  -backend string
    	backend to use (default "etcd")

confd的后端可以是consul、etcd、redis等

• 部署etcd参考：Etcd 单节点安装部署
• 初始化数据

etcdctl put "/apps/upstreams/p.1" '["10.0.0.2:8081", "10.0.0.2:8082"]'
etcdctl put "/apps/upstreams/p.2" '["10.0.0.2:8081", "10.0.0.2:8082"]'

• 查看数据

etcdctl get /apps/upstreams/p.1

启动confd

• 参数

  -node value
    	list of backend nodes

• 命令

# enable watch support
confd -watch -backend etcdv3 -node http://127.0.0.1:2379 & #或采用-interval 5
# run once and exit
confd -onetime -backend etcdv3 -node http://127.0.0.1:2379 &

• 日志

2018-06-09T11:03:29+08:00 a1.xiexianbin.cn confd[15395]: INFO Backend set to etcdv3
2018-06-09T11:03:29+08:00 a1.xiexianbin.cn confd[15395]: INFO Starting confd
2018-06-09T11:03:29+08:00 a1.xiexianbin.cn confd[15395]: INFO Backend source(s) set to http://127.0.0.1:2379
2018-06-09T11:03:29+08:00 a1.xiexianbin.cn confd[15395]: INFO Target config /etc/tengine/conf.d/app_upstreams.conf out of sync
2018-06-09T11:03:29+08:00 a1.xiexianbin.cn confd[15395]: INFO Target config /etc/tengine/conf.d/app_upstreams.conf has been updated
...

• 目标文件

#for prod env
upstream p.1 {
   server 10.0.0.2:8081 max_fails=1 fail_timeout=30s;  server 10.0.0.2:8082 max_fails=1 fail_timeout=30s;
}

upstream p.2 {
   server 10.0.0.2:8081 max_fails=1 fail_timeout=30s;  server 10.0.0.2:8082 max_fails=1 fail_timeout=30s;
}

• 更新

etcdctl put "/apps/upstreams/p.3" '["10.0.0.2:8081", "10.0.0.2:8082"]'

配置文件自动更新

# etcdctl put "/apps/upstreams/p.3" '["10.0.0.2:8081", "10.0.0.2:8082"]'
2018-06-09T14:14:04+08:00 a1.xiexianbin.cn confd[25974]: INFO /etc/tengine/conf.d/app_upstreams.conf has md5sum ebaf0788995ab350200505ea5b31373a should be 4207db050f6cb76d5cb4c59444034af8
2018-06-09T14:14:04+08:00 a1.xiexianbin.cn confd[25974]: INFO Target config /etc/tengine/conf.d/app_upstreams.conf out of sync
OK
2018-06-09T14:14:04+08:00 a1.xiexianbin.cn confd[25974]: INFO Target config /etc/tengine/conf.d/app_upstreams.conf has been updated

# cat /etc/tengine/conf.d/app_upstreams.conf
#for prod env
upstream p.1 {
   server 10.0.0.2:8081 max_fails=1 fail_timeout=30s;  server 10.0.0.2:8082 max_fails=1 fail_timeout=30s;
}

upstream p.2 {
   server 10.0.0.2:8081 max_fails=1 fail_timeout=30s;  server 10.0.0.2:8082 max_fails=1 fail_timeout=30s;
}

upstream p.3 {
   server 10.0.0.2:8081 max_fails=1 fail_timeout=30s;  server 10.0.0.2:8082 max_fails=1 fail_timeout=30s;
}