分布式架构实战–FastDFS分布式文件系统集群安装-演道网

2017 年 2 月 10 日

样例项目实战参考：http://www.roncoo.com/course/view/85d6008fe77c4199b0cdd2885eaeee53 46-48节

文档修订记录

face/tTKsrKXG6QAYwd6P6pZahWRisX4sRrG5.jpg

跟踪服务器：

Tracker Server1 IP: 10.10.16.201 FASTDFS-TRACKER-01

Tracker Server2 IP: 10.10.16.202 FASTDFS-TRACKER-02

存储服务器：

Storage Server1 IP: 10.10.16.203 FASTDFS-STORAGE-01

Storage Server2 IP: 10.10.16.204 FASTDFS-STORAGE-02

集群中各操作系统环境设置：

(1) fastdfs安装目录：/usr/local/fastdfs

[root@FASTDFS-TRACKER-01 local]# mkdir /usr/local/fastdfs

[root@FASTDFS-TRACKER-02 local]# mkdir /usr/local/fastdfs

[root@FASTDFS-STORAGE-01 local]# mkdir /usr/local/fastdfs

[root@FASTDFS-STORAGE-02 local]# mkdir /usr/local/fastdfs

(2)创建fastdfs用户组及fastdfs用户（四台机同时配置）:

[root@xxx]# groupadd fastdfs

[root@xxx]# useradd -g fastdfs fastdfs

[root@xxx]# passwd fastdfs

## 设置fastdfs用户密码为gzzyzz.com （上生产前要改）

(3) 创建数据存储目录:

#创建tracker目录保存运行日志(Tracker服务器)

[fastdfs@FASTDFS-TRACKER-01 ~]$ mkdir -p /home/fastdfs/tracker

[fastdfs@FASTDFS-TRACKER-02 ~]$ mkdir -p /home/fastdfs/tracker

#创建Storage目录保存运行日志及其data数据（Storage服务器）

[fastdfs@FASTDFS-STORAGE-01 ~] $ mkdir -p /home/fastdfs/storage

[fastdfs@FASTDFS-STORAGE-02 ~] $ mkdir -p /home/fastdfs/storage

注：因为要源码编译安装软件，因而各系统要事先安装好:make cmake gcc gcc-c++

[root@FASTDFS-TRACKER-01 yum.repos.d]# yum install make cmake gcc gcc-c++

1、首先安装libevent（集群中所有服务器都执相同的安装）:

fastdfs在编译源程序时fastdfs内部调用libevent的处理机制,需要用到libevent一些依赖文件,否则编译fastdfs会出错。

# 卸载系统自带libevent,自带版本过低,要不然安装fastdfs会出错

[root@FASTDFS-TRACKER-01 fastdfs]# rpm -qa|grep libevent

libevent-1.4.13-4.el6.x86_64

（由上可以系统自带了libevent-1.4.13-4.el6.x86_64）

卸载：

[root@FASTDFS-TRACKER-01 fastdfs]# rpm -e –nodeps libevent-1.4.13-4.el6.x86_64

#下载安装libevent

进入/usr/local/src目录

# cd /usr/local/src/

[root@FASTDFS-TRACKER-01 src]#

wget http://cloud.github.com/downloads/libevent/libevent/libevent-2.0.19-stable.tar.gz

[root@FASTDFS-TRACKER-01 src]# tar -zxvf libevent-2.0.19-stable.tar.gz

[root@FASTDFS-TRACKER-01 src]# cd libevent-2.0.19-stable

#make clean;

[root@FASTDFS-TRACKER-01 libevent-2.0.19-stable]$ ./configure –prefix=/usr/local/libevent

[root@FASTDFS-TRACKER-01 libevent-2.0.19-stable]# make

[root@FASTDFS-TRACKER-01 libevent-2.0.19-stable]# make && make install

#为libevent创建软链接到/lib库下，64位系统对应/lib64

[root@FASTDFS-TRACKER-01 libevent-2.0.19-stable]#

ln -s /usr/local/libevent/lib/libevent-2.0.so.5 /usr/lib/libevent-2.0.so.5

[root@FASTDFS-TRACKER-01 libevent-2.0.19-stable]#

ln -s /usr/local/libevent/lib/libevent-2.0.so.5 /usr/lib64/libevent-2.0.so.5

2、安装FastDFS（集群中所有服务器都执相同的安装）:

进入/usr/local/src目录

[root@FASTDFS-TRACKER-01 src]#

wget https://fastdfs.googlecode.com/files/FastDFS_v4.06.tar.gz

[root@FASTDFS-TRACKER-01 src]# tar -zxvf FastDFS_v4.06.tar.gz

[root@FASTDFS-TRACKER-01 src]# cd FastDFS

#由于定义/usr/local/fastdfs为fastdfs安装目录,所以需要修改make.sh

[gw@register2 FastDFS]$ vi make.sh

# /etc/fdfs 全部替换为 /usr/local/fastdfs/conf

# TARGET_PREFIX=/usr/local 修改为 /usr/local/fastdfs

# TARGET_CONF_PATH=/etc/fdfs 修改为 /usr/local/fastdfs/conf

附件：

#安装

[root@FASTDFS-TRACKER-01 FastDFS]$

./make.sh C_INCLUDE_PATH=/usr/local/libevent/include LIBRARY_PATH=/usr/local/libevent/lib

# 切换到超级管理员

[root@FASTDFS-TRACKER-01 FastDFS]# ./make.sh install

注意：如果安装时提示找不到命令，请查看是不是没有装perl安装环境

wget http://www.cpan.org/src/5.0/perl-5.18.2.tar.gz

tar -zxvf perl-5.18.2.tar.gz

cd perl-5.18.2

./Configure

make

make install

3、配置Tracker Server（10.10.16.201，10.10.16.202）:

进入/usr/local/fastdfs/conf

#修改tracker.conf配置

[root@FASTDFS-TRACKER-01 conf]# vi /usr/local/fastdfs/conf/tracker.conf

绑定IP:

bind_addr=10.10.16.201 和 bind_addr=10.10.16.202

#设置日志目录

由base_path=/home/yuqing/fastdfs 改为 base_path=/home/fastdfs/tracker

store_group=group1

run_by_group= 改为 run_by_group=fastdfs

run_by_user= 改为 run_by_user=fastdfs

trunk_create_file_space_threshold = 20G

#开启自trunk_create_file_space_threshold定义server ID取代ip形式,方便内部网络服务器更换ip#**此方式要重点理解，4.0以后新特性

use_storage_id = true #使用server ID作为storage server标识

storage_ids_filename = storage_ids.conf #

id_type_in_filename = id #文件名反解析中包含server ID,以前是ip

复制storage_ids.conf文件

[root@FASTDFS-TRACKER-01 conf]#

cp /usr/local/src/FastDFS/conf/storage_ids.conf /usr/local/fastdfs/conf/

#编辑storage服务器ID与IP地址的对应关系

[root@register1 conf]# vi /usr/local/fastdfs/conf/storage_ids.conf

100001 group1 10.10.16.203

100002 group1 10.10.16.204

防火墙打开8080和22122端口：

# vi /etc/sysconfig/iptables

增加：

-A INPUT -m state –state NEW -m tcp -p tcp –dport 8080 -j ACCEPT

-A INPUT -m state –state NEW -m tcp -p tcp –dport 22122 -j ACCEPT

重启防火墙:

# /etc/init.d/iptables restart

自定义tracker服务启动、关闭、重启脚本：

# su fastdfs

切换回fastdfs用户后

$ cd /home/fastdfs/tracker/

(1)启动脚本：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ vi start_tracker.sh

/usr/local/fastdfs/bin/fdfs_trackerd /usr/local/fastdfs/conf/tracker.conf

(2)关闭脚本（使用FastDFS自带的stop.sh脚本）：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ vi stop_tracker.sh

/usr/local/fastdfs/bin/stop.sh /usr/local/fastdfs/bin/fdfs_trackerd /usr/local/fastdfs/conf/tracker.conf

(3)重启脚本（使用FastDFS自带的restart.sh脚本）：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ vi restart_tracker.sh

/usr/local/fastdfs/bin/restart.sh /usr/local/fastdfs/bin/fdfs_trackerd /usr/local/fastdfs/conf/tracker.conf

注意：千万不要使用kill -9参数强杀进程，否则可能会导致binlog数据丢失的问题

给自定义脚本赋予可执行权限：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ chmod +x *_tracker.sh

启动tracker服务：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ ./start_tracker.sh

查看是否启用成功：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ ps -ef | grep fastdfs

root 27253 26511 0 10:03 pts/0 00:00:00 su fastdfs

fastdfs 27254 27253 0 10:03 pts/0 00:00:00 bash

fastdfs 27580 1 0 10:25 ? 00:00:00 /usr/local/fastdfs/bin/fdfs_trackerd /usr/local/fastdfs/conf/tracker.conf

fastdfs 27588 27254 5 10:25 pts/0 00:00:00 ps -ef

fastdfs 27589 27254 0 10:25 pts/0 00:00:00 grep fastdfs

[fastdfs@FASTDFS-TRACKER-01 tracker]$

（首次正常启动后会在/home/fastdfs/tracker目录下会产生data目录和logs目录）

[2014-01-12 16:33:54] INFO – local_host_ip_count: 2, 127.0.0.1 10.10.16.201

[2014-01-12 16:34:11] INFO – file: tracker_service.c, line: 920, the tracker leader is 10.10.16.202:22122

停用tracker服务：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ ./stop_tracker.sh

重启tracker服务：

[fastdfs@FASTDFS-TRACKER-01 tracker]$ ./restart_tracker.sh

4、配置Storage Server(10.10.16.203、10.10.16.204，两台机执行相同的配置操作)：

# cd /usr/local/fastdfs/conf

# vi storage.conf

绑定IP分别为:

bind_addr=10.10.16.203 和 bind_addr=10.10.16.204

base_path=/home/yuqing/fastdfs 改为 base_path=/home/fastdfs/storage

store_path0=/home/yuqing/fastdfs 改为 store_path0=/home/fastdfs/storage

tracker_server=192.168.209.121:22122 改为

tracker_server=10.10.16.201:22122

tracker_server=10.10.16.202:22122

（这里我们配了两个跟踪器服务）

run_by_group= 改为 run_by_group=fastdfs

run_by_user= 改为 run_by_user=fastdfs

# 与Nginx端口相同

http.server_port=80

防火墙打开23000和8888端口：

# vi /etc/sysconfig/iptables

增加：

-A INPUT -m state –state NEW -m tcp -p tcp –dport 23000 -j ACCEPT

-A INPUT -m state –state NEW -m tcp -p tcp –dport 8888 -j ACCEPT

重启防火墙:

# /etc/init.d/iptables restart

自定义storage服务启动、关闭、重启脚本：

# su fastdfs

切换回fastdfs用户后

$ cd /home/fastdfs/storage/

(1)启动脚本：

[fastdfs@FASTDFS-STORAGE-01 storage]$ vi start_storage.sh

/usr/local/fastdfs/bin/fdfs_storaged /usr/local/fastdfs/conf/storage.conf

(2)关闭脚本（使用FastDFS自带的stop.sh脚本）：

[fastdfs@FASTDFS-STORAGE-01 storage]$ vi stop_tracker.sh

/usr/local/fastdfs/bin/stop.sh /usr/local/fastdfs/bin/fdfs_storaged /usr/local/fastdfs/conf/storage.conf

(3)重启脚本（使用FastDFS自带的restart.sh脚本）：

[fastdfs@FASTDFS-STORAGE-01 storage]$ vi restart_storage.sh

/usr/local/fastdfs/bin/restart.sh /usr/local/fastdfs/bin/fdfs_storaged /usr/local/fastdfs/conf/storage.conf

注意：千万不要使用kill -9参数强杀进程，否则可能会导致binlog数据丢失的问题

给自定义脚本赋予可执行权限：

[fastdfs@FASTDFS-STORAGE-01 storage]$ chmod +x *_storage.sh

启动Storage Server:

[fastdfs@FASTDFS-STORAGE-01 storage]$ ./start_storage.sh

data path: /home/fastdfs/storage/data, mkdir sub dir…

mkdir data path: 00 …

mkdir data path: 01 …

mkdir data path: 02 …

mkdir data path: 03 …

（首次成功启动会初始化数据存储目录）

重启storage服务：

[fastdfs@FASTDFS-STORAGE-01 storage]$ ./restart_storage.sh

停止storage服务：

[fastdfs@FASTDFS-STORAGE-01 storage]$ ./stop_storage.sh

Storage Server（10.10.16.203和10.10.16.204）安装Nginx:

#安装Nginx使用fastdfs用户

#创建nginx日志目录

# mkdir -p /home/fastdfs/nginx/logs

# chmod a+w /home/fastdfs/nginx/logs

# chown -R fastdfs:fastdfs /home/fastdfs/nginx/logs

安装nginx1.4.4:

# cd /usr/local/src

#下载 nginx

[root@FASTDFS-STORAGE-01 src]#

wget http://nginx.org/download/nginx-1.4.4.tar.gz

解压：

# tar -zxvf nginx-1.4.4.tar.gz

# cd nginx-1.4.4

# ./configure –user=fastdfs –group=fastdfs –prefix=/usr/local/nginx –with-http_stub_status_module

如果产生如下错误提示：

checking for PCRE library … not found

checking for PCRE library in /usr/local/ … not found

checking for PCRE library in /usr/include/pcre/ … not found

checking for PCRE library in /usr/pkg/ … not found

checking for PCRE library in /opt/local/ … not found

./configure: error: the HTTP rewrite module requires the PCRE library.

You can either disable the module by using –without-http_rewrite_module

option, or install the PCRE library into the system, or build the PCRE library

statically from the source with nginx by using –with-pcre=

（大概意思是：HTTP重定向模块的安装需要PCRE库，网上搜了一下，是需要安装pcre-devel与openssl-devel）

# yum install pcre-devel openssl openssl-devel

（RHEL6.3 x64 桌面版中已自带以上包）

再重新执行：

# ./configure –user=fastdfs –group=fastdfs –prefix=/usr/local/nginx –with-http_stub_status_module

#–with-http_stub_status_module 用来监控nginx的当前状态

Configuration summary

+ using system PCRE library

+ OpenSSL library is not used

+ md5: using system crypto library

+ sha1: using system crypto library

+ using system zlib library

nginx path prefix: “/usr/local/nginx”

nginx binary file: “/usr/local/nginx/sbin/nginx”

nginx configuration prefix: “/usr/local/nginx/conf”

nginx configuration file: “/usr/local/nginx/conf/nginx.conf”

nginx pid file: “/usr/local/nginx/logs/nginx.pid”

nginx error log file: “/usr/local/nginx/logs/error.log”

nginx http access log file: “/usr/local/nginx/logs/access.log”

nginx http client request body temporary files: “client_body_temp”

nginx http proxy temporary files: “proxy_temp”

nginx http fastcgi temporary files: “fastcgi_temp”

nginx http uwsgi temporary files: “uwsgi_temp”

nginx http scgi temporary files: “scgi_temp”

执行安装：# make && make install

安装fastdfs-nginx-module插件：

# cd /usr/local/src

# wget http://fastdfs.googlecode.com/files/fastdfs-nginx-module_v1.15.tar.gz

# tar -zxvf fastdfs-nginx-module_v1.15.tar.gz

#修改插件配置文件

# vi /usr/local/src/fastdfs-nginx-module/src/config

ngx_addon_name=ngx_http_fastdfs_module

HTTP_MODULES=”$HTTP_MODULES ngx_http_fastdfs_module”

NGX_ADDON_SRCS=”$NGX_ADDON_SRCS $ngx_addon_dir/ngx_http_fastdfs_module.c”

CORE_INCS=”$CORE_INCS /usr/local/fastdfs/include/fastdfs /usr/local/fastdfs/include/fastcommon/“

CORE_LIBS=”$CORE_LIBS -L/usr/local/fastdfs/lib -lfastcommon -lfdfsclient“

CFLAGS=”$CFLAGS -D_FILE_OFFSET_BITS=64 -DFDFS_OUTPUT_CHUNK_SIZE=’256*1024′ -DFDFS_MOD_CONF_FILENAME='”/usr/local/fastdfs/conf/mod_fastdfs.conf”‘“

#复制mod_fastdfs.conf到/usr/local/fastdfs/conf/目录下

# cp /usr/local/src/fastdfs-nginx-module/src/mod_fastdfs.conf /usr/local/fastdfs/conf/

上面编译时使用的动态链接库：

#启动nginx报错

#./nginx: error while loading shared libraries: libfastcommon.so: cannot open shared object file: No such file or directory

#解决办法 —> 将/usr/local/fastdfs/lib 加入系统文件/etc/ld.so.conf中

# vi /etc/ld.so.conf

include ld.so.conf.d/*.conf

/usr/local/fastdfs/lib

#更新库文件缓存ld.so.cache

# /sbin/ldconfig -v

编译fastdfs-nginx-module模块：

# 重新编译安装nginx

# cd /usr/local/src/nginx-1.4.4

# ./configure –add-module=/usr/local/src/fastdfs-nginx-module/src

# make && make install

修改mod_fastdfs.conf配置:

# vi /usr/local/fastdfs/conf/mod_fastdfs.conf

# connect timeout in seconds

# default value is 30s

connect_timeout=30

# the base path to store log files

base_path=/tmp

# if load FastDFS parameters from tracker server

# since V1.12

# default value is false

load_fdfs_parameters_from_tracker=true

# if use storage ID instead of IP address

# same as tracker.conf

# valid only when load_fdfs_parameters_from_tracker is false

# default value is false

# since V1.13

use_storage_id = true

# FastDFS tracker_server can ocur more than once, and tracker_server format is

# “host:port”, host can be hostname or ip address

# valid only when load_fdfs_parameters_from_tracker is true

tracker_server=10.10.16.201:22122

tracker_server=10.10.16.202:22122

# the port of the local storage server

# the default value is 23000

storage_server_port=23000

# the group name of the local storage server

group_name=group1

# if the url / uri including the group name

# set to false when uri like /M00/00/00/xxx

# set to true when uri like ${group_name}/M00/00/00/xxx, such as group1/M00/xxx

# default value is false

url_have_group_name = true

# path(disk or mount point) count, default value is 1

# must same as storage.conf

store_path_count=1

# store_path#, based 0, if store_path0 not exists, it’s value is base_path

# the paths must be exist

# must same as storage.conf

store_path0=/home/fastdfs/storage

# set the log filename, such as /usr/local/apache2/logs/mod_fastdfs.log

# empty for output to stderr (apache and nginx error_log file)

log_filename=/home/fastdfs/nginx/logs/mod_fastdfs.log

注意, 下载时如发现老报404. 将nginx.conf第一行 usr nobody 修改为 user root 重新启动后解决.

给 storage 的存储目录做一个软连接

# ln -s /home/fastdfs/storage/data /home/fastdfs/storage/data/M00

访问路径带group名（storage对应有多个group的情况），如/group1/M00/00/00/xxx:
   location ~ /group([0-9])/M00 {
         ngx_fastdfs_module;
      }

nginx配置简洁版本:

# vi /usr/local/nginx/conf/nginx.conf

user root root;

worker_processes 2;

error_log /home/fastdfs/nginx/logs/error.log notice;

pid /home/fastdfs/nginx/logs/nginx.pid;

worker_rlimit_nofile 5120;

events {

use epoll;

worker_connections 5120;

}

http {

include mime.types;

default_type application/octet-stream;

sendfile on;

tcp_nopush on;

keepalive_timeout 65;

tcp_nodelay on;

server {

listen 80;

server_name localhost;

#charset koi8-r;

location ~/group([0-9])/M00 {

#alias /home/fastdfs/storage/data;

ngx_fastdfs_module;

}

#启动nginx：

# /usr/local/nginx/sbin/nginx

#重启nginx：

# /usr/local/nginx/sbin/nginx -s reload

用Java客户端上传一个文件：

浏览

face/ASNGBn74BGS5aypebfs2y4x7ydRzHRST.png

face/yAeQkkjTrfs2rDjP6WTc6rFi5pynY26F.png

文件的访问通过硬件做均衡负载

VIP:10.10.16.209 (做负载均衡的虚拟机网关要设为10.10.16.8)

face/Z4HRRkipn3GsaXwTH4T2aP2i8FHdZiDy.png

5、 a

6、 a

7、 a

8、 a

9、 a

关注微信：

face/JQ8mbPNjdrZyCKGxNhNfcQnBcnpYs6Fd.png

转载自演道,想查看更及时的互联网产品技术热点文章请点击http://go2live.cn

About The Author

bjmayor

程序员，码农，php,python,ios,android,go，产品经理，创业。

2025年六月
M	T	W	T	F	S	S
« Jan
						1
2	3	4	5	6	7	8
9	10	11	12	13	14	15
16	17	18	19	20	21	22
23	24	25	26	27	28	29
30

Related Posts

About The Author

bjmayor