Linux下FastDFS分布式存储-总结及部署记录
一、分布式文件系统介绍
分布式文件系统:Distributed file system, DFS,又叫做网络文件系统:Network File System。一种允许文件通过网络在多台主机上分享的文件系统,可让多机器上的多用户分享文件和存储空间。
特点:在一个分享的磁盘文件系统中,所有节点对数据存储区块都有相同的访问权,在这样的系统中,访问权限就必须由客户端程序来控制。分布式文件系统可能包含的功能有:透通的数据复制与容错。
分布式文件系统是被设计用在局域网。而分布式数据存储,则是泛指应用分布式运算技术的文件和数据库等提供数据存储服务的系统。
决定因素:数据的存储方式、数据的读取速率、数据的安全机制。
发展历史:大致分为三个发展阶段,网络文件系统(1980s)、共享SAN文件系统(1990s)、面向对象的并行文件系统(2000s)。
二、FastDFS分布式系统架构介绍
FastDFS:是一个开源的轻量级分布式文件系统,功能包括:文件存储、文件同步、文件访问(文件上传、文件下载)等,解决了大容量存储和负载均衡的问题。特别适合中小文件(建议范围:4KB < file_size <500MB),对以文件为载体的在线服务,如相册网站、视频网站等。
FastDFS是为互联网应用量身定做的分布式文件系统,充分考虑了冗余备份、负载均衡、线性扩容等机制,并注重高可用、高性能等指标。和现有的类Google FS分布式文件系统相比,FastDFS的架构和设计理念有其独到之处,主要体现在轻量级、分组方式和对等结构三个方面。
FastDFS架构图
FastDFS服务端有两个角色:跟踪器(tracker)和存储节点(storage)。跟踪器主要做调度工作,在访问上起负载均衡的作用。
FastDFS模块介绍
1)tracker server:跟踪服务器,用来调度来自客户端的请求,且在内存中记录所有存储组和存储服务器的信息状态。
2)storage server:存储服务器,用来存储文件(data)和文件属性(metadata)。
3)client:客户端,业务请求发起方,通过专用接口基于TCP协议与tracker server和storage server进行交互。
4)group:组,也可称为卷,同组内上的文件是完全相同的。
5)文件标识:包括两部分,组名(group)和文件名(含路径)
6)文件相关属性:键值对(Key Value Pair)方式
7)文件名:与原文件名并不相同。由storage server根据特定信息生成,并且可逆,文件名包含:源存储服务器的IP地址、文件创建时间戳、文件大小、随机数和文件扩展名等。
FastDFS由跟踪服务器(Tracker Server)、存储服务器(Storage Server)和客户端(Client)构成,其中:
跟踪服务器Tracker Server
主要做调度工作,起到均衡的作用;负责管理所有的 storage server和 group,每个 storage 在启动后会连接 Tracker,告知自己所属 group 等信息,并保持周期性心跳。tracker根据storage的心跳信息,建立group==>[storage serverlist]的映射表。
Tracker需要管理的元信息很少,会全部存储在内存中;另外tracker上的元信息都是由storage汇报的信息生成的,本身不需要持久化任何数据,这样使得tracker非常容易扩展,直接增加tracker机器即可扩展为tracker cluster来服务,cluster里每个tracker之间是完全对等的,所有的tracker都接受stroage的心跳信息,生成元数据信息来提供读写服务。
存储服务器Storage Server
主要提供容量和备份服务;以 group 为单位,每个 group 内可以有多台 storage server,数据互为备份。以group为单位组织存储能方便的进行应用隔离、负载均衡、副本数定制(group内storage server数量即为该group的副本数),比如将不同应用数据存到不同的group就能隔离应用数据,同时还可根据应用的访问特性来将应用分配到不同的group来做负载均衡;缺点是group的容量受单机存储容量的限制,同时当group内有机器坏掉时,数据恢复只能依赖group内地其他机器,使得恢复时间会很长。
group内每个storage的存储依赖于本地文件系统,storage可配置多个数据存储目录,比如有10块磁盘,分别挂载在/data/disk1-/data/disk10,则可将这10个目录都配置为storage的数据存储目录。storage接受到写文件请求时,会根据配置好的规则选择其中一个存储目录来存储文件。为了避免单个目录下的文件数太多,在storage第一次启动时,会在每个数据存储目录里创建2级子目录,每级256个,总共65536个文件,新写的文件会以hash的方式被路由到其中某个子目录下,然后将文件数据作为本地文件存储到该目录中。
客户端Client
主要是上传下载数据的服务器,也就是我们自己的项目所部署在的服务器。每个客户端服务器都需要安装Nginx
基本架构如下图所示。
FastDFS的存储策略
为了支持大容量,存储节点(服务器)采用了分卷(或分组)的组织方式。存储系统由一个或多个卷组成,卷与卷之间的文件是相互独立的,所有卷的文件容量累加就是整个存储系统中的文件容量。一个卷可以由一台或多台存储服务器组成,一个卷下的存储服务器中的文件都是相同的,卷中的多台存储服务器起到了冗余备份和负载均衡的作用。
在卷中增加服务器时,同步已有的文件由系统自动完成,同步完成后,系统自动将新增服务器切换到线上提供服务。当存储空间不足或即将耗尽时,可以动态添加卷。只需要增加一台或多台服务器,并将它们配置为一个新的卷,这样就扩大了存储系统的容量。
FastDFS的上传过程
FastDFS向使用者提供基本文件访问接口,比如upload、download、append、delete等,以客户端库的方式提供给用户使用。
Storage Server会定期的向Tracker Server发送自己的存储信息。当Tracker Server Cluster中的Tracker Server不止一个时,各个Tracker之间的关系是对等的,所以客户端上传时可以选择任意一个Tracker。
当Tracker收到客户端上传文件的请求时,会为该文件分配一个可以存储文件的group,当选定了group后就要决定给客户端分配group中的哪一个storage server。当分配好storage server后,客户端向storage发送写文件请求,storage将会为文件分配一个数据存储目录。然后为文件分配一个fileid,最后根据以上的信息生成文件名存储文件。文件名的格式如下:
FastDFS的文件同步
写文件时,客户端将文件写至group内一个storage server即认为写文件成功,storage server写完文件后,会由后台线程将文件同步至同group内其他的storage server。
每个storage写文件后,同时会写一份binlog,binlog里不包含文件数据,只包含文件名等元信息,这份binlog用于后台同步,storage会记录向group内其他storage同步的进度,以便重启后能接上次的进度继续同步;进度以时间戳的方式进行记录,所以最好能保证集群内所有server的时钟保持同步。
storage的同步进度会作为元数据的一部分汇报到tracker上,tracke在选择读storage的时候会以同步进度作为参考。
客户端上传文件后存储与服务器将文件ID返回给客户端,此文件ID用于以后访问该文件的索引信息。文件索引信息包括:组名,虚拟磁盘路径,数据两级目录,文件名。
路径格式:
FastDFS的文件下载
客户端uploadfile成功后,会拿到一个storage生成的文件名,接下来客户端根据这个文件名即可访问到该文件。
跟upload file一样,在downloadfile时客户端可以选择任意tracker server。tracker发送download请求给某个tracker,必须带上文件名信息,tracke从文件名中解析出文件的group、大小、创建时间等信息,然后为该请求选择一个storage用来服务读请求。
FastDFS性能方案
三、Centos下部署FastDFS分布式存储操作记录
1)服务器信息(centos6.9)
描述 ip地址 主机名称 分组 跟踪服务器1 192.168.10.200 Fastdfs_tracker_t1 跟踪服务器2 192.168.10.201 Fastdfs_tracker_t2 存储服务器1 192.168.10.202 Fastdfs_storage_s1 group1 存储服务器2 192.168.10.203 Fastdfs_storage_s2 group2 存储服务器3 192.168.10.204 Fastdfs_storage_s3 group3 客户机1 192.168.10.205 Fastdfs_client
2)FastDFS安装(所有tracker服务器、storage服务器、客户端机器都要有如下操作。这里以Fastdfs_tracker_t1服务器操作为例)
1)编译和安装所需的依赖包 [root@Fastdfs_tracker_t1 ~]# yum install make cmake gcc gcc-c++ 2)安装libfastcommon [root@Fastdfs_tracker_t1 ~]# cd /usr/local/src/ [root@Fastdfs_tracker_t1 src]# wget https://codeload.github.com/happyfish100/libfastcommon/tar.gz/V1.0.7 -O libfastcommon-1.0.7.tar.gz [root@Fastdfs_tracker_t1 src]# tar zxf libfastcommon-1.0.7.tar.gz [root@Fastdfs_tracker_t1 src]# cd libfastcommon-1.0.7 [root@Fastdfs_tracker_t1 libfastcommon-1.0.7]# ./make.sh [root@Fastdfs_tracker_t1 libfastcommon-1.0.7]# ./make.sh install libfastcommon 默认安装到了/usr/lib64/libfastcommon.so和/usr/lib64/libfdfsclient.so 因为FastDFS主程序设置的lib目录是/usr/local/lib,所以需要创建软链接 [root@Fastdfs_tracker_t1 libfastcommon-1.0.7]# ln -s /usr/lib64/libfastcommon.so /usr/local/lib/libfastcommon.so [root@Fastdfs_tracker_t1 libfastcommon-1.0.7]# ln -s /usr/lib64/libfastcommon.so /usr/lib/libfastcommon.so [root@Fastdfs_tracker_t1 libfastcommon-1.0.7]# ln -s /usr/lib64/libfdfsclient.so /usr/local/lib/libfdfsclient.so [root@Fastdfs_tracker_t1 libfastcommon-1.0.7]# ln -s /usr/lib64/libfdfsclient.so /usr/lib/libfdfsclient.so 3)安装FastDFS 先提前安装依赖 [root@Fastdfs_tracker_t1 ~]# yum install gettext gettext-devel libXft libXft-devel libXpm libXpm-devel autoconf libxtst-devel zlib-devel libpng-devel glib-devel [root@Fastdfs_tracker_t1 ~]# cd /usr/local/src/ [root@Fastdfs_tracker_t1 src]# wget https://github.s3.amazonaws.com/downloads/libevent/libevent/libevent-2.0.20-stable.tar.gz [root@Fastdfs_tracker_t1 src]# tar -zvxf libevent-2.0.20-stable.tar.gz [root@Fastdfs_tracker_t1 src]# cd libevent-2.0.20-stable [root@Fastdfs_tracker_t1 libevent-2.0.20-stable]# ./configure && make && make install 接着编译安装FastDFS(编译前要确保已经成功安装了 libfastcommon)。百度云盘下载:https://pan.baidu.com/s/1x_CTcRqZr_BTptXF9kHOng 提取密码:9ku2 [root@Fastdfs_tracker_t1 src]# pwd /usr/local/src [root@Fastdfs_tracker_t1 src]# ls FastDFS_v5.05.tar.gz FastDFS_v5.05.tar.gz [root@Fastdfs_tracker_t1 src]# tar -zvxf FastDFS_v5.05.tar.gz [root@Fastdfs_tracker_t1 src]# cd FastDFS [root@Fastdfs_tracker_t1 FastDFS]# ./make.sh [root@Fastdfs_tracker_t1 FastDFS]# ./make.sh install 采用默认安装的方式安装,安装后的相应文件与目录: 服务脚本在: /etc/init.d/fdfs_storaged /etc/init.d/fdfs_trackerd 配置文件在(样例配置文件): /etc/fdfs/client.conf.sample /etc/fdfs/storage.conf.sample /etc/fdfs/tracker.conf.sample 命令工具在/usr/bin/目录下的: fdfs_appender_test fdfs_appender_test1 fdfs_append_file fdfs_crc32 fdfs_delete_file fdfs_download_file fdfs_file_info fdfs_monitor fdfs_storaged fdfs_test fdfs_test1 fdfs_trackerd fdfs_upload_appender fdfs_upload_file stop.sh restart.sh 因为FastDFS服务脚本设置的bin目录是/usr/local/bin,但实际命令安装在/usr/bin,可以进入/user/bin 目录使用以下命令查看 fdfs 的相关命令: [root@Fastdfs_tracker_t1 ~]# cd /usr/bin/ [root@Fastdfs_tracker_t1 bin]# ls | grep fdfs 因此需要修改FastDFS服务脚本中相应的命令路径,也就是把/etc/init.d/fdfs_storaged和/etc/init.d/fdfs_trackerd 两个脚本中的/usr/local/bin修改成/usr/bin: [root@Fastdfs_tracker_t1 bin]# vim /etc/init.d/fdfs_trackerd 使用查找替换命令进统一修改::%s+/usr/local/bin+/usr/bin [root@Fastdfs_tracker_t1 bin]# vim /etc/init.d/fdfs_storaged 使用查找替换命令进统一修改::%s+/usr/local/bin+/usr/bin
3)配置FastDFS跟踪器Tracker(两台机器Fastdfs_tracker_t1和Fastdfs_tracker_t2都需要操作)
1)复制 FastDFS 跟踪器样例配置文件,并重命名 [root@Fastdfs_tracker_t1 ~]# cd /etc/fdfs/ [root@Fastdfs_tracker_t1 fdfs]# cp tracker.conf.sample tracker.conf 2)编辑跟踪器配置文件 [root@Fastdfs_tracker_t1 fdfs]# vim /etc/fdfs/tracker.conf // 修改的内容如下(其它参数保留默认配置) ...... disabled=false //启用配置文件 port=22122 //tracker 的端口号,一般采用 22122 这个默认端口 base_path=/fastdfs/tracker //tracker 的数据文件和日志目录 # the method of selecting group to upload files # 0: round robin # 1: specify group # 2: load balance, select the max free space group to upload file store_lookup=0 //采取轮巡方式选择要上传的组进行存储,默认2 选择最大空闲空间的组 3)创建基础数据目录(参考基础目录 base_path 配置) [root@Fastdfs_tracker_t1 fdfs]# mkdir -p /fastdfs/tracker 4)关闭iptables防火墙 [root@Fastdfs_tracker_t1 fdfs]# /etc/init.d/iptables stop ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 温馨提示: 如果iptables防火墙中开启的话,则需要打开跟踪器端口(默认为 22122) [root@Fastdfs_tracker_t1 fdfs]# vim /etc/sysconfig/iptables ## FastDFS Tracker Port -A INPUT -m state --state NEW -m tcp -p tcp --dport 22122 -j ACCEPT [root@Fastdfs_tracker_t1 fdfs]# service iptables restart ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 5)启动Tracker服务 [root@Fastdfs_tracker_t1 fdfs]# /etc/init.d/fdfs_trackerd start Starting FastDFS tracker server: 初次成功启动,会在/fastdfs/tracker 目录下创建 data、logs 两个目录;可以查看tracker是否启动成功: [root@Fastdfs_tracker_t1 fdfs]# ls /fastdfs/tracker/ data logs [root@Fastdfs_tracker_t1 fdfs]# netstat -unltp|grep fdfs tcp 0 0 0.0.0.0:22122 0.0.0.0:* LISTEN 10469/fdfs_trackerd [root@Fastdfs_tracker_t1 fdfs]# lsof -i:22122 COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME fdfs_trac 10469 root 5u IPv4 214098 0t0 TCP *:22122 (LISTEN) [root@Fastdfs_tracker_t1 fdfs]# tail -100f /fastdfs/tracker/logs/trackerd.log 关闭Tracker [root@Fastdfs_tracker_t2 fdfs]# /etc/init.d/fdfs_trackerd stop 设置FastDFS跟踪器开机启动 [root@Fastdfs_tracker_t1 fdfs]# chkconfig fdfs_trackerd on [root@Fastdfs_tracker_t1 fdfs]# chkconfig --list|grep fdfs_trackerd fdfs_trackerd 0:off 1:off 2:on 3:on 4:on 5:on 6:off [root@Fastdfs_tracker_t1 fdfs]# vim /etc/rc.d/rc.local [root@Fastdfs_tracker_t1 fdfs]# cat /etc/rc.d/rc.local ....... /etc/init.d/fdfs_trackerd start
4)配置FastDFS存储服务(三台机器Fastdfs_storage_s1、Fastdfs_storage_s2、Fastdfs_storage_s3都需要操作)
1)复制FastDFS存储器样例配置文件,并重命名。 [root@Fastdfs_storage_s1 FastDFS]# cd /etc/fdfs/ [root@Fastdfs_storage_s1 fdfs]# cp storage.conf.sample storage.conf [root@Fastdfs_storage_s1 fdfs]# vim /etc/fdfs/storage.conf //修改的内容如下(其它参数保留默认配置): ....... disabled=false //启用配置文件 group_name=group1 //组名(第一组为 group1,第二组为 group2) port=23000 //storage 的端口号,同一个组的 storage 端口号必须相同 base_path=/fastdfs/storage //设置 storage 的日志目录 store_path0=/fastdfs/storage //存储路径 store_path_count=1 //存储路径个数,需要和 store_path 个数匹配 tracker_server=192.168.10.200:22122 //tracker 服务器的 IP 地址和端口 tracker_server=192.168.1.201:22122 //多个 tracker 直接添加多条配置 http.server_port=8888 //设置 http 端口号 2)创建基础数据目录(参考基础目录 base_path 配置) [root@Fastdfs_storage_s1 fdfs]# mkdir -p /fastdfs/storage 4)关闭iptables防火墙 [root@Fastdfs_storage_s1 fdfs]# /etc/init.d/iptables stop ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 温馨提示: 如果iptables防火墙中开启的话,则需要打开23000端口 [root@Fastdfs_tracker_t1 fdfs]# vim /etc/sysconfig/iptables ## FastDFS Tracker Port -A INPUT -m state --state NEW -m tcp -p tcp --dport 23000 -j ACCEPT [root@Fastdfs_storage_s1 fdfs]# service iptables restart ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 5)启动 Storage [root@Fastdfs_storage_s1 fdfs]# /etc/init.d/fdfs_storaged start Starting FastDFS storage server: 初次成功启动,会在/fastdfs/storage 目录下创建数据目录 data 和日志目录 logs)各节点启动动, 使用 tail -f /fastdfs/storage/logs/storaged.log 命令监听存储节点日志,可以看到存储节点链接到跟踪器, 并提示哪一个为 leader 跟踪器。同时也会看到同一组中的其他节点加入进来的日志信息。 查看23000端口监听情况: [root@Fastdfs_storage_s1 fdfs]# netstat -unltp|grep fdfs tcp 0 0 0.0.0.0:23000 0.0.0.0:* LISTEN 28347/fdfs_storaged [root@Fastdfs_storage_s1 fdfs]# lsof -i:23000 COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME fdfs_stor 28347 root 5u IPv4 211801 0t0 TCP *:inovaport1 (LISTEN) [root@Fastdfs_storage_s1 fdfs]# tail -f /fastdfs/storage/logs/storaged.log mkdir data path: FD ... mkdir data path: FE ... mkdir data path: FF ... data path: /fastdfs/storage/data, mkdir sub dir done. [2018-02-26 14:27:46] INFO - file: storage_param_getter.c, line: 191, use_storage_id=0, id_type_in_filename=ip, storage_ip_changed_auto_adjust=1, store_path=0, reserved_storage_space=10.00%, use_trunk_file=0, slot_min_size=256, slot_max_size=16 MB, trunk_file_size=64 MB, trunk_create_file_advance=0, trunk_create_file_time_base=02:00, trunk_create_file_interval=86400, trunk_create_file_space_threshold=20 GB, trunk_init_check_occupying=0, trunk_init_reload_from_binlog=0, trunk_compress_binlog_min_interval=0, store_slave_file_use_link=0 [2018-02-26 14:27:46] INFO - file: storage_func.c, line: 254, tracker_client_ip: 192.168.10.202, my_server_id_str: 192.168.10.202, g_server_id_in_filename: -905271104 [2018-02-26 14:27:51] INFO - local_host_ip_count: 2, 127.0.0.1 192.168.10.202 [2018-02-26 14:27:51] INFO - file: tracker_client_thread.c, line: 310, successfully connect to tracker server 192.168.10.201:22122, as a tracker client, my ip is 192.168.10.202 [2018-02-26 14:27:51] INFO - file: tracker_client_thread.c, line: 1235, tracker server 192.168.10.201:22122, set tracker leader: 192.168.10.201:22122 [2018-02-26 14:27:52] INFO - file: tracker_client_thread.c, line: 310, successfully connect to tracker server 192.168.10.200:22122, as a tracker client, my ip is 192.168.10.202 ....... 关闭storage [root@Fastdfs_storage_s1 fdfs]# /etc/init.d/fdfs_storaged stop 设置storage开机启动 [root@Fastdfs_storage_s1 fdfs]# chkconfig fdfs_storaged on [root@Fastdfs_storage_s1 fdfs]# chkconfig --list|grep fdfs_storaged fdfs_storaged 0:off 1:off 2:on 3:on 4:on 5:on 6:off [root@Fastdfs_storage_s1 fdfs]# vim /etc/rc.d/rc.local ...... /etc/init.d/fdfs_storaged start 所有Storage节点都启动之后,可以在任一 Storage 节点上使用如下命令查看集群信息: [root@Fastdfs_storage_s1 fdfs]# /usr/bin/fdfs_monitor /etc/fdfs/storage.conf [2018-02-26 14:30:22] DEBUG - base_path=/fastdfs/storage, connect_timeout=30, network_timeout=60, tracker_server_count=2, anti_steal_token=0, anti_steal_secret_key length=0, use_connection_pool=0, g_connection_pool_max_idle_time=3600s, use_storage_id=0, storage server id count: 0 server_count=2, server_index=0 tracker server is 192.168.10.200:22122 group count: 3 Group 1: group name = group1 disk total space = 50268 MB disk free space = 45029 MB trunk free space = 0 MB storage server count = 1 active server count = 1 storage server port = 23000 storage HTTP port = 8888 store path count = 1 subdir count per path = 256 current write server index = 0 current trunk file id = 0 Storage 1: id = 192.168.10.202 ip_addr = 192.168.10.202 ACTIVE http domain = version = 5.05 join time = 2018-02-26 14:27:07 up time = 2018-02-26 14:27:07 total storage = 50268 MB free storage = 45029 MB upload priority = 10 store_path_count = 1 subdir_count_per_path = 256 storage_port = 23000 storage_http_port = 8888 current_write_path = 0 source storage id = if_trunk_server = 0 connection.alloc_count = 256 connection.current_count = 0 connection.max_count = 0 total_upload_count = 0 success_upload_count = 0 total_append_count = 0 success_append_count = 0 total_modify_count = 0 success_modify_count = 0 total_truncate_count = 0 success_truncate_count = 0 total_set_meta_count = 0 success_set_meta_count = 0 total_delete_count = 0 success_delete_count = 0 total_download_count = 0 success_download_count = 0 total_get_meta_count = 0 success_get_meta_count = 0 total_create_link_count = 0 success_create_link_count = 0 total_delete_link_count = 0 success_delete_link_count = 0 total_upload_bytes = 0 success_upload_bytes = 0 total_append_bytes = 0 success_append_bytes = 0 total_modify_bytes = 0 success_modify_bytes = 0 stotal_download_bytes = 0 success_download_bytes = 0 total_sync_in_bytes = 0 success_sync_in_bytes = 0 total_sync_out_bytes = 0 success_sync_out_bytes = 0 total_file_open_count = 0 success_file_open_count = 0 total_file_read_count = 0 success_file_read_count = 0 total_file_write_count = 0 success_file_write_count = 0 last_heart_beat_time = 2018-02-26 14:29:52 last_source_update = 1970-01-01 08:00:00 last_sync_update = 1970-01-01 08:00:00 last_synced_timestamp = 1970-01-01 08:00:00 Group 2: group name = group2 disk total space = 49086 MB disk free space = 44696 MB trunk free space = 0 MB storage server count = 1 active server count = 1 storage server port = 23000 storage HTTP port = 8888 store path count = 1 subdir count per path = 256 current write server index = 0 current trunk file id = 0 Storage 1: id = 192.168.10.203 ip_addr = 192.168.10.203 ACTIVE http domain = version = 5.05 join time = 2018-02-26 14:27:09 up time = 2018-02-26 14:27:09 total storage = 49086 MB free storage = 44696 MB upload priority = 10 store_path_count = 1 subdir_count_per_path = 256 storage_port = 23000 storage_http_port = 8888 current_write_path = 0 source storage id = if_trunk_server = 0 connection.alloc_count = 256 connection.current_count = 0 connection.max_count = 0 total_upload_count = 0 success_upload_count = 0 total_append_count = 0 success_append_count = 0 total_modify_count = 0 success_modify_count = 0 total_truncate_count = 0 success_truncate_count = 0 total_set_meta_count = 0 success_set_meta_count = 0 total_delete_count = 0 success_delete_count = 0 total_download_count = 0 success_download_count = 0 total_get_meta_count = 0 success_get_meta_count = 0 total_create_link_count = 0 success_create_link_count = 0 total_delete_link_count = 0 success_delete_link_count = 0 total_upload_bytes = 0 success_upload_bytes = 0 total_append_bytes = 0 success_append_bytes = 0 total_modify_bytes = 0 success_modify_bytes = 0 stotal_download_bytes = 0 success_download_bytes = 0 total_sync_in_bytes = 0 success_sync_in_bytes = 0 total_sync_out_bytes = 0 success_sync_out_bytes = 0 total_file_open_count = 0 success_file_open_count = 0 total_file_read_count = 0 success_file_read_count = 0 total_file_write_count = 0 success_file_write_count = 0 last_heart_beat_time = 2018-02-26 14:30:15 last_source_update = 1970-01-01 08:00:00 last_sync_update = 1970-01-01 08:00:00 last_synced_timestamp = 1970-01-01 08:00:00 Group 3: group name = group3 disk total space = 49086 MB disk free space = 44962 MB trunk free space = 0 MB storage server count = 1 active server count = 1 storage server port = 23000 storage HTTP port = 8888 store path count = 1 subdir count per path = 256 current write server index = 0 current trunk file id = 0 Storage 1: id = 192.168.10.204 ip_addr = 192.168.10.204 ACTIVE http domain = version = 5.05 join time = 2018-02-26 14:27:10 up time = 2018-02-26 14:27:10 total storage = 49086 MB free storage = 44962 MB upload priority = 10 store_path_count = 1 subdir_count_per_path = 256 storage_port = 23000 storage_http_port = 8888 current_write_path = 0 source storage id = if_trunk_server = 0 connection.alloc_count = 256 connection.current_count = 0 connection.max_count = 0 total_upload_count = 0 success_upload_count = 0 total_append_count = 0 success_append_count = 0 total_modify_count = 0 success_modify_count = 0 total_truncate_count = 0 success_truncate_count = 0 total_set_meta_count = 0 success_set_meta_count = 0 total_delete_count = 0 success_delete_count = 0 total_download_count = 0 success_download_count = 0 total_get_meta_count = 0 success_get_meta_count = 0 total_create_link_count = 0 success_create_link_count = 0 total_delete_link_count = 0 success_delete_link_count = 0 total_upload_bytes = 0 success_upload_bytes = 0 total_append_bytes = 0 success_append_bytes = 0 total_modify_bytes = 0 success_modify_bytes = 0 stotal_download_bytes = 0 success_download_bytes = 0 total_sync_in_bytes = 0 success_sync_in_bytes = 0 total_sync_out_bytes = 0 success_sync_out_bytes = 0 total_file_open_count = 0 success_file_open_count = 0 total_file_read_count = 0 success_file_read_count = 0 total_file_write_count = 0 success_file_write_count = 0 last_heart_beat_time = 2018-02-26 14:29:55 last_source_update = 1970-01-01 08:00:00 last_sync_update = 1970-01-01 08:00:00 last_synced_timestamp = 1970-01-01 08:00:00
5)文件上传测试(在客户机192.168.10.205上测试即可,也可以直接在Fastdfs_tracker服务器如192.168.10.200上测试,只要在安装fastdfs_client机器上测试就行)
1)修改客户端配置文件 [root@Fastdfs_client ~]# cp /etc/fdfs/client.conf.sample /etc/fdfs/client.conf [root@Fastdfs_client ~]# vim /etc/fdfs/client.conf //修改的内容如下(其它参数保留默认配置): ...... base_path=/fastdfs/tracker tracker_server=192.168.10.200:22122 tracker_server=192.168.10.201:22122 2)创建基础数据目录(参考基础目录 base_path 配置) [root@Fastdfs_client ~]# mkdir -p /fastdfs/tracker 3)上传(可以通过Xftp5上传)一张图片(比如test.png)到客户机192.168.10.205目录/usr/local/src/,用作上传测试 [root@Fastdfs_client src]# pwd /usr/local/src [root@Fastdfs_client src]# ll test.png -rw-r--r--. 1 root root 165712 Feb 26 14:53 test.png 然后使用FastDFS上传命令,将测试图片上传到FastDFS存储里。 [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/test.png group1/M00/00/00/wKgKylqTsFCAdeEuAAKHUIZ3K9s287.png 如上能返回文件ID号,说明上传成功。由返回的信息可知,test.png图片被上传到group1下的数据目录下了。 到group1组内机器(即Fastdfs_storage_s1)的数据目录下可找到刚才上传的文件: [root@Fastdfs_storage_s1 ~]# ll /fastdfs/storage/data/00/00/ total 164 -rw-r--r--. 1 root root 165712 Feb 26 14:59 wKgKylqTsFCAdeEuAAKHUIZ3K9s287.png 同理,再上传一张图片进行测试 [root@Fastdfs_client ~]# cd /usr/local/src/ [root@Fastdfs_client src]# ll haha.png -rw-r--r--. 1 root root 12331 Feb 26 16:40 haha.png [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/haha.png group2/M00/00/00/wKgKy1qTyCKAFGzSAAAwK2jUNaY428.png 即haha.png图片被上传到group2组内机器的数据目录下了 [root@Fastdfs_storage_s2 ~]# ll /fastdfs/storage/data/00/00/ total 16 -rw-r--r--. 1 root root 12331 Feb 26 16:41 wKgKy1qTyCKAFGzSAAAwK2jUNaY428.png 再上传一次 [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/haha.png group3/M00/00/00/wKgKzFqTyRaAeDLJAAAwK2jUNaY992.png [root@Fastdfs_storage_s3 ~]# ll /fastdfs/storage/data/00/00/ total 16 -rw-r--r--. 1 root root 12331 Feb 26 16:45 wKgKzFqTyRaAeDLJAAAwK2jUNaY992.png 这里,我测试的时候,一个group组内只有一台存储服务器,测试图片上传到group内,只上传到一台存储服务器上了。 正常来说,应该要在同一个group组内添加2台以上的存储服务器,这样图片就是被上传到多台存储服务器上, 那么当一台存储服务器挂掉后,图片还能从其他的存储服务器上读出来。
6)在各个存储节点上安装Nginx(Fastdfs_storage_s1、Fastdfs_storage_s2、Fastdfs_storage_s3)
1)fastdfs-nginx-module 作用说明 FastDFS 通过Tracker服务器,将文件放在Storage服务器存储,但是同组存储服务器之间需要进入文件复制,有同步延迟的问题。 假设Tracker服务器将文件上传到了A机器上,上传成功后文件ID已经返回给客户端。此时FastDFS存储集群机制会将这个文件同步到同组存储B机器上, 在文件还没有复制完成的情况下,客户端如果用这个文件ID在B机器上取文件,就会出现文件无法访问的错误。 而fastdfs-nginx-module可以重定向文件连接到源服务器取文件,避免客户端由于复制延迟导致的文件无法访问错误。 (解压后的fastdfs-nginx-module在nginx安装时使用) 2)安装nginx前,需要配置以下依赖 [root@Fastdfs_storage_s1 ~]# yum install gcc gcc-c++ make automake autoconf libtool pcre pcre-devel zlib zlib-devel openssl openssl-devel 3)安装nginx和fastdfs-nginx-module [root@Fastdfs_storage_s1 ~]# cd /usr/local/src/ [root@Fastdfs_storage_s1 src]# wget http://nginx.org/download/nginx-1.13.9.tar.gz 下载fastdfs-nginx-module地址:https://pan.baidu.com/s/1i6Jqsi5 提取密码:g1ps [root@Fastdfs_storage_s1 src]# ll fastdfs-nginx-module_v1.16.tar.gz -rw-r--r--. 1 root root 17510 Feb 26 16:58 fastdfs-nginx-module_v1.16.tar.gz [root@Fastdfs_storage_s1 src]# ls nginx-1.13.9.tar.gz nginx-1.13.9.tar.gz [root@Fastdfs_storage_s1 src]# tar -zvxf nginx-1.13.9.tar.gz [root@Fastdfs_storage_s1 src]# tar -zvxf fastdfs-nginx-module_v1.16.tar.gz [root@Fastdfs_storage_s1 src]# vim /usr/local/src/fastdfs-nginx-module/src/config 将 CORE_INCS="$CORE_INCS /usr/local/include/fastdfs /usr/local/include/fastcommon/" 修改为: CORE_INCS="$CORE_INCS /usr/include/fastdfs /usr/include/fastcommon/" 千万注意:这个路径修改是很重要的,不然在 nginx 编译的时候会报错的! 紧接着进行nginx的编译安装操作 [root@Fastdfs_storage_s1 src]# cd nginx-1.13.9 [root@Fastdfs_storage_s1 nginx-1.13.9]# ./configure --prefix=/usr/local/nginx --add-module=/usr/local/src/fastdfs-nginx-module/src [root@Fastdfs_storage_s1 nginx-1.13.9]# make && make install 4)复制 fastdfs-nginx-module 源码中的配置文件到/etc/fdfs 目录,并修改 [root@Fastdfs_storage_s1 src]# cp /usr/local/src/fastdfs-nginx-module/src/mod_fastdfs.conf /etc/fdfs/ [root@Fastdfs_storage_s1 src]# vim /etc/fdfs/mod_fastdfs.conf ...... connect_timeout=10 base_path=/tmp tracker_server=192.168.10.200:22122 tracker_server=192.168.10.201:22122 storage_server_port=23000 group_name=group1 url_have_group_name = true store_path0=/fastdfs/storage group_count = 3 [group1] group_name=group1 storage_server_port=23000 store_path_count=1 store_path0=/fastdfs/storage [group2] group_name=group2 storage_server_port=23000 store_path_count=1 store_path0=/fastdfs/storage [group3] group_name=group3 storage_server_port=23000 store_path_count=1 store_path0=/fastdfs/storage 特别注意: 第二组内的Storage服务器(这里指Fastdfs_storage_s2)的mod_fastdfs.conf配置与第一组配置只有group_nam不同,即: group_name=group2 第三组内的Storage服务器(这里指Fastdfs_storage_s3)的mod_fastdfs.conf 配置与第一组配置只有group_name不同,即: group_name=group3 5)复制 FastDFS 的部分配置文件到/etc/fdfs 目录 [root@Fastdfs_storage_s1 src]# cd /usr/local/src/FastDFS/conf [root@Fastdfs_storage_s1 conf]# cp http.conf mime.types /etc/fdfs/ 6)在/fastdfs/storage 文件存储目录下创建软连接,将其链接到实际存放数据的目录 [root@Fastdfs_storage_s1 conf]# ln -s /fastdfs/storage/data/ /fastdfs/storage/data/M00 7)配置 Nginx,简洁版 nginx 配置样例 [root@Fastdfs_storage_s1 src]# cd /usr/local/nginx/conf/ [root@Fastdfs_storage_s1 conf]# cp nginx.conf nginx.conf.bak [root@Fastdfs_storage_s1 conf]# cat nginx.conf user www; //如果url访问图片时出现404,可以将这个修改为"user root" worker_processes 8; #error_log logs/error.log; #error_log logs/error.log notice; #error_log logs/error.log info; #pid logs/nginx.pid; events { worker_connections 65535; } http { include mime.types; default_type application/octet-stream; charset utf-8; ###### ## set access log format ###### log_format main '$http_x_forwarded_for $remote_addr $remote_user [$time_local] "$request" ' '$status $body_bytes_sent "$http_referer" ' '"$http_user_agent" "$http_cookie" $host $request_time'; ####### ## http setting ####### sendfile on; tcp_nopush on; tcp_nodelay on; keepalive_timeout 65; proxy_cache_path /var/www/cache levels=1:2 keys_zone=mycache:20m max_size=2048m inactive=60m; proxy_temp_path /var/www/cache/tmp; fastcgi_connect_timeout 3000; fastcgi_send_timeout 3000; fastcgi_read_timeout 3000; fastcgi_buffer_size 256k; fastcgi_buffers 8 256k; fastcgi_busy_buffers_size 256k; fastcgi_temp_file_write_size 256k; fastcgi_intercept_errors on; # client_header_timeout 600s; client_body_timeout 600s; # client_max_body_size 50m; client_max_body_size 100m; #允许客户端请求的最大单个文件字节数 client_body_buffer_size 256k; #缓冲区代理缓冲请求的最大字节数,可以理解为先保存到本地再传给用户 gzip on; gzip_min_length 1k; gzip_buffers 4 16k; gzip_http_version 1.1; gzip_comp_level 9; gzip_types text/plain application/x-javascript text/css application/xml text/javascript application/x-httpd-php; gzip_vary on; ## includes vhosts include vhosts/*.conf; } [root@Fastdfs_storage_s1 conf]# mkdir vhosts [root@Fastdfs_storage_s1 conf]# cd vhosts/ [root@Fastdfs_storage_s1 vhosts]# vim fastdfs.conf [root@Fastdfs_storage_s1 vhosts]# cat fastdfs.conf server { listen 8888; server_name localhost; location /group1/M00 { //注意,Fastdfs_storage_s2机器内此处修改为/group2/M00,Fastdfs_storage_s3机器内此处修改为/group3/M00 #alias /fastdfs/storage/data; ngx_fastdfs_module; } error_page 500 502 503 504 /50x.html; location = /50x.html { root html; } } [root@Fastdfs_storage_s1 vhosts]#useradd www [root@Fastdfs_storage_s1 vhosts]# mkdir -p /var/www/cache [root@Fastdfs_storage_s1 vhosts]# /usr/local/nginx/sbin/nginx -t [root@Fastdfs_storage_s1 vhosts]# /usr/local/nginx/sbin/nginx [root@Fastdfs_storage_s1 vhosts]# /usr/local/nginx/sbin/nginx -s reload [root@Fastdfs_storage_s1 vhosts]# lsof -i:8888 COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME nginx 31543 root 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31617 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31618 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31619 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31620 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31621 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31622 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31623 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) nginx 31624 www 10u IPv4 218910 0t0 TCP *:ddi-tcp-1 (LISTEN) +++++++++++++++++++++++++++++++++++++++++++特别注意++++++++++++++++++++++++++++++++++++++++++++++++++ A、8888 端口值是要与/etc/fdfs/storage.conf 中的 http.server_port=8888 相对应, 因为 http.server_port 默认为 8888,如果想改成 80,则要对应修改过来。 B、Storage 对应有多个 group 的情况下,访问路径带 group 名,如/group1/M00/00/00/xxx, 对应的 Nginx 配置为: location ~/group([0-9])/M00 { ngx_fastdfs_module; } C、如查下载时如发现老报 404,将 nginx.conf 第一行 user www 修改为 user root 后重新启动!这个要特别注意!!!!! ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 8)设置Nginx开机启动 [root@Fastdfs_storage_s1 ~]# vim /etc/rc.d/rc.local ........ /usr/local/nginx/sbin/nginx 9)关闭iptables防火墙 [root@Fastdfs_storage_s1 conf]# /etc/init.d/iptables stop ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 温馨提示: 如果打开了iptables防火墙功能,则需要开通8888端口访问 [root@Fastdfs_storage_s1 ~]# vim /etc/sysconfig/iptables ## FastDFS Tracker Port -A INPUT -m state --state NEW -m tcp -p tcp --dport 8888 -j ACCEPT [root@Fastdfs_storage_s1 ~]# service iptables restart ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 10)通过浏览器访问测试时上传的文件(对应上传到FastDFS存储里的路径) 在客户端机器上传图片 [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/haha.png group1/M00/00/00/wKgKylqT3OCAUrqYAAAwK2jUNaY262.png [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/test.png group2/M00/00/00/wKgKy1qT3P-Ae-vQAAKHUIZ3K9s459.png [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/test.png group3/M00/00/00/wKgKzFqT3SWAaunZAAKHUIZ3K9s849.png 然后访问图片的url http://192.168.10.202:8888/group1/M00/00/00/wKgKylqT3OCAUrqYAAAwK2jUNaY262.png http://192.168.10.203:8888/group2/M00/00/00/wKgKy1qT3P-Ae-vQAAKHUIZ3K9s459.png http://192.168.10.204:8888/group3/M00/00/00/wKgKzFqT3SWAaunZAAKHUIZ3K9s849.png
7)在各个跟踪节点服务器上安装Nginx(Fastdfs_tracker_t1、Fastdfs_tracker_t2)
在tracker上安装的nginx主要为了提供 http 访问的反向代理、负载均衡以及缓存服务
1)安装编译 Nginx 所需的依赖包 [root@Fastdfs_tracker_t1 ~]# yum install gcc gcc-c++ make automake autoconf libtool pcre pcre-devel zlib zlib-devel openssl openssl-devel 2)上传 ngx_cache_purge-2.3.tar.gz和nginx-1.13.9.tar.gz到/usr/local/src,解压 [root@Fastdfs_tracker_t1 ~]# cd /usr/local/src/ [root@Fastdfs_tracker_t1 src]# wget http://labs.frickle.com/files/ngx_cache_purge-2.3.tar.gz [root@Fastdfs_tracker_t1 src]# wget http://nginx.org/download/nginx-1.13.9.tar.gz [root@Fastdfs_tracker_t1 src]# tar -zvxf ngx_cache_purge-2.3.tar.gz [root@Fastdfs_tracker_t1 src]# tar -zvxf nginx-1.13.9.tar.gz [root@Fastdfs_tracker_t1 src]# cd nginx-1.13.9 [root@Fastdfs_tracker_t1 nginx-1.13.9]# ./configure --prefix=/usr/local/nginx --add-module=/usr/local/src/ngx_cache_purge-2.3 && make && make install 3)配置 Nginx,设置负载均衡以及缓存 [root@Fastdfs_tracker_t1 conf]# cp nginx.conf nginx.conf.conf [root@Fastdfs_tracker_t1 conf]# cat nginx.conf user www; worker_processes 8; #error_log logs/error.log; #error_log logs/error.log notice; #error_log logs/error.log info; #pid logs/nginx.pid; events { worker_connections 65535; } http { include mime.types; default_type application/octet-stream; charset utf-8; ###### ## set access log format ###### log_format main '$http_x_forwarded_for $remote_addr $remote_user [$time_local] "$request" ' '$status $body_bytes_sent "$http_referer" ' '"$http_user_agent" "$http_cookie" $host $request_time'; ####### ## http setting ####### sendfile on; tcp_nopush on; tcp_nodelay on; keepalive_timeout 65; proxy_cache_path /fastdfs/cache/nginx/proxy_cache levels=1:2 keys_zone=mycache:20m max_size=2048m inactive=60m; proxy_temp_path /fastdfs/cache/nginx/proxy_cache/tmp; fastcgi_connect_timeout 3000; fastcgi_send_timeout 3000; fastcgi_read_timeout 3000; fastcgi_buffer_size 256k; fastcgi_buffers 8 256k; fastcgi_busy_buffers_size 256k; fastcgi_temp_file_write_size 256k; fastcgi_intercept_errors on; # client_header_timeout 600s; client_body_timeout 600s; # client_max_body_size 50m; client_max_body_size 100m; client_body_buffer_size 256k; gzip on; gzip_min_length 1k; gzip_buffers 4 16k; gzip_http_version 1.1; gzip_comp_level 9; gzip_types text/plain application/x-javascript text/css application/xml text/javascript application/x-httpd-php; gzip_vary on; ## includes vhosts include vhosts/*.conf; } [root@Fastdfs_tracker_t1 conf]# mkdir vhosts [root@Fastdfs_tracker_t1 vhosts]# cat lb.conf //如果group组内有多个存储服务器,则这里的负载就起作用了。 upstream fdfs_group1 { server 192.168.10.202:8888 weight=1 max_fails=2 fail_timeout=30s; } upstream fdfs_group2 { server 192.168.10.203:8888 weight=1 max_fails=2 fail_timeout=30s; } upstream fdfs_group3 { server 192.168.10.204:8888 weight=1 max_fails=2 fail_timeout=30s; } server{ listen 8000; server_name localhost; access_log /usr/local/nginx/logs/lb_access.log main; error_log /usr/local/nginx/logs/lb_error.log; location /group1/M00 { proxy_pass http://fdfs_group1; } location /group2/M00 { proxy_pass http://fdfs_group2; } location /group3/M00 { proxy_pass http://fdfs_group3; } #设置清除缓存的访问权限 location ~/purge(/.*) { allow 127.0.0.1; allow 114.242.193.201; allow 192.168.10.0/24; deny all; proxy_cache_purge mycache $host$1$is_args$args ; } } 按以上 nginx 配置文件的要求,创建对应的缓存目录 [root@Fastdfs_tracker_t1 conf]# useradd www [root@Fastdfs_tracker_t1 conf]# ulimit 65535 [root@Fastdfs_tracker_t1 conf]# mkdir -p /fastdfs/cache/nginx/proxy_cache [root@Fastdfs_tracker_t1 conf]# mkdir -p /fastdfs/cache/nginx/proxy_cache/tmp [root@Fastdfs_tracker_t1 conf]# /usr/local/nginx/sbin/nginx -t [root@Fastdfs_tracker_t1 conf]# /usr/local/nginx/sbin/nginx [root@Fastdfs_tracker_t1 conf]# /usr/local/nginx/sbin/nginx -s reload 4)关闭iptables防火墙 [root@Fastdfs_tracker_t1 conf]# /etc/init.d/iptables stop ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 温馨提示: 如果iptables防火墙中开启的话,则需要打开8000端口 [root@Fastdfs_tracker_t1 fdfs]# vim /etc/sysconfig/iptables ## FastDFS Tracker Port -A INPUT -m state --state NEW -m tcp -p tcp --dport 8000 -j ACCEPT [root@Fastdfs_tracker_t1 conf]# service iptables restart ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ 5)设置Nginx开机启动 [root@Fastdfs_storage_s1 ~]# vim /etc/rc.d/rc.local ........ /usr/local/nginx/sbin/nginx 6)文件访问测试 在客户端机器上传图片(或使用上面测试中已经上传过的图片路径进行访问测试) [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/test.jpg group1/M00/00/00/wKgKylqUxBOAFo8hAAKHUIZ3K9s443.jpg [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/haha.png group2/M00/00/00/wKgKy1qUxBeAT0PxAAAwK2jUNaY575.png [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/test.jpg group3/M00/00/00/wKgKzFqUxBmATFurAAKHUIZ3K9s282.jpg 然后访问图片的url http://192.168.10.200:8000/group1/M00/00/00/wKgKylqUxBOAFo8hAAKHUIZ3K9s443.jpg http://192.168.10.201:8000/group1/M00/00/00/wKgKylqUxBOAFo8hAAKHUIZ3K9s443.jpg http://192.168.10.200:8000/group2/M00/00/00/wKgKy1qUxBeAT0PxAAAwK2jUNaY575.png http://192.168.10.201:8000/group2/M00/00/00/wKgKy1qUxBeAT0PxAAAwK2jUNaY575.png http://192.168.10.200:8000/group3/M00/00/00/wKgKzFqUxBmATFurAAKHUIZ3K9s282.jpg http://192.168.10.201:8000/group3/M00/00/00/wKgKzFqUxBmATFurAAKHUIZ3K9s282.jpg 以上是简单配置,多次测试可以发现,每一个Tracker中的 Nginx都单独对后端的Storage存储服务器组做了负载均衡, 但整套FastDFS 集群如果想对外提供统一的文件访问地址,还需要对两个Tracker中的Nginx进行HA集群。 即对Fastdfs_tracker_t1(192.168.10.200)和Fastdfs_tracker_t2(192.168.10.201)做keepalived+nginx的高可用。 192.168.10.200 Master 192.168.10.201 Slave 192.168.10.210 VIP 也就是说整套FastDFS 集群对外提供统一的文件访问(VIP)地址192.168.10.210. Keepalived+Nginx 主从部署记录可以参考:http://www.cnblogs.com/kevingrace/p/6138185.html,部署记录这里就不介绍了
8)fdfs常用命令
1)上传文件命令:fdfs_upload_file 不仅可以上传图片文件,也可以上传其他文件。 [root@Fastdfs_client src]# echo "hahaha" > test.txt [root@Fastdfs_client src]# /usr/bin/fdfs_upload_file /etc/fdfs/client.conf /usr/local/src/test.txt group3/M00/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt 上传成功后,返回一串带有组和路径标识的字符串,这是文件的id,然后到group3组内的存储服务器上查看这个文件的id信息 [root@Fastdfs_storage_s3 ~]# ll /fastdfs/storage/data/00/00/ total 84 -rw-r--r--. 1 root root 7 Feb 27 11:02 wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt [root@Fastdfs_storage_s3 ~]# cat /fastdfs/storage/data/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt hahaha [root@Fastdfs_storage_s3 ~]# md5sum /fastdfs/storage/data/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt 5a6d311c0d8f6d1dd03c1c129061d3b1 /fastdfs/storage/data/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt 2)下载文件命令:fdfs_download_file 在客户端机器上将上面上传到Fastdfs存储里文文件下载下来 [root@Fastdfs_client ~]# cd /mnt/ [root@Fastdfs_client mnt]# /usr/bin/fdfs_download_file /etc/fdfs/client.conf group3/M00/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt [root@Fastdfs_client mnt]# ls wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt [root@Fastdfs_client mnt]# cat wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt hahaha [root@Fastdfs_client mnt]# md5sum wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt 5a6d311c0d8f6d1dd03c1c129061d3b1 wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt 下载后可以对比文件的md5值。 3)查看文件信息命令:fdfs_file_info [root@Fastdfs_client ~]# /usr/bin/fdfs_file_info /etc/fdfs/client.conf group3/M00/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt source storage id: 0 source ip address: 192.168.10.204 file create timestamp: 2018-02-27 11:02:02 file size: 7 file crc32: 3909919332 (0xE90CA264) 4)删除文件命令:fdfs_delete_file [root@Fastdfs_client ~]# /usr/bin/fdfs_delete_file /etc/fdfs/client.conf group3/M00/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt 删除之后没有信息提示,我们再进行下载和查找,结果如下: [root@Fastdfs_client mnt]# /usr/bin/fdfs_download_file /etc/fdfs/client.conf group3/M00/00/00/wKgKzFqUyiqAabn8AAAAB-kMomQ229.txt [2018-02-27 14:21:08] ERROR - file: tracker_proto.c, line: 48, server: 192.168.10.204:23000, response status 2 != 0 download file fail, error no: 2, error info: No such file or directory 到group3组内的存储服务器上查看,发现没有这个文件的id信息 [root@Fastdfs_storage_s3 ~]# ll /fastdfs/storage/data/00/00/ [root@Fastdfs_storage_s3 ~]# 5)追加文件内容命令:fdfs_upload_appender 和 fdfs_append_file 先使用命令上传一个appender文件 [root@Fastdfs_client src]# echo "123456" > haha.txt [root@Fastdfs_client src]# /usr/bin/fdfs_upload_appender /etc/fdfs/client.conf /usr/local/src/haha.txt group1/M00/00/00/wKgKylqU_myEa194AAAAACXX7BE787.txt 到group1组内的存储服务器上查看这个文件 [root@Fastdfs_storage_s1 ~]# ll /fastdfs/storage/data/00/00/ total 12 -rw-r--r--. 1 root root 7 Feb 27 14:45 wKgKylqU_myEa194AAAAACXX7BE787.txt [root@Fastdfs_storage_s1 ~]# cat /fastdfs/storage/data/00/00/wKgKylqU_myEa194AAAAACXX7BE787.txt 123456 然后进行内容追加 [root@Fastdfs_client src]# echo "abcdefgh" > heihei.txt [root@Fastdfs_client src]# /usr/bin/fdfs_append_file /etc/fdfs/client.conf group1/M00/00/00/wKgKylqU_myEa194AAAAACXX7BE787.txt /usr/local/src/heihei.txt 再次到group1组内的存储服务器上查看这个文件内容 [root@Fastdfs_storage_s1 ~]# cat /fastdfs/storage/data/00/00/wKgKylqU_myEa194AAAAACXX7BE787.txt 123456 abcdefgh 或者将文件下载到本地查看 [root@Fastdfs_client src]# cd /mnt/ [root@Fastdfs_client mnt]# /usr/bin/fdfs_download_file /etc/fdfs/client.conf group1/M00/00/00/wKgKylqU_myEa194AAAAACXX7BE787.txt [root@Fastdfs_client mnt]# ls wKgKylqU_myEa194AAAAACXX7BE787.txt [root@Fastdfs_client mnt]# cat wKgKylqU_myEa194AAAAACXX7BE787.txt 123456 abcdefgh 6)监控服务器状态命令:fdfs_monitor [root@Fastdfs_client src]# /usr/bin/fdfs_monitor /etc/fdfs/client.conf [2018-02-27 14:49:06] DEBUG - base_path=/fastdfs/tracker, connect_timeout=30, network_timeout=60, tracker_server_count=2, anti_steal_token=0, anti_steal_secret_key length=0, use_connection_pool=0, g_connection_pool_max_idle_time=3600s, use_storage_id=0, storage server id count: 0 server_count=2, server_index=0 tracker server is 192.168.10.200:22122 group count: 3 Group 1: group name = group1 disk total space = 50268 MB disk free space = 44987 MB trunk free space = 0 MB storage server count = 1 active server count = 1 storage server port = 23000 storage HTTP port = 8888 store path count = 1 subdir count per path = 256 current write server index = 0 current trunk file id = 0 Storage 1: id = 192.168.10.202 ip_addr = 192.168.10.202 ACTIVE http domain = ........ ........ 7)移除故障节点示例 如果一个group组内有多个节点,其中一个节点的上的fastdfs出了问题,需要将它从组中移除,可根据下面的命令操作: a)停掉fastdfs上的storage服务 # /etc/init.d/fdfs_storaged stop # lsof -i :23000 //确保storage服务停止 b)将问题节点(比如group1组内的节点storage-node3,ip为192.168.10.23)从组中移除 # /usr/bin/fdfs_monitor /etc/fdfs/client.conf delete group1 192.168.10.23 c)移除成功,使用fdfs_monitor命令进行监控 # /usr/bin/fdfs_monitor /etc/fdfs/client.conf 通过监控,可以发现移除之后,storage-node3节点的状态为DELETED。 d)当storage-node3上的fastdfs修复后,可以再把它加入到fastdfs组中,加入的操作只需要把storage-node3上的fastdfs服务正常启动就好, 因为配置文件已经做了配置,如果服务正常的话,修复后的节点可以自动加入到fastdfs组中 启动fastdfs_storaged服务 # /etc/init.d/fdfs_storaged stop # lsof -i :23000 使用fdfs_monitor命令进行监控 # /usr/bin/fdfs_monitor /etc/fdfs/client.conf 修复后,再次监控,可以看到,此时storage-node3已经变为ACTIVE状态了。