前提：所用镜像image: luhuiguo/fastdfs
其他镜像命令自行修改
思维流程导向：old-tracker storage ——>加一个新的storage做同步用——>查看状态等待数据同步完成——>断开连接——>修改tracker为新的地址——>修改storage0/data/.data_init_flag配置文件（sync_src_server=）置为空——>重新启动storage——>成功！！！
1、安装fastdfs（开始）
旧tracker：192.168.180.46
新tracker：192.168.180.48
eg：存储一个文件返回路径为-做为后续查找使用：http://192.168.180.46/group1/M00/00/00/wKi0Ll3bji6AKxJiAAAABncc3SA697_big.txt
2、修改新的Storage配置文件，部署在新的服务器②192.168.180.48上（只安装storage，tracker先用旧的tracker）
vim leo_docker-compose.yaml
version: '3.0'
services:
fastdfs:
image: luhuiguo/fastdfs
container_name: tracker
network_mode: host
command: tracker
volumes:
- ./fdfs/tracker:/var/fdfs
storage0:
container_name: storage0
image: luhuiguo/fastdfs
command: storage
network_mode: host
environment:
- TRACKER_SERVER=192.168.180.46:22122
- GROUP_NAME=group1
volumes:
- ./fdfs/storage0:/var/fdfs
注：TRACKER_SERVER配置为旧的tracker
3、启动storage和新的tracker
docker-compose -f leo_docker-compose.yaml up -d
4、查看监控状态
新的storage节点已经加进来了
且ip_addr = 192.168.180.48 (test3) ACTIVE
如果ip_addr = 192.168.180.48 (test3) SYNCING表示正在同步数据
ip_addr = 192.168.180.48 (test3) WAIT_SYNC等待同步
docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
[root@test1 luhuiguo_fastdfs]# docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
[2019-11-25 08:31:31] DEBUG - base_path=/var/fdfs, connect_timeout=30, network_timeout=60, tracker_server_count=1, 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=1, server_index=0
tracker server is 192.168.180.46:22122
group count: 2
Group 1:
group name = group1
disk total space = 120819 MB
disk free space = 100004 MB
trunk free space = 0 MB
storage server count = 2
active server count = 2
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 = 192.168.180.46
ip_addr = 192.168.180.46 (test1) ACTIVE
……
Storage 2:
id = 192.168.180.48
ip_addr = 192.168.180.48 (test3) ACTIVE
……
Group 2:
group name = group2
disk total space = 120819 MB
disk free space = 100004 MB
trunk free space = 0 MB
storage server count = 1
active server count = 1
storage server port = 22222
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 = 192.168.180.46
ip_addr = 192.168.180.46 (test1) ACTIVE
……
5、这时我们在新的服务器上查看之前在旧的fastdfs上传的文件（根据之前存储的返回路径http://192.168.180.46/group1/M00/00/00/wKi0Ll3bji6AKxJiAAAABncc3SA697_big.txt）
[root@test3 data]# cat 00/00/wKi0Ll3bji6AKxJiAAAABncc3SA697_big.txt
hello
可以看到已经同步到这台新的服务器上了
6、同步完成后,停掉新的Storage服务器进程
docker-compose -f leo_docker-compose.yaml stop storage0
再次查看状态
docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
# 可以看到已经下线了ip_addr = 192.168.180.48 (test3) OFFLINE
[2019-11-25 08:31:31] DEBUG - base_path=/var/fdfs, connect_timeout=30, network_timeout=60, tracker_server_count=1, 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=1, server_index=0
tracker server is 192.168.180.46:22122
group count: 2
Group 1:
group name = group1
disk total space = 120819 MB
disk free space = 100004 MB
trunk free space = 0 MB
storage server count = 2
active server count = 2
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 = 192.168.180.46
ip_addr = 192.168.180.46 (test1) ACTIVE
……
Storage 2:
id = 192.168.180.48
ip_addr = 192.168.180.48 (test3) OFFLINE
……
Group 2:
group name = group2
disk total space = 120819 MB
disk free space = 100004 MB
trunk free space = 0 MB
storage server count = 1
active server count = 1
storage server port = 22222
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 = 192.168.180.46
ip_addr = 192.168.180.46 (test1) ACTIVE
……
7、（docker-compose启动不用执行此步骤，二进制需要）修改.data_init_flag文件，将文件中sync_src_server配置项留空，其它配置项不变
sync_src_server=
[root@test3 data]# pwd
/home/fastdfs/fdfs/storage0/data
[root@test3 data]# ll -a
[root@test3 data]# vim .data_init_flag
storage_join_time=1574668867
sync_old_done=1
sync_src_server=
sync_until_timestamp=1574668870
last_ip_addr=192.168.180.48
last_server_port=23000
last_http_port=8888
current_trunk_file_id=0
trunk_last_compress_time=0
8、重新启动storage
docker-compose -f leo_docker-compose.yaml up -d storage0
9、查看状态已经切换到了新的tracker（结束）
docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
[root@test3 fastdfs]# docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
[2019-11-25 09:46:45] DEBUG - base_path=/var/fdfs, connect_timeout=30, network_timeout=60, tracker_server_count=1, 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=1, server_index=0
tracker server is 192.168.180.48:22122
group count: 1
Group 1:
group name = group1
disk total space = 120819 MB
disk free space = 107597 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 = 192.168.180.48
ip_addr = 192.168.180.48 (test3) ACTIVE
http domain =
验证1、如果想要再次验证，就再找一台服务器创建storage连接新的tracker。
查看连接状态（ip_addr = 192.168.180.47 (test2) ACTIVE ）为验证的服务器
[root@test3 fastdfs]# docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
[2019-11-25 09:53:30] DEBUG - base_path=/var/fdfs, connect_timeout=30, network_timeout=60, tracker_server_count=1, 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=1, server_index=0
tracker server is 192.168.180.48:22122
group count: 1
Group 1:
group name = group1
disk total space = 120819 MB
disk free space = 101093 MB
trunk free space = 0 MB
storage server count = 2
active server count = 2
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 = 192.168.180.47
ip_addr = 192.168.180.47 (test2) ACTIVE
……
Storage 2:
id = 192.168.180.48
ip_addr = 192.168.180.48 (test3) ACTIVE
http domain =
version = 5.11
……
1.1、在新加的storage上查看我们迁移过来的数据——因为是同一个group所以是同步的，根据之前返回的数据路径，在新加的storage服务器上查看。
[root@test2 cluster]# cat fdfs/storage0/data/00/00/wKi0Ll3bji6AKxJiAAAABncc3SA697_big.txt
hello
可以看到数据已经同步过来了！
使用备注：
docker-compose修改配置文件需要重启其中一个service需要
docker-compose -f leo_docker-compose.yaml stop
docker-compose -f leo_docker-compose.yaml up -d storage0 #不能用start启动
验证2、
①、上传
eg：
/root@localhost:/# usr/bin/fdfs_test /etc/fdfs/client.conf upload a.txt
②、下载
eg：[root@test1 fastdfs]# /usr/bin/fdfs_test /etc/fdfs/client.conf download group1 M00/00/00/wKi0Ll3bji6AKxJiAAAABncc3SA697_big.txt
Usage: /usr/bin/fdfs_test <config_file> download <group_name> <remote_filename>
③、删除（所有storage中的文件都会删除）
eg：[root@test1 fastdfs]# /usr/bin/fdfs_test /etc/fdfs/client.conf delete group1 M00/00/00/wKi0Ll3cwVCAHq1oAAAADj9UaMQ484.txt
Usage: /usr/bin/fdfs_test <config_file> delete <group_name> <remote_filename>
验证3、不同情况下同步合并验证
===========46===========
①上传example file url: http://192.168.180.46/group1/M00/00/00/wKi0Ll3c2tGACRe1AAAACfdYteI663_big.txt
leojiang
②上传example file url: http://192.168.180.46/group1/M00/00/00/wKi0Ll3c4vGAeuJnAAAACBgItVc245_big.txt
maosiyu
③同步lihongbao
④上传example file url: http://192.168.180.46/group1/M00/00/00/wKi0Ll3c6C2AUFlvAAAAA7UVEzU990_big.txt
46
④同步48
===========48===========
①同步leojiang
②同步maosiyu
③上传example file url: http://192.168.180.48/group1/M00/00/00/wKi0MF3c5R2AA5oKAAAACit7gaU604_big.txt
lihongbao
④上传example file url: http://192.168.180.48/group1/M00/00/00/wKi0MF3c52SAVkcXAAAAA9SWPrs177_big.txt
48
④同步46
命令参考：
查看tracker的配置
docker exec -it tracker fdfs_monitor /etc/fdfs/client.conf
参数说明：
tracker_server_count：2 --表示2个Tracker Server
tracker server is 198.168.1.121:22122 --表示Leader Tracker
group count: 1 --表示有1个group
group name = group1 --组名称是group1
storage server count = 2 --组内有2个storage
active server count = 2 --活动的storage有2个
storage server port = 23002 --storage的端口
storage HTTP port = 9101 --storage的文件访问端口
store path count = 1 --storage只挂了一个存储目录
total_upload_count = 11 --总共上传了多少个文件
total_upload_bytes = 691405 --总共上传了多少字节
success_upload_bytes = 691405 --成功上传了多少字节
total_download_count = 2 --总共下载了多少文件（使用java客户端）
查看状态：
docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf
# 从集群中删除storage
docker exec -it storage0 fdfs_monitor /etc/fdfs/client.conf delete group1 10.1.8.101
创建文件：
root@localhost:/# echo hello>a.txt
上传文件：
/usr/bin/fdfs_test /etc/fdfs/client.conf upload a.txt
/usr/bin/fdfs_test /etc/fdfs/client.conf download group1 M00/02/32/CgAUD1s6716ATKtBAAMd3jd5Lxc105.png
1）wKjyglxkBuSAbT8iAAAAAc3XBpM982.txt 是主文件 。
2）wKjyglxkBuSAbT8iAAAAAc3XBpM982_big.txt 是从文件。
3)wKjyglxkBuSAbT8iAAAAAc3XBpM982.txt-m 文件 是 wKjyglxkBuSAbT8iAAAAAc3XBpM982.txt文件 属性文件，意思就是：wKjyglxkBuSAbT8iAAAAAc3XBpM982.txt-m中存放的是wKjyglxkBuSAbT8iAAAAAc3XBpM982.txt文件的各项属性的信息，比如文件的类型，高度和宽度等等属性的信息。
upload：上传普通文件，包括主文件
upload_appender：上传appender文件，后续可以对其进行append、modify和truncate操作
upload_slave：上传从文件
download：下载文件
delete：删除文件
append：在appender文件后追加内容
modify：appender文件修改
set_metadata：设置文件附加属性
get_metadata：获取文件附加属性
重启跟踪器：
/usr/bin/restart.sh /usr/bin/fdfs_trackerd /etc/fdfs/tracker.conf
重启存储器
/usr/bin/restart.sh /usr/bin/fdfs_storaged /etc/fdfs/storage.conf
参数形式如下：
volume0/M00/00/02/Cs8b8lFJIIyAH841AAAbpQt7xVI4715674
volume0：组名
M00：磁盘名
00/02：目录
Cs8b8lFJIIyAH841AAAbpQt7xVI4715674：文件名，采用base64编码，信息包含源storage server Ip、文件创建时间、文件大小、文件CRC32效验码和随机数