FastDFS FAQ （欢迎反馈，我将及时整理）

uvain

回复 139# duoduoluo_z


    你这group的问题解决了么，我也同样的问题，没想明白

duoduoluo_z

什么问题啊？我记得就是注释掉echo 命令，好像就ok了。。。
回复 141# uvain


   

开源大将军

来看看分布式文件系统学习资料，学习中

mm8m6m7m

FastDFS中Trunk文件使用 出现新数据和旧数据无法同步的现象。请问可以手动同步吗？可以的话能不能提供一下操作步骤。谢谢！回复 1# happy_fish100


   

liupanhu

V5.05 版本 压测访问可以达到3000并发，上传1000并发就报错。

客户端，连接不是及时关闭。
# ss -4 |grep :23000|awk '{print $1}'|sort |uniq -c|sort -rn|more
  61858 ESTAB
  58594 CLOSE-WAIT


# cat trackerd.log

[2016-07-22 22:48:47] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:10] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:11] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:12] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0
[2016-07-22 22:59:12] ERROR - file: tracker_nio.c, line: 213, client ip: 192.168.5.24, recv timeout, recv offset: 0, expect length: 0


# cat storaged.log

[2016-07-22 22:59:16] ERROR - file: tracker_proto.c, line: 37, server: 192.168.2.125:23000, recv data fail, errno: 107, error info: Transport endpoint is not connected
[2016-07-22 22:59:17] ERROR - file: storage_sync.c, line: 2710, connect to storage server 192.168.2.125:23000 fail, errno: 111, error info: Connection refused

# cat fastdfs_client.log

[2016-07-22 22:59:16] ERROR - file: ../tracker/tracker_proto.c, line: 37, server: 192.168.2.125:23000, recv data fail, errno: 107, error info: Transport endpoint is not connected
[2016-07-22 22:59:16] ERROR - file: ../tracker/tracker_proto.c, line: 37, server: 192.168.2.125:23000, recv data fail, errno: 107, error info: Transport endpoint is not connected
[2016-07-22 22:59:16] ERROR - file: ../tracker/tracker_proto.c, line: 37, server: 192.168.2.125:23000, recv data fail, errno: 107, error info: Transport endpoint is not connected
[2016-07-22 22:59:16] ERROR - file: connection_pool.c, line: 130, connect to 192.168.2.125:23000 fail, errno: 111, error info: Connection refused
[2016-07-22 22:59:16] ERROR - file: connection_pool.c, line: 130, connect to 192.168.2.125:23000 fail, errno: 111, error info: Connection refused
[2016-07-22 22:59:16] ERROR - file: connection_pool.c, line: 130, connect to 192.168.2.125:23000 fail, errno: 111, error info: Connection refused
[2016-07-22 22:59:16] ERROR - file: ../tracker/tracker_proto.c, line: 37, server: 192.168.2.125:23000, recv data fail, errno: 107, error info: Transport endpoint is not connected

305835227

@happy_fish100   鱼大大好，想问下我使用了连接池（java client），并且tracker.conf和storage.conf都启动了连接池（use_connection_pool = true），为何在我的机器上（java client）看到和trakcer建立的连接很少很少，我设置的连接池是1000个，fdfs版本5.08 服务器系统为CentOS6.5 64bit

gbzdyf

回复 126# haoyunkgq

你好，，，我的跟你是同样的问题，，，也是在另外一个机房单独增加一个storage，这台storage是内网机器，没外网，做了NAT上网，
给机器映射了23000端口，在tracker上也添加了id配置，启动storage报错,,,,你的怎样解决的，，，

forbreak

鱼大：你好，我在使用 append接口上传的文件时，fastDHT 不会对这种类型的文件进行去重?  有没有修改文件标识的接口，我先传一个文件为可以append的文件，等我把后续的文件都上传完成的时候，在改变文件类型为 普通文件。这样fastDHT 在对它进行去重？

雨漠夏末

跪求解决啊，用fastdfs一段时间了，http访问资源（tracker访问方式）在一个时间点突然不行了，在此之前上传的文件通过tracker都访问不了，访问通过日志看就会说找不到文件，非tracker依然可以访问，奇怪的是重新上传的文件通过tracker访问依然是可以的。 fastdfs  http解决方案用的是apache，两台跟踪，三台存储（两个分组），我把配置贴出来，老大看看是怎么回事啊。
tracker机器上的http扩展配置：
两台跟踪 10.12.5.54、10.12.5.57除了主机IP，http配置都是一样的：
LoadModule fastdfs_module modules/mod_fastdfs.so
<VirtualHost *:80>
Servername 10.12.5.54
<Location /group1/M00>
sethandler fastdfs
</Location>
<Location /group2/M00>
sethandler fastdfs
</Location>
</VirtualHost>

三台存储除了ip 以及配置分组都是一样的：
LoadModule fastdfs_module modules/mod_fastdfs.so
<VirtualHost *:80>
        Servername 10.12.5.55
        <Location /group1/M00>
                sethandler fastdfs
        </Location>
</VirtualHost>

雨漠夏末

fastdfs http扩展，无论nginx还是apache，刚上传的文件使用tracker访问可以正常访问，文件上传大概一天多，使用tracker访问方式就不行了，非tracker访问还是可以的，是什么问题啊鱼大大