MOOSEFS 安装简易向导--Installing MooseFS Step by Step Tutorial

Core Technology Development & Support Team

翻译:田逸(sery@163.com)

blob.png 

 

January 7, 2017

© 2013-2017 v. 1.5.1 Piotr Robert Konopelko, Core Technology Development & Support Team. All rights reserved.

 

Proofread by Agata Kruszona-Zawadzka Coordination & layout by Piotr Robert Konopelko. Please send corrections to peter@mfs.io.

 

 

 

 

 

Contents—目录

1 Introduction 3

1.1 1.6.2x 与 2.0.x版本间的主要差异 . . . . . . . . . . . . . . . . . . 3

1.2 多主服务器 – 工作方式? . . . . . . . . . . . . . . . . . . . . . . 3

2 安装前的准备工作………………………………………………………….. 5

 2.1 配置域名服务 . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.2 添加源码库 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

 2.2.1 源码库分支 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.3 MooseFS 与 MooseFS Pro 软件包命名差异  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

3 MooseFS 在专用机器/系统上的安装处理…………………………………………………7

 3.1 安装Master Server(s)  . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

 3.2 MooseFS CGI and CGI Server 安装 . . . . . . . . . . . . . . . . . . . . . . 9

3.3 MooseFS CLI 安装 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.4 Metadata backup servers (Metaloggers) 安装 . . . . . . . . . . . . . . . . . 11

 3.5 Chunkservers 安装 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.6 客户端安装 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

4 MooseFS 基本用法………………………………………………. 15

5停止 MooseFS 17

6附录: Setting up DNS server on Debian/Ubuntu 18

6.1 Setting up DNS server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

 6.2 Setting up revDNS server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

 

 

 

 

 

 

 

 

 

 

 

第一章

 介绍

 

Notice: there is one dependency to resolve: users’ computers need FUSE package to mount MooseFS. It can be downloaded and installed from repositories.

注意:需要解决一个依赖关系:客户端挂载moosfs时,需要FUSE(用户空间文件系统)支持。该软件包能够从源码库下载及安装。

 

1.1 Key differences between versions 1.6.2x and 2.0.x

Moosefs 1.6.2X与版本2.0.x版本之间的主要差异

1.     Master host(s) configuration is done solely via DNS – it is no longer possible to list master(s) IP address(es) in clients’ and Chunkservers’ configuration; default name for master domain is mfsmaster, it can be changed in configuration files;

主控制系统命名配置是单独通过dns来完成—它不再出现在客户端及存储节点端的配置文件之中。主控制器缺省命名为mfsmaster,它也可以在配置文件里进行更改。

2. In Pro version Metaloggers become optional, they can be replaced by additional Master Servers; in Community Edition it is still strongly recommended to set up Metaloggers.

专业版本(付费)日志服务器为可选项,它们可以被附加的主控服务器所替代;社区免费版本还是建议强烈配备日志服务器Metaloggers.

3. Mfsmetarestore tool is no longer present in the system; instead, it is enough to start the master process with -a switch;

恢复工具Mfsmetarestore不再提供,由master带选项-a 来代替。

4. Configuration files now sit in mfs subdirectory inside the /etc directory (this change was introduced in 1.6.27).

配置文件路径统一为/etc/mfs(从版本1.6.27开始)

1.2   Many Master Servers – how does it work?

多主控服务器—如何工作?

In previous MooseFS versions you had only one master process and any number of Metaloggers. In the event of master failure, system administrator was able to retrieve ”metadata” information from the Metalogger and start a new master (on a new machine, if necessary), so the file system was up and running again. But this was always causing the system to be unavailable to clients for a period of time and required manual work to bring it back up.

在以前的版本中,只有一个主控服务器(Master)及多个日志服务器(metalogger).一旦主控服务器发生异常/故障,系统管理员只能从日志服务器(metalogger)检索信息,启动一个新的主控服务器并进行数据恢复(如果有必要的话)。但这总会造成一段时间的服务挂起,并且需要手工从备份数据中进行恢复。

New MooseFS Pro version introduces many Master Servers working together in multiple roles. One role is ”leader”. The Leader Master is acting as it used to for the Chunkservers and clients. There is never more than one leader in any working system.

新的专业版本的Moosefs支持多个主控服务器协同工作,这些主控服务器可以同时充当多种角色。其中一种角色称为“领导者”,它只为存储服务器(chunkserver)及客户端提供服务,这个角色在整个系统里,只能有唯一的一个。

The other role is ”follower”. The follower master is doing what Metaloggers used to do – it downloads metadata from the leader master and keeps it. But unlike a Metalogger, if a leader master stops working, a follower master is immediately ready to take on the role of leader. If the leader master fails, a new candidate for leader is chosen from the followers. The candidate assumes a role of ”elect”, that automatically converts to ”leader” as soon as more than half of the Chunkservers connect to elect. There can be more than one follower in the system

另外一个角色是“追随者”。这些“追随者”服务器类似日志服务器(metalogger),从领导者服务器(leader)下载元数据并保存下来。但它与日志服务器的区别在于,一旦主控“领导者”服务器停止服务,某个“追随者”主控服务器能立即准备接替“领导者leader”的任务。如果“领导者”主控服务器故障或者异常,一个新的“领导者”服务器将从“追随者”主控服务器中产生。候选资格可以假定为一个选举角色,只要超过一半数量的数据存储节点(chunkserver)选举某个“追随者”主控服务器,该追随者就立马切换为“领导者”主控服务器。

The whole switching operation is almost invisible to the system users, as it usually takes between a couple to a dozen or so seconds. When/if the former leader master starts working again, it assumes the role of follower. If a follower master fails, it has no effect on the whole system. If such a master starts working again, it again assumes the role of follower.

整个“领导者”切换过程对于系统用户来说,几乎是不可见的,因为它花费的时间就是那么一瞬间而已。假如前“领导者”主控服务器得以恢复,假如到系统中来,那么刚接替称为“领导者”的那个服务器自动交权,降级称为原来的角色“追随者”主控服务器。如果其中的某个“追随者”主控服务器在运行中发生故障,它不对整个服务造成影响。恢复服务以后,它可以继续作为“追随者”存在于系统之中。


继续翻译中......