Glusterfs Build and install

Glusterfs is mainly used in cluster system, and it has good expansibility. The structure of the software is well designed, easy to expand and configure, and the flexible collocation of each module to get the specific solution. Addresses the following issues: networked storage, federated storage (fusing storage space on multiple nodes), redundant backups, large file load Balancing (chunking).

Due to the lack of key features, reliability has not been tested for a long time and is not suitable for use in a product environment that requires 24 hours of uninterrupted service. Currently suitable for off-line applications for large data volumes, here's a look at the installation configuration of Glusterfs Distributed File system

Glusterfs is an open-source Distributed file system that allows users to use multiple servers and interconnect via Ethernet or InfiniBand RDMA to form a glusterfs cluster

。

The Glusterfs cluster provides an interface to Nfs,cifs and Gluster Native (mounted via fuse) for users to access the Glusterfs storage pool.

Glusterfs uses an elastic hashing algorithm to locate the location of the file store. Because of the elastic hashing algorithm, Glusterfs does not require a dedicated Meta-data server to hold the metadata, so you can avoid the entire cluster being unavailable because of a metadata server outage.

Because the metadata server is not needed, the overall performance of the glusterfs is outstanding when the data is read and written at multiple mount points.

Environmental centos6.2

Fuse-2.9.3.tar.gz #依赖于fuse

Glusterfs-3.6.7.tar.gz #本文用的版本

Prepare two hosts

192.168.5.110 G1

192.168.5.111 G2

And do IP mapping in the hosts

To start the installation installation environment:

Install on two machines, respectively

[email protected] ~]# Yum install-y gcc gcc-c++ flex flex-devel bison bison-devel OpenSSL openssl-devel libxml2 libxml2 -devel

Compiling and installing Fuse,glusterfs

CD fuse-2.9.3 &&/configure && make && make INSTALLCD glusterfs-3.6.7 &&/configure--PR Efix=/usr/local/glusterfs && make && make install

Both G1 and G2 perform the above operations

G1 and G2 start the Gluster:

[Email protected] ~]#/etc/init.d/glusterd restart

Make a command soft connect:

[Email protected] ~]# ln-s/usr/local/glusterfs/sbin/gluster/usr/bin/gluster

Execute on the G2

Gluster peer probe G1 adds G1 to the Gluster cluster, and the native (G2) does not need to be joined.

To view cluster information:

[Email protected] ~]# Gluster peer status

Number of Peers:1

Hostname:g1

uuid:32338373-972e-4c64-9489-548710070c5a

State:peer in Cluster (Connected)

Reject machine:

Gluster Peer Detach G1

To create a volume:

Gluster Volume Create Test-volume Replica 2 transport TCP G1:/data G2:/data force

To view a volume:

Gluster Volume Info

Client 1 mount (also install fuse and glusterfs to support Glusterfs file system):

Same hosts under want IP map

Mkdir/mnt/gfs

Mount-t Glusterfs g1:test-volume/mnt/gfs/

Df-h can view mounts after successful mount

Http://www.poluoluo.com/server/201502/338951.html

This article is from the "Fluffy Duck" blog, please be sure to keep this source http://woshitieren.blog.51cto.com/2466034/1731730

Glusterfs Build and install