虚拟机搭建Lustre文件系统

目标

使用虚拟机搭建Lustre系统,体验学习Lustre文件系统。具体来说,该文件系统有四个节点,一个mds节点,两个oss节点,一个client节点;下图展示了该系统的全貌。

背景

  • 虚拟机镜像使用 CentOS-7-x86_64-DVD-1511.iso,该镜像的官方下载地址请点击 这里
  • 使用VMware ESXi虚拟化平台管理虚拟机
  • 每个虚拟机分配两块磁盘,一块用于操作系统的安装,另一块用于Lustre文件系统存储
  • 需要同时起五个虚拟机(有一个用于本地repo功能),所以尽量选用一个高性能的宿主机
  • 支持LDISKFS和ZFS

注意

使用VMware Workstation,或者Virtual Box等虚拟机管理软件也可以。本文以VMware ESXi为例做演示。

安装配置虚拟机

首先完成一台虚拟机的安装,然后在该虚拟机上针对Lustre系统做一些基础配置,最后关闭该虚拟机,使用VMware ESXi的 导出OVF模板 功能(其他虚拟机管理软件也有类似功能)导出该虚拟机的示例模板。最终利用虚拟机模板再创建四个虚拟机,用于构建Lustre文件系统。

虚拟机安装

根据宿主机的性能安排虚拟机的硬件配置,不确定的配置使用默认值即可;需要注意的是给虚拟机分配两块磁盘,在本例中,系统盘分配20G,Lustre存储盘分配16G。

加载 CentOS-7-x86_64-DVD-1511.iso 镜像,上电后选择 minimal 安装方式安装即可。

简单配置虚拟机

  • 查看Linux版本
[root@repo ~]# uname -a
Linux repo 3.10.0-693.11.6.el7_lustre.x86_64 #1 SMP Wed Jan 31 20:03:36 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
  • 安装一些基础包
yum groupinstall “Development Tools” -y
yum install epel-release quilt libselinux-devel python-docutils xmlto asciidoc elfutils-libelf-devel elfutils-devel zlib-devel rng-tools binutils-devel python-devel sg3_utils newt-devel perl-ExtUtils-Embed audit-libs-devel lsof hmaccalc -y

  • 关闭SELINUX(Lustre不支持SELINUX)
Set SELINUX=disabled in /etc/sysconfig/selinux
  • 关闭防火墙,或者添加iptables规则
systemctl stop firewalld.service
systemctl disable firewalld.service
# iptables -A IN_public_allow -m state --state NEW -m tcp -p tcp --dport 988 -j ACCEPT

导出虚拟机模板

正常关闭该虚拟机后,导出该虚拟机模板。

获得五个虚拟机

利用虚拟机模板,使用 部署OVF模板 功能创建四个虚拟机作为Lustre文件系统的四个节点。

使用 hostnamectl set-hostname mds/oss1/oss2/client1 命令设置这四个节点的主机名,便于分辨。起始的虚拟机作为repo主机(主机名设置为repo),为这四个节点提供Lustre repo功能。

构建本地Lustre repo库

该小节严格遵循 Using YUM to Manage Local Software Provisioning 的步骤来完成本地repo库的搭建。基本原理就是在repo主机上起一个web服务器,把从官方源上拷贝过来的Lustre安装包放到web服务目录中;mds/oss1/oss2/client1这四个节点安装Lustre时就可以直接使用repo主机的web服务提供的repo了,从而更方便地统一部署安装Lustre了。

创建repo定义

本例使用了2.10.3版本(注意,该版本有bug,建议使用最新版本),从 官方repo 中选择目标版本,将目标版本的URL分别填到对应的baseurl变量中,可以排除掉用于debug功能的rpm包从而减小安装。

[root@repo ~]# cat lustre-repo.conf
[lustre-server]
name=lustre-server
baseurl=https://downloads.hpdd.intel.com/public/lustre/lustre-2.10.3/el7/server
exclude=*debuginfo*
gpgcheck=0

[lustre-client]
name=lustre-client
baseurl=https://downloads.hpdd.intel.com/public/lustre/lustre-2.10.3/el7/client
exclude=*debuginfo*
gpgcheck=0

[e2fsprogs-wc]
name=e2fsprogs-wc
baseurl=https://downloads.hpdd.intel.com/public/e2fsprogs/latest/el7
exclude=*debuginfo*
gpgcheck=0

同步源到本地

使用 reposync 命令将官方源中的包同步到本地(repo主机),目标目录是 /var/www/html/repo,即Apache服务器的默认目录。

mkdir -p /var/www/html/repo
cd /var/www/html/repo
reposync -c ~/lustre-repo.conf -n \
-r lustre-server \
-r lustre-client \
-r e2fsprogs-wc

创建repo元数据

创建本地repo的元数据:

cd /var/www/html/repo
for i in e2fsprogs-wc lustre-client lustre-server; do
(cd $i && createrepo .)
done

创建repo访问文件

该文件定义了repo主机提供的本地源服务,其中 192.168.199.143 是repo主机的IP地址,确保Lustre文件系统的四个节点都可以访问到该地址。

[root@repo ~]# cat lustre.repo

[lustre-server]
name=lustre-server
baseurl=http://192.168.199.143/repo/lustre-server
enabled=0
gpgcheck=0
proxy=_none_

[lustre-client]
name=lustre-client
baseurl=http://192.168.199.143/repo/lustre-client
enabled=0
gpgcheck=0

[e2fsprogs-wc]
name=e2fsprogs-wc
baseurl=http://192.168.199.143/repo/e2fsprogs-wc
enabled=0
gpgcheck=0
  1. 在repo主机上 service httpd start 启动web服务,关闭防火墙或开放80端口,使web服务可用
  2. 将该 lustre.repo 文件拷贝到mds/oss1/oss2/client1这四个节点的 /etc/yum.repos.d/ 目录中
  3. 可在mds/oss1/oss2/client1四节点中使用 curl 命令验证本地repo的可用性

搭建Lustre

至此,准备工作完成,可以开始Lustre的安装了。

mds/oss1/oss2三节点安装

在mds/oss1/oss2这三个节点上执行以下命令,完成Lustre服务端的软件包安装和模块插入。

[root@mds ~]# yum --nogpgcheck --disablerepo=* --enablerepo=e2fsprogs-wc install e2fsprogs -y
[root@mds ~]# yum -y install epel-release
[root@mds ~]# yum install http://download.zfsonlinux.org/epel/zfs-release.el7_3.noarch.rpm
[root@mds ~]# yum upgrade linux-firmware dracut -y # 升级冲突包
[root@mds ~]# yum upgrade xfsprogs kmod kexec-tools -y # 升级冲突包
[root@mds ~]# yum --nogpgcheck --disablerepo=base,extras,updates --enablerepo=lustre-server install kernel kernel-devel kernel-headers kernel-tools kernel-tools-libs kernel-tools-libs-devel -y
[root@mds ~]# hid=`[ -f /etc/hostid ] && od -An -tx /etc/hostid|sed 's/ //g'` [ "$hid" = `hostid` ] || genhostid
[root@mds ~]# reboot
[root@mds ~]# yum --nogpgcheck --enablerepo=lustre-server install kmod-lustre-osd-ldiskfs lustre-dkms lustre-osd-ldiskfs-mount lustre-osd-zfs-mount lustre lustre-resource-agents zfs -y
[root@mds ~]# modprobe -v zfs
[root@mds ~]# modprobe -v lustre

具体插入的内核模块可参看以下输出:

[root@mds ~]# modprobe -v zfs
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/spl.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/znvpair.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/zcommon.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/icp.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/zavl.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/zunicode.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/zfs.ko.xz
[root@mds ~]# modprobe -v lustre
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/libcfs.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/lnet.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/obdclass.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/ptlrpc.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/fld.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/fid.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/lov.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/mdc.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/lmv.ko.xz
insmod /lib/modules/3.10.0-693.11.6.el7_lustre.x86_64/extra/lustre.ko.xz

troubleshooting

在上述命令中有对某些包的升级操作,主要原因是它们的版本太低,所以需要升级。

[root@mds ~]# yum --nogpgcheck --disablerepo=base,extras,updates --enablerepo=lustre-server install kernel kernel-devel kernel-headers kernel-tools kernel-tools-libs kernel-tools-libs-devel
Loaded plugins: fastestmirror
Loading mirror speeds from cached hostfile
* epel: mirrors.ustc.edu.cn
Package matching kernel-headers-3.10.0-693.11.6.el7_lustre.x86_64 already installed. Checking for update.
Resolving Dependencies
--> Running transaction check
---> Package kernel.x86_64 0:3.10.0-693.11.6.el7_lustre will be installed
---> Package kernel-devel.x86_64 0:3.10.0-693.11.6.el7_lustre will be installed
---> Package kernel-tools.x86_64 0:3.10.0-327.el7 will be updated
---> Package kernel-tools.x86_64 0:3.10.0-693.11.6.el7_lustre will be an update
---> Package kernel-tools-libs.x86_64 0:3.10.0-327.el7 will be updated
---> Package kernel-tools-libs.x86_64 0:3.10.0-693.11.6.el7_lustre will be an update
---> Package kernel-tools-libs-devel.x86_64 0:3.10.0-693.11.6.el7_lustre will be installed
--> Processing Conflict: kernel-3.10.0-693.11.6.el7_lustre.x86_64 conflicts xfsprogs  Processing Conflict: kernel-3.10.0-693.11.6.el7_lustre.x86_64 conflicts kmod  Processing Conflict: kernel-3.10.0-693.11.6.el7_lustre.x86_64 conflicts kexec-tools  Finished Dependency Resolution
Error: kernel conflicts with xfsprogs-3.2.2-2.el7.x86_64
Error: kernel conflicts with kexec-tools-2.0.7-38.el7.x86_64
Error: kernel conflicts with kmod-20-5.el7.x86_64
You could try using --skip-broken to work around the problem
You could try running: rpm -Va --nofiles --nodigest
[root@mds ~]# rpm -qa |grep xfsprogs
xfsprogs-3.2.2-2.el7.x86_64

诸如此类的:

[root@mds ~]# yum --nogpgcheck --disablerepo=base,extras,updates \
> --enablerepo=lustre-server install \
> kernel \
> kernel-devel \
> kernel-headers \
> kernel-tools \
> kernel-tools-libs \
> kernel-tools-libs-devel
Loaded plugins: fastestmirror
lustre-server | 2.9 kB 00:00:00
zfs | 2.9 kB 00:00:00
(1/2): lustre-server/primary_db | 499 kB 00:00:00
(2/2): zfs/x86_64/primary_db | 86 kB 00:00:02
Loading mirror speeds from cached hostfile
* epel: mirrors.aliyun.com
Package matching kernel-headers-3.10.0-693.11.6.el7_lustre.x86_64 already installed. Checking for update.
Resolving Dependencies
--> Running transaction check
---> Package kernel.x86_64 0:3.10.0-693.11.6.el7_lustre will be installed
--> Processing Dependency: linux-firmware >= 20170606-55 for package: kernel-3.10.0-693.11.6.el7_lustre.x86_64
--> Processing Dependency: dracut >= 033-502 for package: kernel-3.10.0-693.11.6.el7_lustre.x86_64
---> Package kernel-devel.x86_64 0:3.10.0-693.11.6.el7_lustre will be installed
---> Package kernel-tools.x86_64 0:3.10.0-327.el7 will be updated
---> Package kernel-tools.x86_64 0:3.10.0-693.11.6.el7_lustre will be an update
---> Package kernel-tools-libs.x86_64 0:3.10.0-327.el7 will be updated
---> Package kernel-tools-libs.x86_64 0:3.10.0-693.11.6.el7_lustre will be an update
---> Package kernel-tools-libs-devel.x86_64 0:3.10.0-693.11.6.el7_lustre will be installed
--> Processing Conflict: kernel-3.10.0-693.11.6.el7_lustre.x86_64 conflicts xfsprogs  Processing Conflict: kernel-3.10.0-693.11.6.el7_lustre.x86_64 conflicts kmod  Processing Conflict: kernel-3.10.0-693.11.6.el7_lustre.x86_64 conflicts kexec-tools  Finished Dependency Resolution
Error: Package: kernel-3.10.0-693.11.6.el7_lustre.x86_64 (lustre-server)
Requires: linux-firmware >= 20170606-55
Installed: linux-firmware-20150904-43.git6ebf5d5.el7.noarch (@anaconda)
linux-firmware = 20150904-43.git6ebf5d5.el7
Error: Package: kernel-3.10.0-693.11.6.el7_lustre.x86_64 (lustre-server)
Requires: dracut >= 033-502
Installed: dracut-033-359.el7.x86_64 (@anaconda)
dracut = 033-359.el7
Error: kernel conflicts with xfsprogs-3.2.2-2.el7.x86_64
Error: kernel conflicts with kexec-tools-2.0.7-38.el7.x86_64
Error: kernel conflicts with kmod-20-5.el7.x86_64
You could try using --skip-broken to work around the problem
You could try running: rpm -Va --nofiles --nodigest
[root@mds ~]#

client1节点安装

在client1节点执行以下命令,完成Lustre客户端的软件包安装和模块插入。

[root@client1 ~]# yum install kernel kernel-devel kernel-headers kernel-abi-whitelists kernel-tools kernel-tools-libs kernel-tools-libs-devel -y
[root@client1 ~]# reboot
[root@client1 ~]# yum install epel-release
[root@client1 ~]# yum --nogpgcheck --enablerepo=lustre-client install lustre-client-dkms lustre-client -y
[root@client1 ~]# modprobe -v lustre
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/libcfs.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/lnet.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/obdclass.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/ptlrpc.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/lov.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/fid.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/mdc.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/fld.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/lmv.ko.xz
insmod /lib/modules/3.10.0-862.3.2.el7.x86_64/extra/lustre.ko.xz

配置mds

终于可以配置mds了,将之前添加的第二块盘(/dev/sdb)作为Lustre的存储盘,格式化为lustre格式。

[root@mds ~]# mkfs.lustre --fsname=lustrefs --mgs --mdt --index=0 /dev/sdb

Permanent disk data:
Target: lustrefs:MDT0000
Index: 0
Lustre FS: lustrefs
Mount type: ldiskfs
Flags: 0x65
(MDT MGS first_time update )
Persistent mount opts: user_xattr,errors=remount-ro
Parameters:

checking for existing Lustre data: not found
device size = 16384MB
formatting backing filesystem ldiskfs on /dev/sdb
target name lustrefs:MDT0000
4k blocks 4194304
options -J size=655 -I 1024 -i 2560 -q -O dirdata,uninit_bg,^extents,dir_nlink,quota,huge_file,flex_bg -E lazy_journal_init -F
mkfs_cmd = mke2fs -j -b 4096 -L lustrefs:MDT0000 -J size=655 -I 1024 -i 2560 -q -O dirdata,uninit_bg,^extents,dir_nlink,quota,huge_file,flex_bg -E lazy_journal_init -F /dev/sdb 4194304
Writing CONFIGS/mountdata

最后创建挂载点,挂载该盘。

[root@mds ~]# mkdir /mnt/mdt
[root@mds ~]# mount -t lustre /dev/sdb /mnt/mdt
mount.lustre: increased /sys/block/sdb/queue/max_sectors_kb from 512 to 4096

配置oss1

在oss1节点将第二块盘作为ost类型的存储盘,其中 192.168.199.138 为mds节点的IP。

[root@oss1 ~]# mkfs.lustre --fsname=lustrefs --mgsnode=192.168.199.138@tcp0 --ost --index=0 /dev/sdb

Permanent disk data:
Target: lustrefs:OST0000
Index: 0
Lustre FS: lustrefs
Mount type: ldiskfs
Flags: 0x62
(OST first_time update )
Persistent mount opts: ,errors=remount-ro
Parameters: mgsnode=192.168.199.138@tcp

checking for existing Lustre data: not found
device size = 16384MB
formatting backing filesystem ldiskfs on /dev/sdb
target name lustrefs:OST0000
4k blocks 4194304
options -J size=400 -I 512 -i 69905 -q -O extents,uninit_bg,dir_nlink,quota,huge_file,flex_bg -G 256 -E resize="4290772992",lazy_journal_init -F
mkfs_cmd = mke2fs -j -b 4096 -L lustrefs:OST0000 -J size=400 -I 512 -i 69905 -q -O extents,uninit_bg,dir_nlink,quota,huge_file,flex_bg -G 256 -E resize="4290772992",lazy_journal_init -F /dev/sdb 4194304
Writing CONFIGS/mountdata
[root@oss1 ~]#

最后创建挂载点,挂载sdb。

[root@oss1 ~]# mkdir /mnt/ost0
[root@oss1 ~]# mount -t lustre /dev/sdb /mnt/ost0/
mount.lustre: increased /sys/block/sdb/queue/max_sectors_kb from 512 to 4096

配置oss2

同理,oss2的配置与oss1一样。

[root@oss2 ~]# mkfs.lustre --fsname=lustrefs --mgsnode=192.168.199.138@tcp0 --ost --index=1 /dev/sdb

Permanent disk data:
Target: lustrefs:OST0001
Index: 1
Lustre FS: lustrefs
Mount type: ldiskfs
Flags: 0x62
(OST first_time update )
Persistent mount opts: ,errors=remount-ro
Parameters: mgsnode=192.168.199.138@tcp

checking for existing Lustre data: not found
device size = 16384MB
formatting backing filesystem ldiskfs on /dev/sdb
target name lustrefs:OST0001
4k blocks 4194304
options -J size=400 -I 512 -i 69905 -q -O extents,uninit_bg,dir_nlink,quota,huge_file,flex_bg -G 256 -E resize="4290772992",lazy_journal_init -F
mkfs_cmd = mke2fs -j -b 4096 -L lustrefs:OST0001 -J size=400 -I 512 -i 69905 -q -O extents,uninit_bg,dir_nlink,quota,huge_file,flex_bg -G 256 -E resize="4290772992",lazy_journal_init -F /dev/sdb 4194304
Writing CONFIGS/mountdata
[root@oss2 ~]# mkdir /mnt/ost1
[root@oss2 ~]# mount -t lustre /dev/sdb /mnt/ost1
mount.lustre: increased /sys/block/sdb/queue/max_sectors_kb from 512 to 4096
[root@oss2 ~]#

配置client1

在client1节点创建挂载点,挂载Lustre文件系统。

[root@client1 ~]# mkdir /lustre
[root@client1 ~]# mount -t lustre 192.168.199.138@tcp0:/lustrefs /lustre/

验证Lustre

在client1节点可以查看Lustre文件系统的所有磁盘情况。

[root@client1 ~]# lfs df -h
UUID bytes Used Available Use% Mounted on
lustrefs-MDT0000_UUID 8.9G 46.2M 8.0G 1% /lustre[MDT:0]
lustrefs-OST0000_UUID 15.4G 45.1M 14.5G 0% /lustre[OST:0]
lustrefs-OST0001_UUID 15.4G 45.1M 14.5G 0% /lustre[OST:1]

filesystem_summary: 30.7G 90.3M 29.0G 0% /lustre

[root@client1 ~]#

简单测试

试了一下写入速度,同样的128M写入10次,在client1节点本地可达到118MB/s,写入到Lustre文件系统速度在12.1MB/s;当然这肯定不能作为结论,因为毕竟只是一次在虚拟机上的尝试安装,没有考虑网络、磁盘和其他硬件性能等。

另外,不小心发现了Lustre 2.10.3版本的bug,从上图可以看出查看 /lustre 目录时报错了。查了一下,原来是个已知bug,并且在2.10.4上已经修复了。具体bug信息可以查看 RHEL/CentOS 7.5 prevent listing Lustre directory contents

关于DKMS

Using_DKMS 上提到需要安装DKMS的相关开发包(见下面的yum命令),才可以支持DKMS。

yum install \
asciidoc audit-libs-devel automake bc \
binutils-devel bison device-mapper-devel elfutils-devel \
elfutils-libelf-devel expect flex gcc gcc-c++ git \
glib2 glib2-devel hmaccalc keyutils-libs-devel krb5-devel ksh \
libattr-devel libblkid-devel libselinux-devel libtool \
libuuid-devel libyaml-devel lsscsi make ncurses-devel \
net-snmp-devel net-tools newt-devel numactl-devel \
parted patchutils pciutils-devel perl-ExtUtils-Embed \
pesign python-devel redhat-rpm-config rpm-build systemd-devel \
tcl tcl-devel tk tk-devel wget xmlto yum-utils zlib-devel

但本例中,没有执行上述命令,依然有DKMS存在,再具体了没有深入研究;具体输出可查看下面,mds/oss1/oss2输出相同,四节点都没有升级内核(可能与这有关?)。

mds输出

[root@mds ~]# dkms status;rpm -qa |grep dkms
lustre, 2.10.3, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
spl, 0.7.9, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
zfs, 0.7.9, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
dkms-2.6.1-1.el7.noarch
spl-dkms-0.7.9-1.el7_3.noarch
lustre-dkms-2.10.3-1.el7.noarch
zfs-dkms-0.7.9-1.el7_3.noarch
[root@mds ~]#

oss1输出

[root@oss1 ~]# dkms status;rpm -qa |grep dkms
lustre, 2.10.3, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
spl, 0.7.9, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
zfs, 0.7.9, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
dkms-2.6.1-1.el7.noarch
spl-dkms-0.7.9-1.el7_3.noarch
lustre-dkms-2.10.3-1.el7.noarch
zfs-dkms-0.7.9-1.el7_3.noarch

oss2输出

[root@oss2 ~]# dkms status;rpm -qa |grep dkms
lustre, 2.10.3, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
spl, 0.7.9, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
zfs, 0.7.9, 3.10.0-693.11.6.el7_lustre.x86_64, x86_64: installed
dkms-2.6.1-1.el7.noarch
spl-dkms-0.7.9-1.el7_3.noarch
lustre-dkms-2.10.3-1.el7.noarch
zfs-dkms-0.7.9-1.el7_3.noarch

client1输出

[root@client1 ~]# dkms status;rpm -qa |grep dkms
lustre-client, 2.10.3, 3.10.0-862.3.2.el7.x86_64, x86_64: installed
dkms-2.6.1-1.el7.noarch
lustre-client-dkms-2.10.3-1.el7.noarch

参考资料

  1. Creating a Lustre Test System from Source with Virtual Machines : 讲解了虚拟机源码安装Lustre的简要步骤,虽然使用的Lustre、Linux版本较低,但参考价值很大
  2. Installing the Lustre Software : 很全面很具体的安装部署文档,参考价值很大
  3. Lustre* Software Release 2.x Manual : 官方手册,大而全,可翻阅
Advertisements

分类:默认目录

Tagged as: , , , ,

1 reply

发表评论

Fill in your details below or click an icon to log in:

WordPress.com 徽标

You are commenting using your WordPress.com account. Log Out /  更改 )

Google+ photo

You are commenting using your Google+ account. Log Out /  更改 )

Twitter picture

You are commenting using your Twitter account. Log Out /  更改 )

Facebook photo

You are commenting using your Facebook account. Log Out /  更改 )

w

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.