Showing posts with label arch. Show all posts
Showing posts with label arch. Show all posts

Saturday, November 7, 2015

利用arch配置一个pxe server,网络安装各种版本linux系统,各种版本windows系统!

想了半天也没有找到合适的板块发这个,最后还是发这里把!

最近折腾了两个小主机,华硕EeeBox B202,atom n270的u,1g的内存,80g的硬盘,机器挺小巧,就是单核u,功能实在鸡肋,索性一个做了个播放机,这个就做了一个pxe server!

利用wol的功能,网络装系统,至于pxe的原理,如下:


以下是摘自syslinux官网关于pxelinux的介绍,懒得翻译了,大家自己看,应该不难理解.
The basic configuration is the same for all Syslinux variants. This document explains only some of the differences specifically applicable to PXELINUX.
On the TFTP server, create the directory "/tftpboot", and copy "pxelinux.0" (from the Syslinux distribution) and any kernel or initrd images that you want to boot.
[5.00+] Also copy "ldlinux.c32" from the Syslinux distribution to the "/tftpboot" directory on the TFTP server.
Finally, create the directory "/tftpboot/pxelinux.cfg". The configuration file (equivalent of syslinux.cfg -- see the SYSLINUX FAQ for the options here) will live in this

directory.
Because more than one system may be booted from the same server, the configuration file name depends on the IP address of the booting machine.
Before a generic explanation, let's see first an example. When:
the bootloader file name is "/mybootdir/pxelinux.0"; and,
the client UUID is "b8945908-d6a6-41a9-611d-74a6ab80b83d"; and,
the Ethernet MAC address is "88:99:AA:BB:CC:DD"; and,
the IP address is "192.168.2.91", or in uppercase hexadecimal, "C0A8025B";
then PXELINUX will try the following configuration files (in this order):
/mybootdir/pxelinux.cfg/b8945908-d6a6-41a9-611d-74a6ab80b83d
/mybootdir/pxelinux.cfg/01-88-99-aa-bb-cc-dd
/mybootdir/pxelinux.cfg/C0A8025B
/mybootdir/pxelinux.cfg/C0A8025
/mybootdir/pxelinux.cfg/C0A802
/mybootdir/pxelinux.cfg/C0A80
/mybootdir/pxelinux.cfg/C0A8
/mybootdir/pxelinux.cfg/C0A
/mybootdir/pxelinux.cfg/C0
/mybootdir/pxelinux.cfg/C
/mybootdir/pxelinux.cfg/default
Let's see now what exactly the above example represents.
After attempting the file as specified in the DHCP or hardcoded options, PXELINUX will probe the following paths, prefixed with "pxelinux.cfg/", under the initial Working

Directory.
The client UUID, if provided by the PXE stack.
Note that some BIOSes do not have a valid UUID, and it might end up reporting something like all 1's.
This value is represented in the standard UUID format using lowercase hexadecimal digits, e.g. "b8945908-d6a6-41a9-611d-74a6ab80b83d".
The hardware type (using its ARP type code) and address, all in lowercase hexadecimal with dash separators.
For example, for an Ethernet (ARP type "1") with address "88:99:AA:BB:CC:DD", it would search for the filename "01-88-99-aa-bb-cc-dd".
The client's own IPv4 address in uppercase hexadecimal, followed by removing hex characters, one at a time, from the end. For example, "192.168.2.91" → "C0A8025B".
The included program, "gethostip", can be used to compute the hexadecimal IP address for any host.
Lowercase "default".


1:首先,需要的工具包有 dnsmasq,samba, nfs-utils,openssh,syslinux

dnsmasq   =    充当dhcp和tftpserver
samba      =    安装windows使用,另外也为了方便windows机器管理
nfs-utils    =    加载linux镜像用
openssh    =    管理用,ssh登录很方便,毕竟这台机器是个小的服务器吧
syslinux     =   启动pxelinux以及各种启动程序

arch下很简单,一条命令

pacman -S dnsmasq samba nfs-utils openssh syslinux

2:然后就是配置dhcp和tftp server,以及nfs 和 ssh server了


直接贴配置文件比较省事
dnsmasq为了配置方便,我在/etc/dnsmasq.d/重建了一个pxe.conf的配置文件,配置如下

记住要把/etc/dnsmasq.conf里面下面这句话取消注释


# Include another lot of configuration options.
#conf-file=/etc/dnsmasq.more.conf
conf-dir=/etc/dnsmasq.d


#     function as a tftp server 启动tftpserver
enable-tftp
#     root directory for files available via ftp. ftp的root环境
tftp-root=/tftpboot
#     the boot filename
#pxe-service=X86PC,"PXE Boot Menu",pxelinux 启动文件
dhcp-boot=/tftpboot/pxelinux.0
#     kill multicast
#dhcp-option=vendor:PXEClient,6,2b
#     disable reuse of the dhcp servername and filename fields as extra.
#     option space. to avoid confusing some old or broken dhcp clients.
dhcp-no-override
#     This range(s) is for the public interface, where dnsmasq functions
#     as a proxy DHCP server providing boot information but no IP leases.
#     Any ip in the subnet will do, so you may just put your server NIC ip here.
#     I used my router ip address
#dhcp-range=192.168.2.1,proxy,255.255.255.0

#dhcp-range=192.168.2.100,proxy
# do not give IPs that are in pool of DSL routers DHCP 我避开了家里路由器的dhcp缺省池
dhcp-range=192.168.2.151,192.168.2.200,12h
#     router/gateway ip address
dhcp-option=3,192.168.2.1
#     dns server ip address
dhcp-option=6,192.168.2.1
#     enable logging
log-dhcp
log-queries


samba的配置文件smb.conf
在windows下直接在文件浏览器中打入\\192.168.2.19,就可以打开tftpboot文件夹,直接管理里面的文件


# This is the main Samba configuration file. You should read the
# smb.conf(5) manual page in order to understand the options listed
# here. Samba has a huge number of configurable options (perhaps too
# many!) most of which are not shown in this example
#
# Any line which starts with a ; (semi-colon) or a # (hash)
# is a comment and is ignored. In this example we will use a #
# for commentry and a ; for parts of the config file that you
# may wish to enable
#
# NOTE: Whenever you modify this file you should run the command "testparm"
# to check that you have not made any basic syntactic errors.
#
#======================= Global Settings =====================================
[global]

# workgroup = NT-Domain-Name or Workgroup-Name
   workgroup = Workgroup

# server string is the equivalent of the NT Description field
   server string = Samba Server

# This option is important for security. It allows you to restrict
# connections to machines which are on your local network. The
# following example restricts access to two C class networks and
# the "loopback" interface. For more examples of the syntax see
# the smb.conf man page
;   hosts allow = 192.168.1. 192.168.2. 127.

# if you want to automatically load your printer list rather
# than setting them up individually then you'll need this
   printcap name = /etc/printcap
   load printers = yes

# It should not be necessary to spell out the print system type unless
# yours is non-standard. Currently supported print systems include:
# bsd, sysv, plp, lprng, aix, hpux, qnx
;   printing = bsd

# Uncomment this if you want a guest account, you must add this to /etc/passwd
# otherwise the user "nobody" is used
;  guest account = pcguest

# this tells Samba to use a separate log file for each machine
# that connects
   log file = /var/log/samba/%m.log

# Put a capping on the size of the log files (in Kb).
   max log size = 50

# Security mode. Most people will want user level security. See
# security_level.txt for details.
   security = user
#   security = share
# Use password server option only with security = server
;   password server =

# Password Level allows matching of _n_ characters of the password for
# all combinations of upper and lower case.
;  username level = 8

# You may wish to use password encryption. Please read
# ENCRYPTION.txt, Win95.txt and WinNT.txt in the Samba documentation.
# Do not enable this option unless you have read those documents
;  encrypt passwords = yes
;  smb passwd file = /etc/samba/smbpasswd

# The following are needed to allow password changing from Windows to
# update the Linux sytsem password also.
# NOTE: Use these with 'encrypt passwords' and 'smb passwd file' above.
# NOTE2: You do NOT need these to allow workstations to change only
#        the encrypted SMB passwords. They allow the Unix password
#        to be kept in sync with the SMB password.
;  unix password sync = Yes
;  passwd program = /usr/bin/passwd %u
;  passwd chat = *New*UNIX*password* %n\n *ReType*new*UNIX*password* %n\n *passwd:*all*authentication*tokens*updated*successfully*

# Unix users can map to different SMB User names
;  username map = /etc/samba/smbusers
# 开启 windows guest登录
   map to guest = Bad User

# Using the following line enables you to customise your configuration
# on a per machine basis. The %m gets replaced with the netbios name
# of the machine that is connecting
;   include = /etc/samba/smb.conf.%m

# Configure Samba to use multiple interfaces
# If you have multiple network interfaces then you must list them
# here. See the man page for details.
;   interfaces = 192.168.12.2/24 192.168.13.2/24

# Configure remote browse list synchronisation here
#  request announcement to, or browse list sync from:
# a specific host or from / to a whole subnet (see below)
;   remote browse sync = 192.168.3.25 192.168.5.255
# Cause this host to announce itself to local subnets here
;   remote announce = 192.168.1.255 192.168.2.44

# Browser Control Options:
# set local master to no if you don't want Samba to become a master
# browser on your network. Otherwise the normal election rules apply
;   local master = no

# OS Level determines the precedence of this server in master browser
# elections. The default value should be reasonable
;   os level = 33

# Domain Master specifies Samba to be the Domain Master Browser. This
# allows Samba to collate browse lists between subnets. Don't use this
# if you already have a Windows NT domain controller doing this job
;   domain master = yes
    domain master = auto

# Preferred Master causes Samba to force a local browser election on startup
# and gives it a slightly higher chance of winning the election
;   preferred master = yes

# Use only if you have an NT server on your network that has been
# configured at install time to be a primary domain controller.
;   domain controller =

# Enable this if you want Samba to be a domain logon server for
# Windows95 workstations.
;   domain logons = yes

# if you enable domain logons then you may want a per-machine or
# per user logon script
# run a specific logon batch file per workstation (machine)
;   logon script = %m.bat
# run a specific logon batch file per username
;   logon script = %U.bat

# Where to store roving profiles (only for Win95 and WinNT)
#        %L substitutes for this servers netbios name, %U is username
#        You must uncomment the [Profiles] share below
;   logon path = \\%L\Profiles\%U

# All NetBIOS names must be resolved to IP Addresses
# 'Name Resolve Order' allows the named resolution mechanism to be specified
# the default order is "host lmhosts wins bcast". "host" means use the unix
# system gethostbyname() function call that will use either /etc/hosts OR
# DNS or NIS depending on the settings of /etc/host.config, /etc/nsswitch.conf
# and the /etc/resolv.conf file. "host" therefore is system configuration
# dependant. This parameter is most often of use to prevent DNS lookups
# in order to resolve NetBIOS names to IP Addresses. Use with care!
# The example below excludes use of name resolution for machines that are NOT
# on the local network segment
# - OR - are not deliberately to be known via lmhosts or via WINS.
; name resolve order = wins lmhosts bcast

# Windows Internet Name Serving Support Section:
# WINS Support - Tells the NMBD component of Samba to enable it's WINS Server
;   wins support = yes

# WINS Server - Tells the NMBD components of Samba to be a WINS Client
# Note: Samba can be either a WINS Server, or a WINS Client, but NOT both
;   wins server = w.x.y.z

# WINS Proxy - Tells Samba to answer name resolution queries on
# behalf of a non WINS capable client, for this to work there must be
# at least one WINS Server on the network. The default is NO.
;   wins proxy = yes

# DNS Proxy - tells Samba whether or not to try to resolve NetBIOS names
# via DNS nslookups. The built-in default for versions 1.9.17 is yes,
# this has been changed in version 1.9.18 to no.
   dns proxy = no

# Case Preservation can be handy - system default is _no_
# NOTE: These can be set on a per share basis
;  preserve case = no
;  short preserve case = no
# Default case is normally upper case for all DOS files
;  default case = lower
# Be very careful with case sensitivity - it can break things!
;  case sensitive = no

#============================ Share Definitions ==============================
[homes]
   comment = Home Directories
   browseable = no
   writable = yes

# Un-comment the following and create the netlogon directory for Domain Logons
; [netlogon]
;   comment = Network Logon Service
;   path = /home/netlogon
;   guest ok = yes
;   writable = no
;   share modes = no


# Un-comment the following to provide a specific roving profile share
# the default is to use the user's home directory
;[Profiles]
;    path = /home/profiles
;    browseable = no
;    guest ok = yes


# NOTE: If you have a BSD-style print system there is no need to
# specifically define each individual printer
[printers]
   comment = All Printers
   path = /var/spool/samba
   browseable = no
# Set public = yes to allow user 'guest account' to print
   guest ok = no
   writable = no
   printable = yes

# This one is useful for people to share files
;[tmp]
;   comment = Temporary file space
;   path = /tmp
;   read only = no
;   public = yes

# A publicly accessible directory, but read only, except for people in
# the "staff" group
;[public]
;   comment = Public Stuff
;   path = /home/samba
;   public = yes
;   read only = yes
;   write list = @staff

# Other examples.
#
# A private printer, usable only by fred. Spool data will be placed in fred's
# home directory. Note that fred must have write access to the spool directory,
# wherever it is.
;[fredsprn]
;   comment = Fred's Printer
;   valid users = fred
;   path = /homes/fred
;   printer = freds_printer
;   public = no
;   writable = no
;   printable = yes

# A private directory, usable only by fred. Note that fred requires write
# access to the directory.
;[fredsdir]
;   comment = Fred's Service
;   path = /usr/somewhere/private
;   valid users = fred
;   public = no
;   writable = yes
;   printable = no

# a service which has a different directory for each machine that connects
# this allows you to tailor configurations to incoming machines. You could
# also use the %u option to tailor it by user name.
# The %m gets replaced with the machine name that is connecting.
;[pchome]
;  comment = PC Directories
;  path = /usr/pc/%m
;  public = no
;  writable = yes

# A publicly accessible directory, read/write to all users. Note that all files
# created in the directory by users will be owned by the default user, so
# any user with access can delete any other user's files. Obviously this
# directory must be writable by the default user. Another user could of course
# be specified, in which case all files would be owned by that user instead.
;[public]
;   path = /usr/somewhere/else/public
;   public = yes
;   only guest = yes
;   writable = yes
;   printable = no

# The following two entries demonstrate how to share a directory so that two
# users can place files there that will be owned by the specific users. In this
# setup, the directory should be writable by both users and should have the
# sticky bit set on it to prevent abuse. Obviously this could be extended to
# as many users as required.
;[myshare]
;   comment = Mary's and Fred's stuff
;   path = /usr/somewhere/shared
;   valid users = mary fred
;   public = no
;   writable = yes
;   printable = no
;   create mask = 0765

[tftpboot]
    comment = TFTP + PXE
    path = /tftpboot/
    public = yes
    only guest = yes
    browseable = yes
    read only = no
    writable = yes
    create mask = 0644
    directory mask = 0755
#    force create mask = 0644
#    force directory mask = 0755
    force user = root
    force group = root





nfs的配置文件 exports


# /etc/exports - exports(5) - directories exported to NFS clients
#
# Example for NFSv2 and NFSv3:
#  /srv/home        hostname1(rw,sync) hostname2(ro,sync)
# Example for NFSv4:
#  /srv/nfs4    hostname1(rw,sync,fsid=0)
#  /srv/nfs4/home   hostname1(rw,sync,nohide)
# Using Kerberos and integrity checking:
#  /srv/nfs4        *(rw,sync,sec=krb5i,fsid=0)
#  /srv/nfs4/home   *(rw,sync,sec=krb5i,nohide)
#
# Use `exportfs -arv` to reload.
#/tftpboot/images/ARCH_201510/ 192.168.2.0/24(ro,no_subtree_check)
/tftpboot/images/ARCH_201510/  *(ro,no_subtree_check,async,no_root_squash)
/tftpboot/images/kali2/         *(ro,no_subtree_check,async,no_root_squash)


3:接着配置tftpboot主目录

makdir /tftpboot/{images,iso,syslinux,pxelinux.cfg}

其中images文件夹是挂载的linux安装镜像和windows安装镜像

iso文件夹里是我各种安装iso文件的收集地点

syslinux文件夹里是syslinux各种模块的收藏地点

pxelinux.cfg目录里存放着pxelinux的启动配置文件,也就是目录文件

然后把各种pxelinux启动相关的文件,拷贝入tftpboot目录,和syslinux目录


cp /usr/lib/syslinux/bios/pxelinux.0  /tftpboot/
cp   /usr/lib/syslinux/bios/{chain.c32,hdt.c32,ldlinux.c32,libcom32.c32,libgpl.c32,libmenu.c32,libutil.c32,poweroff.c32,reboot.c32,vesamenu.c32} /tftpboot/syslinux/

其中
pxelinux.0是启动文件,hdt是用来显示机器的各种配置的,chain.c32是用来启动本地硬盘的,lib文件都是各种库,poweroff和reboot大家都懂的.

最后就是/tftpboot/pxelinux.cfg/default配置文件的内容了,这个内容是根据我的机器挂载的镜像修改的



UI syslinux/vesamenu.c32
TIMEOUT 600
ONTIMEOUT Boot Local
PROMPT 0
NOESCAPE 1
ALLOWOPTIONS 1

menu hshift 5
menu width 59

menu title pxe boot menu

menu color title        * #ffffffff *
menu color border       * #00000000 #00000000 none
menu color sel          * #ffffffff #76a1d0ff *
menu color hotsel       1;7;37;40 #ffffffff #76a1d0ff *
menu color tabmsg       * #ffffffff #00000000 *
menu color help         37;40 #ffdddd00 #00000000 none
menu vshift 2
menu rows 20
menu helpmsgrow 25
# the command line must be at least one line from the bottom.
menu cmdlinerow 26
menu timeoutrow 26
menu tabmsgrow 28
menu tabmsg press enter to boot or tab to edit a menu entry


label boot local
    COM32 syslinux\chain.c32
    APPEND hd0
    text help
        boot to local hard disk
    endtext

label winpe3 32 (cn)
    kernel syslinux/memdisk
    append iso raw
    initrd iso/winpe_x86.iso
    text help
        boot to windows pe 3 32bit iso (~184mb)
    endtext

label winpe3 64 (cn)
    MENU LABEL Autoinstall win10-amd64
    kernel syslinux/memdisk
    append iso raw
    initrd iso/winpe_amd64.iso
    text help
        boot to windows pe 3 64bit iso (~207mb)
        autoinstall win10 64
    endtext

LABEL Arch32_(NFS)
    TEXT HELP
        Boot the Arch Linux (i686) live medium (Using NFS).
    It allows you to install Arch Linux or perform system maintenance.
    ENDTEXT
    MENU LABEL Install Arch Linux (i686) (NFS)
    LINUX images/ARCH_201510/arch/boot/i686/vmlinuz
    INITRD images/ARCH_201510/arch/boot/intel_ucode.img,images/ARCH_201510/arch/boot/i686/archiso.img
    APPEND archisobasedir=arch archiso_nfs_srv=192.168.2.19:/tftpboot/images/ARCH_201510 ip=:::::eth0:dhcp

LABEL Arch64 (NFS)
    TEXT HELP
        Boot the Arch Linux (x86_64) live medium (Using NFS).
    It allows you to install Arch Linux or perform system maintenance.
    ENDTEXT
    MENU LABEL Install Arch Linux (x86_64) (NFS)
    LINUX images/ARCH_201510/arch/boot/x86_64/vmlinuz
    INITRD images/ARCH_201510/arch/boot/intel_ucode.img,images/ARCH_201510/arch/boot/x86_64/archiso.img
    APPEND archisobasedir=arch archiso_nfs_srv=192.168.2.19:/tftpboot/images/ARCH_201510 ip=:::::eth0:dhcp

LABEL kali2 live(amd64)
    TEXT HELP
        Boot to Kali 2.0 (sana) amd64 Live
    ENDTEXT
    MENU LABEL Kali 2.0 Live(amd64)
    KERNEL images/kali2/live/vmlinuz
    INITRD images/kali2/live/initrd.img
    APPEND netboot=nfs nfsroot=192.168.2.19:/tftpboot/images/kali2 boot=live username=root hostname=kali

# http://www.memtest.org/
LABEL memtest
MENU LABEL Run Memtest86+ (RAM test)
LINUX memtest

# http://hdt-project.org/
LABEL hdt
MENU LABEL Hardware Information (HDT)
COM32 syslinux/hdt.c32
APPEND modules_alias=syslinux/hdt/modalias.gz pciids=syslinux/hdt/pciids.gz

LABEL reboot
MENU LABEL Reboot
COM32 syslinux/reboot.c32

LABEL poweroff
MENU LABEL Power Off
COM32 syslinux/poweroff.c32


期间还自己学会了自制winpe,我做了2个版本的winpe,一个32位,一个64位.


最后,只要把要安装的机器修改为网卡启动,就可以进入我配置的pxe server了

如下图


Monday, November 7, 2011

linux中查看系统资源占用情况的命令

top:
主要参数
d:指定更新的间隔,以秒计算。
q:没有任何延迟的更新。如果使用者有超级用户,则top命令将会以最高的优先序执行。
c:显示进程完整的路径与名称。
S:累积模式,会将己完成或消失的子行程的CPU时间累积起来。
s:安全模式。
i:不显示任何闲置(Idle)或无用(Zombie)的行程。
n:显示更新的次数,完成后将会退出to
显示参数:
PID(Process ID):进程标示号。
USER:进程所有者的用户名。
PR:进程的优先级别。
NI:进程的优先级别数值。
VIRT:进程占用的虚拟内存值。
RES:进程占用的物理内存值。
SHR:进程使用的共享内存值。
S:进程的状态,其中S表示休眠,R表示正在运行,Z表示僵死状态,N表示该进程优先值是负数。
%CPU:该进程占用的CPU使用率。
%MEM:该进程占用的物理内存和总内存的百分比。
TIME+:该进程启动后占用的总的CPU时间。
Command:进程启动的启动命令名称,如果这一行显示不下,进程会有一个完整的命令行。
top命令使用过程中,还可以使用一些交互的命令来完成其它参数的功能。这些命令是通过快捷键启动的。
<空格>:立刻刷新。
P:根据CPU使用大小进行排序。
T:根据时间、累计时间排序。
q:退出top命令。
m:切换显示内存信息。
t:切换显示进程和CPU状态信息。
c:切换显示命令名称和完整命令行。
M:根据使用内存大小进行排序。
W:将当前设置写入~/.toprc文件中。这是写top配置文件的推荐方法。

free
1.作用
free命令用来显示内存的使用情况,使用权限是所有用户。

2.格式
free [-b-k-m] [-o] [-s delay] [-t] [-V]

3.主要参数
-b -k -m:分别以字节(KB、MB)为单位显示内存使用情况。
-s delay:显示每隔多少秒数来显示一次内存使用情况。
-t:显示内存总和列。
-o:不显示缓冲区调节列。

uptime
18:59:15 up 25 min,  2 users,  load average: 1.23, 1.32, 1.21
现在的时间
系统开机运转到现在经过的时间
连线的使用者数量
最近一分钟,五分钟和十五分钟的系统负载
参数: -V 显示版本资讯。

vmstat
procs -----------memory---------- ---swap-- -----io---- --system-- ----cpu----
r  b   swpd free   buff   cache    si   so    bi     bo    in    cs   us  sy  id wa
0  1  24980 10792 8296  47316    5   19   205    52  1161   698  26  3  1  70
1 观察磁盘活动情况
磁盘活动情况主要从以下几个指标了解:
bi:表示从磁盘每秒读取的块数(blocks/s)。数字越大,表示读磁盘的活动越多。
bo:表示每秒写到磁盘的块数(blocks/s)。数字越大,表示写磁盘的活动越多。
wa:cpu等待磁盘I/O(未决的磁盘IO)的时间比例。数字越大,表示文件系统活动阻碍cpu的情况越严重,因为cpu在等待慢速的磁盘系统提供数据。wa为0是最理想的。如果wa经常大于10,可能文件系统就需要进行性能调整了。
2 观察cpu活动情况
vmstat比top更能反映出cpu的使用情况:
us:用户程序使用cpu的时间比例。这个数字越大,表示用户进程越繁忙。
sy: 系统调用使用cpu的时间比例。注意,NFS由于是在内核里面运行的,所以NFS活动所占用的cpu时间反映在sy里面。这个数字经常很大的话,就需要注 意是否某个内核进程,比如NFS任务比较繁重。如果us和sy同时都比较大的话,就需要考虑将某些用户程序分离到另外的服务器上面,以免互相影响。
id:cpu空闲的时间比例。
wa:cpu等待未决的磁盘IO的时间比例。

iostat
用于统计CPU的使用情况及tty设备、硬盘和CD-ROM的I/0量
参数:
    -c      只显示CPU行
    -d      显示磁盘行
    -k      以千字节为单位显示磁盘输出
    -t       在输出中包括时间戳
    -x      在输出中包括扩展的磁盘指标

avg-cpu:  %user   %nice    %sys %iowait   %idle
          20.25    0.18    2.61   76.39    0.57
%iowait 等待本地I/O时CPU空闲时间的百分比
%idle 未等待本地I/O时CPU空闲时间的百分比

Device:            tps   Blk_read/s   Blk_wrtn/s   Blk_read   Blk_wrtn
hda               9.86       284.34        84.48     685407     2036
每秒传输数(tps)、每秒512字节块读取数(Blk_read/s)、每秒512字节块写入数(Blk_wrtn/s)和512字节块读取(Blk_read)和写入(Blk_wrtn)的总数量。

一、如何查看服务器的CPU
今天安装了9台Linux服务器,型号完全不一样(有DELL、HP和IBM服务器),又懒得去对清单,如何在Linux下cpu的个数和核数呢? 另外,nginx的cpu工作模式也需要确切的知道linux服务器到底有多少个逻辑cpu,不过现在服务器那是相当的彪悍,直接上 worker_processes 8吧。
判断依据:
1.具有相同core id的cpu是同一个core的超线程。(Physical id and core id are not necessarily consecutive but they are unique. Any cpu with the same core id are hyperthreads in the same core.)
2.具有相同physical id的cpu是同一颗cpu封装的线程或者cores。(Any cpu with the same physical id are threads or cores in the same physical socket.)
以自己的惠普DL380G6为例说明:
①物理cpu个数:
[root@localhost ~]# cat /proc/cpuinfo | grep "physical id" | sort | uniq | wc -l
②每个物理cpu中core的个数(即核数)
[root@localhost ~]# cat /proc/cpuinfo | grep "cpu cores" | uniq
cpu cores : 4
这样可以推算出自己服务器的逻辑CPU为物理个数*核数,如果你的服务器是server2003,你可以在你的资源管理器里以图形化方式看到你的逻辑CPU个数。

二、查看服务器的内存情况
有时候,你运行了许多大的进程,比如你开启了300个fast-cgi。这时候你感觉系统很慢,便需要查看服务器的内存情况:
[root@server ~]# free -m
                     total used free shared buffers cached
Mem:            249   163    86            0        10         94
-/+ buffers/cache:   58  191
Swap:            511       0  511
参数解释:
total 内存总数
used 已经使用的内存数
free 空闲的内存数
shared 多个进程共享的内存总额
buffers Buffer Cache和cached Page Cache 磁盘缓存的大小
-buffers/cache (已用)的内存数:used - buffers - cached
+buffers/cache(可用)的内存数:free + buffers + cached
可用的memory=free memory+buffers+cached
上面的数值是一台我公司内网供PHP开发人员使用的DELL PE2850,内存为2G的服务器,其可使用内存为=217+515+826。记住,Linux的内存使用管理机制是有多少就用多少(特别是在频繁存取文件后),即Linux内存不是拿来看的,是拿来用的。


编辑推荐:Linux系统内存监控全面讲解:free命令

三、服务器磁盘使用情况
有时感觉硬盘反映很慢,或需要查看日志所在分区时,下列命令可以查看磁盘的使用情况,很有用:
①查看硬盘分区情况
fdisk -l
②查看当前硬盘使用情况
df -h
③查看硬盘性能
# iostat -x 1 10
Linux 2.6.18-92.el5xen                          03/01/2010

avg-cpu: %user %nice %system %iowait %steal %idle
                     1.10 0.00 4.82 39.54 0.07 54.46

Device:   rrqm/s wrqm/s       r/s       w/s     rsec/s wsec/s avgrq-sz avgqu-sz await svctm %util

sda              0.00       3.50    0.40      2.50       5.60   48.00      18.48         0.00   0.97   0.97   0.28
sdb              0.00       0.00    0.00      0.00       0.00      0.00       0.00          0.00   0.00   0.00   0.00
sdc               0.00       0.00    0.00      0.00       0.00      0.00       0.00         0.00    0.00   0.00  0.00
sdd              0.00       0.00    0.00      0.00       0.00      0.00       0.00          0.00    0.00   0.00  0.00
sde              0.00       0.10    0.30      0.20       2.40      2.40       9.60          0.00    1.60   1.60  0.08
sdf             17.40       0.50 102.00    0.20 12095.20   5.60   118.40          0.70    6.81  2.09 21.36
sdg          232.40       1.90 379.70    0.50 76451.20 19.20   201.13          4.94 13.78   2.45 93.16
像我们公司,开发人员占多数。有时为了节约成本,会同时采购一些性价比比较高的二手服务器;这时一般将服务器的硬件的工作模式设置为RAID1,同 时对几种型号的服务器作相同文件的写操作,然后各自执行iostat –d,作对比性测试。这样服务器的硬盘性能孰优孰劣,一下子就能对比出来。


图解:
Tps 该设备每秒I/O传输的次数(每秒的I/O请求)
Blk_read/s 表求从该设备每秒读的数据块数量
Blk_wrth/s 表示从该设备每秒写的数据块数量
编辑推荐:Linux系统监控之磁盘I/O篇


四、查看系统内核
查看系统内核主要为了掌握其版本号,为安装LVS等软件做准备。
uname –a
有关查看内核信息的更多指令,可参考Linux查看版本信息及CPU内核、型号等一文。有关Linux的内核优化,则可参考Linux 2.6.31内核优化指南一文。

五、查看服务器使用的Linux发行版的名称、版本号及描述信息等
lsb_release -a
这是我的某台用于SVN实验的vmware机器情况:







六、查看服务器的平均负载
感觉到系统压力较大时用可top或uptime查看下服务器的平均负载。uptime的另一个用法是查看你的Linux服务器已经稳定运行多少天没有重启了,我这边的机器最高记录是360多天。

七、查看系统整体性能情况
如果感觉系统比较繁忙,可以用vmstat查看系统整体性能情况。vmstat不仅仅适应于linux系统,它一样适用于FreeBSD等unix系统。
vmstat 1 2
如果 r经常大于 4 ,且id经常少于40,表示cpu的负荷很重。
如果pi,po 长期不等于0,表示内存不足。
如果disk 经常不等于0, 且在 b中的队列 大于3, 表示 io性能不好。
编辑推荐:Linux系统监控工具之vmstat详解

八、查看系统已载入的相关模块
Linux操作系统的核心具有模块化的特性,应此在编译核心时,务须把全部的功能都放入核心。你可以将这些功能编译成一个个单独的模块,待需要时再分别载入。比如说在安装LVS+Keepalived,下列用法被经常用到:
#检查内核模块,看一下ip_vs是否被加载 lsmod |grep ip_vsip_vs 77313 0
如果要查看当前系统的已加载模块,直接lsmod。

九、Linux下查找PCI设置
有时需要在Linux下查找PCI设置,可用lspci命令,它可以列出机器中的PCI 设备,比如声卡、显卡、Modem、网卡等,主板集成设备也能列出来。lspci 读取的是hwdata 数据库。有的小伙可能和我一样,最关心的还是网卡型号:
[root@mail ~]# lspci | grep Ethernet
03:00.0 Ethernet controller: Broadcom Corporation NetXtreme II BCM5708 Gigabit Ethernet (rev 12)
07:00.0 Ethernet controller: Broadcom Corporation NetXtreme II BCM5708 Gigabit Ethernet (rev 12)

推荐阅读:七大实用命令行工具 玩转Linux网络配置。至于网络流量的监控,建议用centos自带的工具iptraf,其用法可参考这篇文章。

掌握上面这九条,你对你的Linux服务器的硬件信息基本就可以随时掌控了。

freebsd 路由问题

使用virtualbox安装的freebsd配置的nginx+php-fpm+mysql安装的wordpress无法使用no-ip的动态域名访问, 可是同样配置的arch上的wordpress却可以从外网无障碍访问?


使用路由的端口映射,把外网的80端口转给内网的虚拟机上,就是使用的NAT(network address translation)(网络地址转换)


仔细研究了研究,2机的不同点只有ip,freebsd的ip为192.168.1.110,arch的ip为192.168.111

然后就想到是不是路由设置的问题?

2台guest机为了和主机通信,都是使用的2个网卡,一个使用的NAT自动获取地址;1个使用的bridged 桥接网络来实现和 host机的链接.

使用netstat -s 来察看

发现freebsd的设定为192.168.1.110的2号网卡em0没有设置网关....难道就是这个问题?

最后,每个路由的不同属性可以在 Flags 列中看到。下边是个关于这些标志和它们的含义的一个简表:
U Up: 路由处于活动状态。
H Host: 路由目标是单个主机。
G Gateway: 所有发到目的地的网络传到这一远程系统上, 并由它决定最后发到哪里。
S Static: 这个路由是手工配置的,不是由系统自动生成的。
C Clone: 生成一个新的路由, 通过这个路由我们可以连接上这些机子。 这种类型的路由通常用于本地网络。
W WasCloned: 指明一个路由——它是基于本地区域网络 (克隆) 路由自动配置的。
L Link: 路由涉及到了以太网硬件。

通过修改/etc/rc.conf 文件,给em0网卡添加路由

在/etc/rc.conf中,添加
defaultrouter="192.168.1.1"

然后重启网络服务
# /etc/rc.d/netif restart
注意: 如果在 /etc/rc.conf 中配置了默认网关, 还需要运行下面的命令:
# /etc/rc.d/routing restart


然后,使用noip注册的域名来访问,OK啦!


参考  :  设置网卡
             网关和路由

Wednesday, November 2, 2011

修改fstab让ntfs分区挂载文件读写权限,文件夹读写执行权限

一直使用的ntfs-3g来读写win下的分区,没有怎么注意权限的问题,可是看到挂载的分区文件都是可执行权限,对安全性以及读写都有问题,因为使用awesome所以一般很少开文件管理器,最多使用mc来批量复制拷贝文件,看到一片绿色可执行,头大.

察看fstab的参数原来可以制定挂载的时候,对文件以及文件夹采用不同的权限.

比如下面这段挂载的参数:

/dev/ /mnt/windows ntfs-3g gid=users,fmask=113,dmask=002 0 0 

以群组users,文件权限664,文件夹权限774来挂载.

fmask,dmask,以及umask的意思就是扣除相应的权限,777-$(*mask)= 来定义挂载的权限.

umask代表user的mask,fmask代表文件file的mask,当然dmask就代表文件夹的mask了.


一.022表示默认创建新文件权限为755 也就是 rxwr-xr-x(所有者全部权限,属组读写,其它人读写) 
二.027表示默认创建新文件权限为750 也就是rxwr-x---(所有者全部权限,属组读写,其它人无)

下表列出了一些umask值及它们所对应的目录和文件权限
常用的umask值及对应的文件和目录权限
umask值        目录         文件
022               755          644
027               750          640
002               775          664
006               771          660
007               770          660
我现在的fstab配置文件如下
/etc/fstab.html
# 
# /etc/fstab: static file system information
#
# <file system> <dir>   <type>  <options>   <dump>  <pass>
tmpfs       /tmp    tmpfs   nodev,nosuid    0   0
UUID=b2b3abc1-d666-4948-9c17-a5e12c637d63 /boot reiserfs defaults 0 1
UUID=afd39ba2-f3f3-4bac-bbc8-dc5307ba4e89 / ext4 defaults 0 1
UUID=26f23578-3dc8-4394-bcba-7320a4aaff7d swap swap defaults 0 0
UUID=90ab4280-c642-47c0-91e4-d3be7d254d1a /home ext4 defaults,user_xattr 0 1
UUID="4D9E019F07B2E598"  /media/wine ntfs-3g gid=users,fmask=113,dmask=002,nls=utf8 0 2

Tuesday, November 1, 2011

mpd 启动错误 Address family not supported by protocol

这2天arch启动的时候tty1老提示 Error : Socked :  Address family not supported by protocol


放google搜索,找到Address family not supported by protocol - IPv6 issue on machines with disabled v6

This happens on a machine that has socket.has_ipv6 set to true as the linux system supports IPv6, but jason has disabled IPv6 by following https://wiki.archlinux.org/index.php/IPv6_-_Disabling_the_Module

手动启动的时候提示如下错误:

listen: bind to '[::1]:6600' failed: Failed to create socket: Address family not supported by protocol (continuing anyway, because binding to '127.0.0.1:6600' succeeded)
playlist directory is not a directory: "/home/jarodlau/.mpd/playlists"

原来是禁用ipv6的结果,另外我没有把播放列表的目录拼写正确,少了个s,应该是playlists,我的写成了playlist

解决办法很简单 :

编辑 /etc/hosts文件 ,把一些ipv6的地址注释掉就ok了,如下

/etc/hosts.html
#
# /etc/hosts: static lookup table for host names
#

#<ip-address>   <hostname.domain.org>   <hostname>
127.0.0.1   localhost.localdomain   localhost arch-laptop
#::1        localhost.localdomain   localhost

# End of file

Monday, October 31, 2011

2个不错的dns

因为电信的114问题,dns劫持严重,于是使用了pdnsd来本地缓存dns信息,可是google和opendns的延迟很严重,所以一直在找其他不错的dns

google的8.8.8.8和8.8.4.4以及opendns的208.67.222.222和208.67.220.220,用的人太多了,gfw重点照顾,所以不太好用了

推荐个台湾中华电信dns:168.95.192.1 和 168.95.192.2

编辑pdnsd的配置文件  sudo vim /etc/pdnsd.conf


~/pdnsd.conf.html
# The following section is most appropriate if you have a fixed connection to
# the Internet and an ISP which provides good DNS servers.
server {
    label= "mydns";
    ip = 168.95.192.1,168.95.192.2;  # Put your ISP's DNS-server address(es) here.
#   proxy_only=on;     # Do not query any name servers beside your ISP's.
                       # This may be necessary if you are behind some
                       # kind of firewall and cannot receive replies
                       # from outside name servers.
    timeout=4;         # Server timeout; this may be much shorter
               # that the global timeout option.
    #uptest=if;         # Test if the network interface is active.
    #interface=eth0;    # The name of the interface to check.
    #uptest=ping;
    #    ping_timeout=100;
    interval=10m;      # Check every 10 minutes.
    purge_cache=off;   # Keep stale cache entries in case the ISP's
               # DNS servers go offline.
} 
然后重启就ok了 sudo rc.d restart pdnsd

Wednesday, October 26, 2011

利用grub2直读iso镜像 制作多启动u盘

U盘现在容量越来越大,4-8G的U盘只装一个启动系统太浪费了,这几天看了看利用grub2制作多启动U盘的内容,有点心得,这里大概说说.

原理:
利用grub2可以启动iso文件的方法,可以制作大量的多启动U盘,但是很大部分还不支持,比如常用的debian,centos,等等就不支持,不过我常用的systermrescuecd,archboot,freebsd都支持这种启动方式.


操作过程:

首先把U盘分区,我实验用的U盘有2G容量,为了通用性,我就格式化为fat32格式,这样windows和linux,以及bsd下都可以复制iso,修改grub.cfg文件.

1)分区就使用fdisk

首先你要把u盘的mbr清空,避免以前做的mbr影响将要安装的grub2启动

(X代表你的U盘所在的盘符,比如我的u盘就是/dev/sdb)

$ sudo dd if=/dev/zero of=/dev/sdX bs=512  count=1

$ sudo fdisk -l
$ sudo fdisk /dev/sdX

d删除已有分区,p显示分区情况,m察看fdisk帮助,n新建分区,p为主分区,t转换分区格式,b转换为win95vfat格式,l显示可以转换的分区格式,a激活启动分区,w保存分区情况,q不保存关闭fdisk

2) 安装grub2

这里要注意,安装的grub2最好是32位的,比如systermrescuecd和ubuntu livecd都选用32位的,可以保持兼容性;我这里就走了很多弯路,使用机器arch本身的grub2,导致u盘无法启动的情况,应为我的arch是64位的.

首先挂载你的u盘,如果你使用gnome3,或者其他自动挂载的桌面,你就需要找到你的u盘挂载的地方

$ sudo mount -t vfat /dev/sdb1 /mnt

$ sudo grub-install --no-floppy --force --root-directory=/mnt /dev/sdb1


3) 编辑grub.cfg文件,并添加iso文件

$ sudo mkdir /mnt/iso

把你需要启动的iso镜像放到/mnt/iso目录中就可以了

$ sudo vim /mnt/boot/grub/grub.cfg

下面是我的grub.cfg配置文件,你可以自己根据自己的情况修改
~/grub.cfg.html
set color_normal='light-blue/black'
set color_highlight='light-cyan/blue'

menuentry "System Rescue CD" {
 loopback loop /iso/systemrescuecd-x86-2.3.1.iso
 linux (loop)/isolinux/rescuecd isoloop=/iso/systemrescuecd-x86-2.3.1.iso setkmap=us docache dostartx
 initrd (loop)/isolinux/initram.igz
}

menuentry "Parted Magic Disk Utilities" {
 loopback loop /iso/pmagic-6.2.iso
 linux (loop)/pmagic/bzImage iso_filename=/iso/pmagic-6.2.iso edd=off noapic load_ramdisk=1 prompt_ramdisk=0 rwnomce sleep=10 loglevel=0
 initrd (loop)/pmagic/initramfs
}

menuentry "Ubuntu 11.04" {
 loopback loop /iso/ubuntu-11.04-desktop-i386.iso
 linux (loop)/casper/vmlinuz boot=casper iso-scan/filename=/iso/ubuntu-11.04-desktop-i386.iso noeject noprompt --
 initrd (loop)/casper/initrd.lz
}

menuentry "Arch Linux i686" {
 loopback loop /iso/archlinux-2011.10-1-archboot.iso
 linux (loop)/boot/vmlinuz rootdelay=10
 initrd (loop)/boot/initrd.img
}

menuentry "Arch Linux x86_64" {
 loopback loop /iso/archlinux-2011.10-1-archboot.iso
 linux (loop)/boot/vm64 rootdelay=10
 initrd (loop)/boot/initrd64.img
}

menuentry "FreeBSD 8.2 i386" {
 set isofile=/iso/FreeBSD-8.2-RELEASE-i386-disc1.iso
 loopback loop $isofile
 kfreebsd (loop)/boot/kernel/kernel iso-scan/filename=$isofile noeject noprompt splash --
 kfreebsd_module (loop)/boot/mfsroot.gz type=mfs_root
}

menuentry "FreeBSD 8.2 x86_64" {
 set isofile=/iso/FreeBSD-8.2-RELEASE-amd64-disc1.iso
 loopback loop $isofile
 kfreebsd (loop)/boot/kernel/kernel iso-scan/filename=$isofile noeject noprompt splash --
 kfreebsd_module (loop)/boot/mfsroot.gz type=mfs_root
}


参考 :  Restore_Usbstick
       Boot Multiple ISO from USB via Grub2 using Linux
       MultiBootUSB shell
       Installing Grub (2) on a USB flash drive
       Install GRUB2 on USB from Ubuntu Linux

qemu直接使用u盘启动

一直使用virtualbox做虚拟机,今天想做个usb的多启动光盘,但是vbox使用u盘启动的方法太麻烦了.


搜索发现qemu可以支持直接u盘启动,方法也很简单.

首先 , 确保安装qemu  

$ sudo pacman -S qemu


然后,找到你的u盘,可以使用fdisk,df,来察看,比如我的u盘就是/dev/sdb

$ sudo fdisk -l

最后,就可以使用qemu来启动你的u盘了

$ sudo qemu -m 128 -hda /dev/sdb

上面参数的含义是使用128M内存,以hda方式,把/dev/sdb作为启动盘引导


当然,你也可以使用物理硬盘来引导,比如

$ sudo qemu -m 128 -hda /dev/sda

就可以把你本身的物理硬盘系统启动起来




使用最新的archboot制作启动u盘

archboot 是arch脱离出来的使用最新arch源制作非官方的启动镜像,2011-10月版本作成了2合一盘,i686和X86_64位同一个iso,好东西,下面是archboot wiki的介绍.

Archboot Releases

下载下来以后, 直接 运行一下命令就可以

dd if= of=/dev/ bs=1M

恢复u盘的默认状态

Restore Usbstick

Take care about which device actually is your USB stick. The next command will render all data on /dev/sdX inaccessible.
  • First, wipe the bootsector of the USB stick:
dd if=/dev/zero of=/dev/sdX bs=512 count=1
  • Then, create a new FAT32 partition on the stick and write a FAT32 filesystem on it (vfat or type b in fdisk terminology):
fdisk /dev/sdX 
n
p
1


t
b
w


mkdosfs -F32 /dev/sdX1 
or (mkfs.vfat -F 32 -n mutilboot /dev/sdX1)

几个常用开源镜像源推荐

平常经常装开源软件,从国外下载很慢,国内都有很不错的镜像,比如我常用的freebsd,centos,arch,ubuntu,这些镜像都有相应的源.

首先就是国内大学类的镜像代表 BJTU 了,地址  http://mirror.bjtu.edu.cn/cn/


然后是国内比较流行的163网易镜像源,163还有不错的免费邮箱,是我邮箱的第二选择,当然第一选择是 gmail了 , 地址 http://mirrors.163.com/

arch linux 重装win7后恢复grub2启动

今天很郁闷,本来想尝试把u盘使用grub2作成多启动的,结果一不小心,把win7的c盘格式化了,只能进入arch,于是又重新安装了win7,结果又进不了arch,只能使用刚坐好的u盘恢复本机的arch了.


情景:

arch刚装好grub2(安装的grub2-bios),配置文件都还在;
arch的安装镜像使用的是grub,所以需要有支持grub2安装的版本进行恢复,比如ubuntu的livecd,当然你也可以选择和我一样使用archlinux-2011.10-1-archboot.iso(支援i686和x86_64位安装).

操作步奏:


首先使用livecd启动,打开一个终端

$ su -

把自己转为root,察看分区情况

# fdisk -l

这里可以看到你的分区情况,比如我的/分区在/dev/sda10 ,/boot 分区在 /dev/sda8

然后把分区挂载起来

# mount /dev/sda10 /mnt

# mount /dev/sda8 /mnt/boot

接下来安装grub2 (grub2-bios)

# grub_bios-install --root-directory=/mnt/  /dev/sda

接下来重启就可以了

参考:
         记录一下重装Win后grub2的恢复方法
         arch wiki archboot 使用方法

archlinux mkfs.vfat 和 mkfs.ntfs

今天做个U盘的多启动盘,发现没有安装fat32的格式化工具,倒是有ntfs的格式化.

原来需要mkfs支持vfat和ntfs格式化需要2个包
 
dosfstools


$ sudo pacman -S dosfstools 

这个包给你2个命令,mkfs.vfat,mkfs.msdos

如果你也想支持ntfs的格式化mkfs.ntfs就需要安装 


$ sudo pacman -S  ntfsprogs


Tuesday, October 25, 2011

archlinux 升级grub2

GRUB2是下一代的多系统引导器,grub即将退位,项目名也改成了grub-legacy。相比grub(目前的版本是0.97)而言,grub2重写了代码以实现模块化和可移植性。

GRUB 2 has replaced what was formerly known as GRUB (i.e. version 0.9x), which has, in turn, become GRUB Legacy. Enhancements to GRUB are still being made, but the current released versions are quite usable for normal operation.
GRUB Legacy is no longer being developed. For the differences between GRUB Legacy and GRUB, see the Grub Legacy Documentation.

Grub2的Logo:

将archlinux的grub升级到grub2是一件很简单的事情,你只需要通过pacman安装新的grub2软件包,转换menu.lst为grub.cfg就可以了。这篇日志的大部分资料来自Archlinux的wiki,命令截图来自我的笔记本电脑。

首先,搜索一下仓库里的grub2软件包: 


你会发现目前仓库里有4个grub2的软件包,做个简单的介绍吧:

grub2-bios             –> 适用于PC BIOS
grub2-commin      –> 文件集合(包含了所有平台相关的文件)
grub2-efi-i386       –> 适用于32位平台的UEFI
grub2-efi-x86_64  –> 适用于64位平台的UEFI

一般情况下,我们选择安装grub2-bios。什么是UEFI呢?这货全称Unified Extensible Firmware Interface,直译叫做“统一的可扩展固件接口”,这种接口用于操作系统自动从预启动的操作环境,加载到某种操作系统上,从而使开机程序化繁为简, 节省时间。UEFI将替代已存在多年的BIOS,目前市场上的主板已经开始采用UEFI 。

 安装和升级过程:

 1) 安装grub2-bios
安装过程中会提示与现有的grub冲突,直接回答y就可以了; 另外如果你喜欢干净,可以把/boot/grub/ 目录备份到别的地方,然后删除grub目录里的内容.

2) 擦除MBR上grub-legacy的痕迹. 这一步很关键,如果你的grub是安装在MBR上,你就需要做这一步操作. of后面跟的是硬盘设备名,可以使用$ sudo fdisk -l 察看.

3) 安装grub2到MBR.
     $sudo cp /usr/share/grub/{unicode.pf2,ascii.pf2} /boot/grub/

4) 生成grub2的配置文件grub.cfg.(可以从 menu.lst 或者 grub.conf 转换过来)


5)重启系统,检验成果


结束语:

这里只是介绍了Archlinux上如何将grub(又名grub-legacy)升级到grub2,至于grub2的命令操作、grub.cfg的配置方式则没有给出说明 (你可以在wiki上获得详细内容,下面有链接)

参考资料:
1. Archlinux上有详细的介绍,包括BIOS和UEFI的安装方式,GPT的延伸阅读。
https://wiki.archlinux.org/index.php/GRUB2#For_BIOS_Systems
https://wiki.archlinux.org/index.php/GRUB2#During_Arch_Linux_installation
2. GRUB项目主页
http://www.gnu.org/software/grub/index.html

pacman 之 清理缓存脚本

~/code/shell/cleanold.html
#!/bin/bash
#
# 本脚本清理指定文件夹(不知定默认为/var/DIR/pacman/pkg)下过期的软件包,默认保留最新的两个版本.
# Carbon Jiao   2009-09-06 
# Archlive <http://archlive.google.com>
#
# 本脚本最开始算法来自于http://www.linuxsir.org/bbs/thread348435.html
# 原作者  veryxp
#
# 2009-09-21 更新:执行前先等待5秒。。。

APPNAME=$(basename "${0}")
cmdline=$@

KEEPN=2
DIR=""

# usage: usage <exitvalue>
usage ()
{
   echo ""
   echo " 用法: sudo sh ./${APPNAME} [选项] -r <预清理目录>"
   echo "       如果将本脚本放在系统PATH指定的目录中,则:"
   echo "       sudo ${APPNAME} [选项] -r <预清理目录>"
   echo " 选项:"
   echo "    -f               删除前无需确认;"
   echo "    -a               没有安装的所有包也删除;"
   echo "    -r <DIR>         DIR 为预清理的目录;"
   echo "    -n <n>          n: 保留版本数默认为$KEEPN;"
   echo "    -h               本帮助信息."
   exit $1
}
COLORMSG=1
plain() {
    if [ $COLORMSG -eq 1 ]; then
        echo -e "\033[1;2m    $@\033[1;0m"
    else
        echo -e "   $@"
    fi
}

msg() {
    if [ $COLORMSG -eq 1 ]; then
        echo -e "\033[1;32m==>\033[1;0m\033[1;1m $@\033[1;0m"
    else
        echo -e "==> $@"
    fi
}
msg2() {
    if [ $COLORMSG -eq 1 ]; then
        echo -e "\033[1;34m  ->\033[1;0m\033[1;1m $@\033[1;0m"
    else
        echo -e "  -> $@"
    fi
}
warn() {
    if [ $COLORMSG -eq 1 ]; then
        echo -e "\033[1;33m==> $(gettext "WARN:")\033[1;0m\033[1;1m $@\033[1;0m"
    else
        echo -e "==> $(gettext "WARN:") $@"
    fi
}
error() {
    if [ $COLORMSG -eq 1 ]; then
        echo -e "\033[1;31m==> $(gettext "ERROR:")\033[1;0m\033[1;1m $@\033[1;0m"
    else
        echo -e "==> $(gettext "ERROR:") $@"
    fi
}

[ "$EUID" != "0" ] && error "错误: 必须以root用户运行本制作脚本." && usage 1

# 检测、配置输入的命令
while getopts ':fahn:r:' arg; do
   case "${arg}" in
    n) KEEPN="${OPTARG}";;
    r) DIR="${OPTARG}";;
    f) ;;
    a) ;;
        h) usage 0;;
        ?) error "输入$@包含无效参数 "; usage 1;;
   esac
done

if [ "x$DIR" = "x" ]; then
    if [ ! -f /etc/pacman.conf ]; then error "没有指定预清理的目录,且操作系统不是Arch,为安全考虑,程序退出!" && usage 1; fi
    DIR=$(grep CacheDir /etc/pacman.conf | awk '{print $3}')
    if [ "x$DIR" = "x" ]; then error "没有指定预清理的目录,且pacman.conf无缓存相关设置!" && usage 1; fi
fi

msg "输入的参数:$cmdline"
result="清理目录$DIR, 保留${KEEPN}个版本"
case $cmdline in *a*) result="${result}, 删除系统没有安装的软件包"
            if [ ! -f /etc/pacman.conf ]; then error "操作系统不是Arch, 出现问题后果自负!"; fi
            ;;
        esac
case $cmdline in *f*) result="${result}, 删除先无需确认";; esac

msg ${result}

shift `expr $OPTIND - 1`

#if ! [ "x${@}" = "x" ]; then
#   error "无法识别的参数或命令: $@"
#   usage 1
#fi

msg "5秒后继续,Ctrl+C 可以结束本脚本的执行。"
sleep 5

action_clean ()
{
   for ((i=0; i<${#list[@]}; i++)); do
    # 取得软件包名称
    pkgname=${list[$i]% *}

    case $cmdline in *-a*)
        # 如果系统里没有安装此包,将其删除
        #if ! pacman -Qq $pkgname>/dev/null 2>&1; then  #不喜欢调用pacman的用如下语句 Carbonjiao注
        if [ "x$(ls $(grep DBPath /etc/pacman.conf | awk '{print $3}')/local | grep "${list[$i]// /-}")" = "x" ]; then
            plain "系统中没有安装$pkgname, 现在删除$DIR 下的软件包 ${list[$i]// /-}"
            case $cmdline in *-f*) rm -rf ${list[$i]// /-};;
                  *) rm -ri ${list[$i]// /-} ;;
            esac
            continue
        fi
        ;;
    esac

    # 如果出现重复的软件包名称,则更新重复计数器 dup
    if [ "$pkgname" == "$prev" ]; then
        let "dup++"
    else
    # 如果计数器超过设置的 KEEPN,就开始清理
        if [ ${dup:-0} -gt $KEEPN ]; then
            msg "软件 $prev 有 $dup 个版本"
            # 确定需要处理的数组上下标 ub 和 lb,用 let 赋值更会清晰一些
            ub=$(($i-1)); lb=$(($i-$dup))

            # 用冒泡法将要处理的数组元素排序,版本旧的在前
            # 如够偷懒可以不排序,按修改时间判断包的新旧,可能不准
            for ((x=$lb; x<$ub; x++)); do
                for ((y=$ub; y>$x; y--)); do
                    vercmp "${list[$y]// /-}" "${list[$(($y-1))]// /-}" > /dev/null
                    if [ $? -lt 0 ]; then
                        t="${list[$y]}"
                        list[$y]="${list[$(($y-1))]}"
                        list[$(($y-1))]="$t"
                    fi
                done
            done

            # 排序好之后,清理掉多余的包
            for ((z=0; z<$(($dup-$KEEPN)); z++)); do
                plain "删除过期软件包${list[$(($lb+$z))]// /-}"
                case $cmdline in *-f*) rm -rf "${list[$(($lb+$z))]// /-}" ;;
                         *) rm -ri "${list[$(($lb+$z))]// /-}" ;;
                esac
            done
        fi
        dup=1 # 重置计数器 dup
    fi
    prev="$pkgname" # 记录上一个包的名称
   done
}

IFS=$'\n'
SED='s/\(.\{1,\}\)-\([0-9][a-z0-9\._]*-[0-9\.]\{1,\}\)/\1 \2/'
# 用 sed 将缓存中的文件名分割成“软件包名称+空格+其它后缀的形式,并存入数组 list
# 如 xsel-1.2.0-1.pkg.tar.gz 在数组中就是 "xsel -1.2.0-1.pkg.tar.gz"
# 使用下面的 ls 命令请确保文件是按名称排序

cd "$DIR"
if [ "$(ls -1 *-i686.* | wc -l)" -gt 0 ]; then
    list=(`ls -1 *-i686.* | sed -e "$SED"`)
    action_clean
fi
if [ "$(ls -1 *-x86_64.* | wc -l)" -gt 0 ]; then
    list=(`ls -1 *-x86_64.* | sed -e "$SED"`)
    action_clean
fi
if [ "$(ls -1 *-any.* | wc -l)" -gt 0 ]; then
    list=(`ls -1 *-any.* | sed -e "$SED"`)
    action_clean
fi