ntpd install

时间同步

  1. 设置时区和时间

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    [root@[ALL] ~]# [ -f /etc/localtime ] && cp -f /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
    [root@[ALL] ~]# [ -f /etc/sysconfig/clock ] && echo 'ZONE="Asia/Shanghai"' | tee /etc/sysconfig/clock
    [root@[ALL] ~]# [ -f /etc/timezone ] && echo 'Asia/Shanghai' | tee /etc/timezone
    [root@[ALL] ~]# [ -f /etc/sysconfig/ntpd ] && echo 'SYNC_HWCLOCK=yes' | tee -a /etc/sysconfig/ntpd
    [root@[ALL] ~]# ntpdate cn.pool.ntp.org
    [root@[ALL] ~]# cp -f /etc/{ntp.conf,ntp.conf.bak}
    [root@[ALL] ~]# cat > /etc/ntp.conf <<EOF
    > driftfile /var/lib/ntp/drift
    > restrict default nomodify notrap nopeer noquery
    > restrict 127.0.0.1
    > restrict ::1
    > server cn.pool.ntp.org prefer
    > server 0.centos.pool.ntp.org iburst
    > server 1.centos.pool.ntp.org iburst
    > server 2.centos.pool.ntp.org iburst
    > server 3.centos.pool.ntp.org iburst
    > includefile /etc/ntp/crypto/pw
    > keys /etc/ntp/keys
    > disable monitor
    > EOF
    [root@[ALL] ~]# cp -f /etc/ntp/{step-tickers,step-tickers.bak}
    [root@[ALL] ~]# cat > /etc/ntp/step-tickers <<EOF
    > cn.pool.ntp.org
    > 0.centos.pool.ntp.org
    > 1.centos.pool.ntp.org
    > 2.centos.pool.ntp.org
    > 3.centos.pool.ntp.org
    > EOF
    [root@[ALL] ~]# systemctl start ntpd && systemctl enable ntpd
    1
    2
    3
    4
    //不在推荐使用
    [root@[ALL] ~]# systemctl stop ntpd
    [root@[ALL] ~]# systemctl disable ntpd
    [root@[ALL] ~]# echo '1 1 * * * /usr/sbin/ntpdate -s cn.pool.ntp.org' | tee /var/spool/cron/root

  1. ntpd 和 ntpdate 简介

    1
    2
    3
    4
    5
    在Windwos中,系统时间的设置很简单,界面操作,通俗易懂.而且设置后,重启,关机都没关系.系统时间会自动保存在Bios的时钟里面,启动计算机的时候,系统会自动在Bios里面取硬件时间,以保证时间的不间断.

    在Linux下,默认情况下,系统时间和硬件时间,并不会自动同步.在Linux运行过程中,系统时间和硬件时间以异步的方式运行,互不干扰.硬件时间的运行,是靠Bios电池来维持,而系统时间,是用CPU tick来维持的

    在系统开机的时候,会自动从Bios中取得硬件时间,设置为系统时间.
  2. Linux系统时间的设置

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    //查看时间
    [root@localhost ~]# date
    2008年 12月 12日 星期五 14:44:12 CST

    //修改时间
    [root@localhost ~]# date --set "1/1/09 00:01" ==>月/日/年时:分:秒
    2009年 01月 01日 星期四 00:01:00 CST

    //date 有几种时间格式可接受,这样也可以设置时间:
    [root@localhost ~]# date 012501012009.30 ==>月日时分年.秒
    2009年 01月 25日 星期日 01:01:30 CST
  3. Linux硬件时间的设置

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    硬件时间的设置,可以用hwclock或者clock命令.其中,clock和hwclock用法相近,只用一个就行,只不过clock命令除了支持x86硬件体系外,还支持Alpha硬件体系

    //查看硬件时间 可以是用 hwclock ,hwclock --show 或者 hwclock -r
    [root@localhost ~]# hwclock --show
    2008年12月12日 星期五 06时52分07秒 -0.376932 seconds

    //设置硬件时间
    [root@localhost ~]# hwclock --set --date="1/25/09 00:00" <== 月/日/年时:分:秒
    [root@localhost ~]# hwclock
    2009年01月25日 星期日 00时00分06秒 -0.870868 seconds
  4. Linux系统时间和硬件时间的同步

    1
    2
    3
    4
    5
    6
    7
    8
    9
    同步系统时间和硬件时间,可以使用hwclock命令:

    //以系统时间为基准,修改硬件时间
    [root@localhost ~]# hwclock --systohc ==> sys(系统时间)to(写到)hc(Hard Clock)
    [root@localhost ~]# hwclock -w

    //以硬件时间为基准,修改系统时间
    [root@localhost ~]# hwclock --hctosys
    [root@localhost ~]# hwclock -s
  5. 不同机器之间的时间同步

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    为了避免主机时间因为长期运作下所导致的时间偏差,进行时间同步(synchronize)的工作是非常必要的.
    Linux系统下,一般使用ntp服务器来同步不同机器的时间.一台机器,可以同时是ntp服务器和ntp客户机.在网络中,推荐使用像DNS服务器一样分层的时间服务器来同步时间.
    同步时间,可以使用ntpdate命令,也可以使用ntpd服务

    //使用ntpdate比较简单,格式如下:
    [root@linux ~]# ntpdate [-nv] [NTP IP/hostname]
    [root@linux ~]# ntpdate 192.168.0.1
    [root@linux ~]# ntpdate time.ntp.org

    但这样的同步,只是强制性的将系统时间设置为ntp服务器时间.如果cpu tick有问题,只是治标不治本.所以,一般配合cron命令,来进行定期同步设置.比如,在crontab中添加:
    0 12 * * * * /usr/sbin/ntpdate time.ntp.org

    //使用ntpd服务,要好于ntpdate加cron的组合
    因为,ntpdate同步时间,会造成时间的跳跃,对一些依赖时间的程序和服务会造成影响.比如: sleep,timer等.而且,ntpd服务可以在修正时间的同时,修正cpu tick.理想的做法为,在开机的时候,使用ntpdate强制同步时间,在其他时候使用ntpd服务来同步时间.
    要注意的是,ntpd有一个自我保护设置: 如果本机与上源时间相差太大,ntpd不运行.所以新设置的时间服务器一定要先 ntpdate从上源取得时间初值,然后启动ntpd服务.ntpd服务运行后,先是每64秒与上源服务器同步一次,根据每次同步时测得的误差值经复杂计算逐步调整自己的时间,随着误差减小,逐步增加同步的间隔.每次跳动,都会重复这个调整的过程.
  6. ntpd服务的设置

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    40
    41
    42
    43
    44
    45
    46
    47
    48
    49
    50
    51
    52
    53
    54
    55
    56
    57
    58
    59
    60
    //ntpd服务的相关设置文件
    * /etc/ntp.conf
    这个是NTP daemon的主要设文件,也是 NTP 唯一的设定文件
    * /usr/share/zoneinfo/
    在这个目录下的文件其实是规定了各主要时区的时间设定文件(例如:北京地区的时区设定文件在/usr/share/zoneinfo/Asia/Beijing).
    这个目录里面的文件与下面要谈的两个文件(clock与localtime)是有关系的
    * /etc/sysconfig/clock
    这个文件其实也不包含在NTP的daemon当中,因为这个是linux的主要时区设定文件.
    每次开机后,Linux会自动的读取这个文件来设定自己系统所默认要显示的时间
    * /etc/localtime
    这个文件就是"本地端的时间配置文件".
    刚刚那个clock文件里面规定了使用的时间设置文件(ZONE)为"/usr/share/zoneinfo/Asia/Beijing",所以说,这就是本地端的时间了.此时,Linux系统就会将Beijing那个文件另存为一份/etc/localtime文件,所以未来我们的时间显示就会以Beijing那个时间设定文件为准
    * /etc/timezone
    系统时区文件


    下面重点说说"/etc/ntp.conf"文件的设置: 在NTP Server的设定上面,其实最好不要对Internet无限制的开放,尽量仅提供您自己内部的Client端联机进行网络校时就好.此外,NTP Server也是需要网络上面较为准确的主机来自行更新自己的时间,所以在我们的NTP Server上面也要找一部最靠近自己的Time Server来进行自我校正.事实上,NTP这个服务也是C/S的一种模式

    [root@linux ~]# vi /etc/ntp.conf
    # 1.关于权限设定部分
    # 权限设定主要以"restrict"这个参数来设定,主要语法为:
    #  restrict {IP_addr} mask {net_mast} parameter
    #  * IP 可以是软件地址,也可以是"default","default"类似"0.0.0.0"
    #  * paramter则有:
    #   * ignore: 关闭所有的NTP联机服务
    #   * nomodify: 表示Client端不能更改Server端的时间参数.不过,Client端仍然可以透过Server端来进行网络校时
    #   * notrust: 该Client除非通过认证,否则该Client来源将被视为不信任网域
    #   * noquery: 不提供Client端的时间查询
    #   * notrap: 不提供trap这个远程事件登入
    #   * Kod: kod技术可以阻止"Kiss of Death"包对服务器的破坏
    #   * Nopeer: 不与其他同一层的NTP服务器进行时间同步
    #   如果paramter完全没有设定,那就表示该IP(或网域)没有任何限制

    restrict default nomodify notrap noquery  #关闭所有的NTP要求封包
    restrict 127.0.0.1    #这是允许本地查询
    restrict 192.168.0.1 mask 255.255.255.0 nomodify
    # 在192.168.0.1/24网段内的服务器就可以通过这台NTPServer进行时间同步了

    # 2.上层主机的设定
    # 要设定上层主机主要以server这个参数来设定,语法为:
    #  server [IP|HOST Name] [prefer]
    #  * Server后面接的就是我们上层"Time Server",如果Server参数后面加上"perfer"的话,那表示我们的NTP主机主要以该部主机来作为时间校正的对应.另外,为了解决更新时间封包的传送延迟动作,所以可以使用"driftfile"来规定我们的主机在与Time Server"沟通时所花费的时间,可以记录在"driftfile"后面接的文件内.
    # * prefer 表示优先级最高
    # * burst 当一个运程NTP服务器可用时,向它发送一系列的并发包进行检测
    # * iburst 当一个运程NTP服务器不可用时,向它发送一系列的并发包进行检测
    # 默认情况小15分钟后才会与上层NTP服务器进行时间校对
    # 例如:我们的NTP server与cn.pool.ntp.org联机时所花费的时间会记录在"/etc/ntp/drift"文件内
    server 0.pool.ntp.org
    server 1.pool.ntp.org
    server 2.pool.ntp.org
    server cn.pool.ntp.org prefer

    # 3.其他设置值,以系统默认值即可
    server 127.127.1.0 # local clock
    fudge 127.127.1.0 stratum 10
    driftfile /var/lib/ntp/drift
    broadcastdelay 0.008
    keys /etc/ntp/keys

    综上,restrict用来设置访问权限,server用来设置上层时间服务器,driftfile用来设置保存漂移时间的文件
  7. ntp服务的启动与观察

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    30
    31
    32
    33
    34
    35
    36
    37
    38
    39
    40
    41
    42
    43
    44
    45
    46
    在启动NTP服务前,先对提供服务的这台主机手动的校正一次时间.()因为启动服务器,端口会被服务端占用,就不能手动同步时间了)
    [root@linux ~] # ntpdate cn.pool.ntp.org

    然后,启动ntpd服务:
    [root@linux ~] # /etc/init.d/ntpd start

    查看端口:
    [root@linux ~] # netstat -ln|grep 123
    udp 0 0 192.168.228.153:123 0.0.0.0:*
    udp 0 0 127.0.0.1:123 0.0.0.0:*
    udp 0 0 0.0.0.0:123 0.0.0.0:*
    udp 0 0 :::123 :::*

    确认我们的NTP服务器已经更新了自己的时间:
    [root@linux ~] # ntpstat
    synchronized to NTP server(127.127.1.0) at stratum 11
    time correct to within 950ms
    polling server every 64 s

    常见错误:
    25 Apr 15:30:17 ntpdate[11520]: no server suitable for synchronization found
    其实,这不是一个错误.而是由于每次重启NTP服务器之后大约要3-5分钟客户端才能与server建立正常的通讯连接.当此时用客户端连接服务端就会报这样的信息.一般等待几分钟就可以了.

    [root@linux ~] # ntptrace –n 127.0.0.1
    127.0.0.1:stratum 11, offset 0.000000,synch distance 0.950951
    222.73.214.125:stratum 2,offset –0.000787,synch distance 0.108575
    209.81.9.7:stratum 1,offset 0.000028,synch distance 0.00436,refid ‘GPS’
    #这个指令可以列出目前NTP服务器(第一层)与上层NTP服务器(第二层)彼此之间的关系


    指令"ntpq -p"可以列出目前我们的NTP与相关的上层NTP的状态,以上的几个字段的意义如下:
    * remote
    即NTP主机的IP或主机名称
    注意最左边的符号,如果由"+"则代表目前正在作用中的上层NTP,如果是"*"则表示也有连上线,不过是作为次要联机的NTP主机
    * refid 参考的上一层NTP主机的地址
    * st 即stratum阶层
    * when 几秒前曾做过时间同步更新的操作
    * poll 下次更新在几秒之后
    * reach 已经向上层NTP服务器要求更新的次数
    * delay 网络传输过程钟延迟的时间
    * offset 时间补偿的结果
    * jitter Linux系统时间与BIOS硬件时间的差异时间

    最后提及一点,ntp服务,默认只会同步系统时间!!!
    如果想要让ntp同时同步硬件时间,可以设置/etc/sysconfig/ntpd文件
    在/etc/sysconfig/ntpd文件中,添加"SYNC_HWCLOCK=yes".这样,就可以让硬件时间与系统时间一起同步
  8. 快速部署

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    21
    22
    23
    24
    25
    26
    27
    28
    29
    // 私有ntp服务器,同步公网的ntp服务器
    * 编辑配置
    /etc/ntp.conf:
    # 定义服务器
    server pool.ntp.org
    restrict default nomodify notrap noquery

    #从192.168.0.1-192.168.0.254的服务器都可以使用我们的NTP服务器来同步时间
    restrict 192.168.0.0 mask 255.255.255.0 notrust nomodify notrap

    server 127.127.1.0
    fudge 127.127.1.0 stratum 8 当NTP服务不可用时,使用NTP服务器当前时间同步给客户机

    #注释掉
    #restrict default ignore

    * 启动:
    /etc/init.d/ntpd start
    chkconfig ntpd on

    默认情况下,我们配置的NTP服务器不会去时间源那里同步时间,所以必须修改"/etc/ntp/step-tickers"文件,加入我们的时间源,这样每次通过"/etc/init.d/ntpd"来启动服务的时候就会自动更新时间了

    //配置LINUX客户端
    在linux客户端上执行"ntpdate ntp_server_ip"就可以根据时间服务器统一局域网的时间了,将上面的命令放到cron里每天早上3点定期执行,crontab –e 然后输入:
    0 3 * * * /usr/sbin/ntpdate 192.168.0.179


    //防火墙
    iptables -I INPUT -p udp -m udp --sport 123 -j ACCEPT

  1. 安装

    1
    yum -y install ntp ntpdate
  2. 配置参数

    1
    2
    3
    4
    5
    6
    7
    ignore  :关闭所有的 NTP 联机服务
    nomodify:客户端不能更改服务端的时间参数,但是客户端可以通过服务端进行网络校时
    notrust :客户端除非通过认证,否则该客户端来源将被视为不信任子网
    noquery :不提供客户端的时间查询:用户端不能使用ntpq,ntpc等命令来查询ntp服务器
    notrap :不提供trap远端登陆:拒绝为匹配的主机提供模式 6 控制消息陷阱服务。陷阱服务是 ntpdq 控制消息协议的子系统,用于远程事件日志记录程序。
    nopeer :用于阻止主机尝试与服务器对等,并允许欺诈性服务器控制时钟
    kod : 访问违规时发送 KoD 包。
  3. 通用配置文件

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    /etc/ntp.conf:
    driftfile /var/lib/ntp/drift
    restrict default nomodify notrap nopeer noquery
    restrict 127.0.0.1
    restrict ::1
    server 0.centos.pool.ntp.org iburst
    server 1.centos.pool.ntp.org iburst
    server 2.centos.pool.ntp.org iburst
    server 3.centos.pool.ntp.org iburst
    includefile /etc/ntp/crypto/pw
    keys /etc/ntp/keys
    disable monitor
  4. 限制网段访问

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    /etc/ntp.conf:
    driftfile /var/lib/ntp/drift
    restrict default nomodify notrap
    restrict 127.0.0.1
    restrict ::1
    server 0.centos.pool.ntp.org iburst
    server 1.centos.pool.ntp.org iburst
    server 2.centos.pool.ntp.org iburst
    server 3.centos.pool.ntp.org iburst
    includefile /etc/ntp/crypto/pw
    keys /etc/ntp/keys
    disable monitor
  5. 启动ntp服务

    1
    2
    3
    4
    5
    6
    7
    8
    systemctl start ntpd
    systemctl enable ntpd

    iptables -A INPUT -p UDP -i eno16777736 -s 192.168.1.0/24 --dport 123 -j ACCEPT
    setsebool -P ntp_disable_trans

    vi /etc/sysconfig/ntpd #允许BIOS与系统时间同步,添加下面一行。
    SYNC_HWCLOCK=yes
  6. 检测ntp服务是否运行

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    //检测ntp服务是否运行
    netstat -tlunp | grep ntp

    //检测有无和上层ntp联通
    ntpstat

    //检测ntp和上层ntp的联通状态
    ntpq -p

    remote - 本机和上层ntp的ip或主机名,“+”表示优先,“*”表示次优先
    refid - 参考上一层ntp主机地址
    st - stratum阶层
    when - 多少秒前曾经同步过时间
    poll - 下次更新在多少秒后
    reach - 已经向上层ntp服务器要求更新的次数
    delay - 网络延迟
    offset - 时间补偿
    jitter - 系统时间与bios时间差
  7. 客户端配置时间同步

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    //与ntp服务端同步
    ntpdate 192.168.1.101

    //写入到bios
    hclock -w

    //写入到计划任务
    /etc/crontab:
    01 0 * * * root /usr/sbin/ntpdate 192.168.1.101 2>&1 >/dev/null

    //重启crond计划任务
    /etc/init.d/crond restart