[Author]: Kwu
Linux configuration Time server, Big Data cluster time synchronization is very critical, error more than 500ms will have problems, usually configure the intranet time server.
This article takes CentOS6.6 as an example to introduce the configuration of intranet time server:
1. Open NTPD Service
Chkconfig ntpd onchkconfig--list ntpd
2. Configure the service side
Vi/etc/ntp.conf:
Server us.pool.ntp.org minpoll 4fudge 127.127.1.0 stratum 1restrict 127.0.0.1restrict 192.168.0.0 Mask 255.255.255.0 Nomo Dify Notrapdriftfile/var/lib/ntp/drift
Vi/etc/ntp/step-tickers
3. Configure the Client
1) vi/etc/ntp.conf
Server 192.168.2.245fudge 127.127.1.0 stratum 2restrict 127.0.0.1driftfile/var/lib/ntp/driftrestrict 192.168.2.245 Mask 255.255.255.255
Vi/etc/ntp/step-tickers file contents to the clock server IP address
192.168.2.245
2) or
* * * * * */usr/sbin/ntpdate 192.168.2.245 | Logger-t NTP
4, restart the service side and the customer's NTP service
Service NTPD Restart
5, Test time synchronization
Modification time
Date-s "2007-08-03 18:21:15"
Restore the correct time after the sync time, view
Date
Linux Configuration Time Server