Cloudera recommends using NTP to synchronize the Hadoop cluster time

The host's NTP service did not respond to clock skew requests. Suggestions

This is a host health test that checks whether the host's system clock is out of sync with its NTP server. This test checks to see if the absolute value of the host clock deviation reported by the "ntpdc-c loopinfo" command is not too large. If the command fails or the NTP daemon for the host is not running, the test returns a health "bad".

If NTP is not used on the host, the check should be disabled for hosts that use the following configuration options. Cloudera recommends using NTP to synchronize the Hadoop cluster time.

This health test failure may indicate a problem with the host's NTP service or configuration.

You can configure the test by using host clock skew threshold host configuration settings.

Time zone change: (East Zone eight)

Cp-f/usr/share/zoneinfo/asia/shanghai/etc/localtime

#reboot

Select the Node1 node for the clock synchronization server, and the other nodes to synchronize the time to that node for the client.

Installing NTP

$ yum install ntp

启动ntp
#service ntpd restart    <===启动ntp服务

说明：NTP服务启动需要大约五分钟时间，服务启动之前，若客户端同步时间，则会出现错误“no server suitable for synchronization found”

定时同步时间：
在work02和 work03上配置crontab定时同步时间

crontab -e
00 12 * * * root /usr/sbin/ntpdate 192.168.249.131 >> /root/ntpdate.log 2>&1 

View Tasks
Crontab-l

Cloudera recommends using NTP to synchronize the Hadoop cluster time