Vertica Clock Skewed

The network time protocol (NTP) daemon must be running on all of the hosts in the cluster so that their clocks are synchronized. The spread daemon relies on all of the nodes to have their clocks synchronized for timing purposes.

MC monitor's the clock skew among the hosts and will warn you if a problem emerges, so that you can adjust the clock accordingly.

There are two types of the clock skew that are important to MC and Vertica:

• MC-agent clock skew: The clock differences between the MC and the hosts over certain threshold. The default threshold is now 60 sec which will cause MC polling incorrect data from agents to monitor the system.
• DB node clock skew: the clock differences among database node's over the console.clockskew.mchost.threshold predefined in the console.properties file (default = 2 seconds).

See:
https://my.vertica.com/docs/9.1.x/HTML/index.htm#Authoring/InstallationGuide/BeforeYouInstall/ntp.htm
https://my.vertica.com/docs/9.1.x/HTML/index.htm#Authoring/InstallationGuide/BeforeYouInstall/chrony.htm

The file /opt/vconsole/config/console.properties stores the management console's configurable parameters. There are three pertinent parameters for the clock skew check:

1. console.clockskew.check: A boolean value. True: Will enable the user to run the clock check schedule job; False: The user can opt out the clock check if their hosts are ntpd-synced.
2. console.clockskew.dbnode.threshold: Numeric value representing the seconds the dbnode clock skew threshold from the medium value clock value among the db nodes. The default is 2 sec, which allows up to 4 seconds the differences between maximum to the minimum.
3. console.clockskew.job.schedule: String value that describes when the clock skew job will start. The format follows that of crontab: minute, hour, day of month, month and day of week (using space as separator). The default value is 0 15 0 * * ?, which means 12:15AM every day.