Network Time Protocol (NTP) is the standard for keeping all servers on the Internet on time. TelVue servers reference a minimum of two NTP servers at any given time, however those NTP servers can occasionally become unavailable due to Internet traffic issues or an NTP server failure. We typically reference the NTP Pool, as the pool will hand out an NTP server address to you, and keep track of when those NTP servers become unavailable due to the aforementioned issues. Your server will then reach back to the pool and will receive a new address. The pool however is an imperfect system and can occasionally hand out bad servers repeatedly, thus some customers opt to define static servers as their NTP sources. This too is an imperfect system as those servers can also change or become unavailable.
NTP Configuration
You can ensure that you have NTP servers defined within the server interface under Config>Network>Time. Typically you’ll see entries from the NTP pool, unless something has been statically defined.
Typical pool servers that TelVue recommends are:
0.pool.ntp.org
1.pool.ntp.org
2.pool.ntp.org
3.pool.ntp.org
ntp.ubuntu.com
0.ubuntu.pool.ntp.org
1.ubuntu.pool.ntp.org
or any referenced here: http://tf.nist.gov/tf-cgi/servers.cgi
Once you set a time zone or a new NTP server, if your system time is out of sync with the NTP time, you may want to restart your server to get the server back to the proper time.
NTP Health and Email Alerts
If your server is currently referencing the pool or a static NTP server that is not available the health status will display WARNING, telling you to either check or change NTP servers. In the instance that an NTP server goes down or your server loses Internet connectivity, the server clock will go in to “free-run” mode where we continue to keep track of time-of-day on the server locally. If the server is left in free-run mode for extended periods of time it can cause time-drift, however once re-referenced to NTP the TelVue server will stabilize and again reference NTP.
The WARNING and CRITICAL NTP health message are generally benign as NTP servers control time drift, which changes very slowly in the event none of the servers are available. This is not a major factor unless you have many devices that should be synced to the same source that start drifting visibly.
Since many of the NTP servers used in production are public, they will go down for maintenance at random times. The status will typically clear on its own within a 24-hour period.