Little did I know that typing “NTP issues” into a Google search would return 14 million responses. While using an NTP server to synchronize the clocks in devices across an environment seems like a basic practice, I didn’t realize how many people would be interested in what happens when a server is unreachable, or clock drift occurs.
Mismatched timestamps can potentially create a cascading effect that can hamper everyone. Problems frequently experienced by IT and network users will range from not being able to run scheduled jobs or use cloud hosted apps, to IT finding log information where the time is so off, that you may not find the data as useful due to the mismatch in time. And, if you decide to use the data you spend of extra time manually correlating the information. If managing a network within the following two verticals, there are also these considerations:
- In a financial institution the FDA, FINRA's OATS Reporting Technical Specifications, and MiFID II mandate keeping accurate time, as possible issues can lead to fines and lengthy audits.
- Inaccurate timestamps in a military environment can lead to security issues or worse.
Due to the impact created by NTP issues, we looked at incoming TAC cases and saw a significant number of instances related to NTP issues yearly. This led us to explore the idea of using AI networking to help. Once down this path, we also saw that half of the APs being managed by HPE Aruba Networking Central are experiencing some form of NTP issue, unbeknownst to their IT organizations.
New AI-powered NTP feature in Central 2.5.7
This new feature is designed to use telemetry from each AP to look for instances where the NTP server is unreachable, which equates to a site and APs experiencing clock drift. Possible causes could be due to NTP servers going off-line, incorrectly using the wrong IP address for an intended NTP server, mismatched settings on new devices (i.e., different time zone settings), or issues within a network.
As of version 2.5.7 of HPE Aruba Networking Central you may see an AI Insight that shows when the configured NTP server is unreachable by our access points. The Insight includes the number of sites and APs affected. Drilling into the sites themselves will then allow you to see the deviation in time from AP to AP.
Recommended troubleshooting guidance is then provided based on the possible reasons for the problem. For instance, checking on the validity of the IP address entered or if the server is reachable.
NTP Server Unreachable AI Insight
If you’re new to using AIOps or HPE Aruba Networking Central, you can learn more from the following links:
