|
Dear all,
I try to setup automatic mechanism to check the time accuracy and health condition from ntp.log, peerstats and graph generation. Something I want to share with you.
1. I can write small program to check the offset (or something) automatically from the log file. If the offset is greater than (for example) 1 millisecond, it will send alert to me for taking action. However, as there is network delay or machine load problem, it may have FALSE alarm. Then, this mechanism may not be useful. I still think how to guarantee and solve this problem.
2. In order to confirm the health condition continuously, it may need to have daily check on NTP statistical data. Although the offset is most likely not greater than the benchmark, the curve (unstable) cannot define the time server as in health condition. Please check the graphs here.
Do anyone have idea on automatic checking and monitoring NTP service in health condition and good response in team of the objective of 7 x 24 if possible.
Thanks and Regards,
Curtis
|
_______________________________________________ timekeepers mailing list [email protected] https://fortytwo.ch/mailman/cgi-bin/listinfo/timekeepers
