Yesterday we received alert from zabbix that one of our CHR's restarted, but in Hyper-V is showing 497 days of uptime.
CHR itself shows 8h uptime, realizing that Zabbix read the uptime and predict that it has restarted.
But it did not restarted indeed, because all other monitors that pings to that CHR and that is connected to it did not failed.
VRRP backup also does not show that it changed to master.
I am not able to read logs because we are not sending it to greylog os something like that.
Have you seen that behavior before?