Hello,
Hello,
unfortunately the problem with the faulty monitoring is still not solved.
This suggests that my servers are unreliable and removed from the pool.
For me as an admin, every email with the messages means checking the services. All messages of the last time turned out to be wrong.
I will therefore log out my servers from the pool to relieve the monitoring.
CU
Jörg
Hi Jörg,
Over the last weeks @stevesommars has been doing a bunch of analytics work to track it down. He has found some suspicious network paths through Germany where NTP packets are sometimes dropped in a pattern that matches what the monitoring system sees.
If you stick it out a little longer we hopefully can figure it out. If you are able to work with Steve on debugging it, that might be helpful too!
Ask
Your server isn’t bad.
There are simply not enough monitors that check if you are online or not.
It’s just one and it’s failing all over the place.
I have 4 IPv4 NTP-servers and 2 of them also serve IPv6.
All IPv4 have Time-outs and go up and down like dancers, de IPv6 dance at 20 all the time.
But those servers are the same.
We simply need more monitors and not just 1, that is the whole issue.
And 1 monitor seeing an NTP-server online all the time should be enough to +10 score.
My servers are in Belgium, Netherlands, Germany and France, all in trouble when it comes to IPv4.
Some worse, some better…all jumping…it’s a monitor issue, there are simply not enough of them.
Bas.