Dynatrace cluster nodes clock out of sync

[dtdeer]

I go the warning like "Server time of server1 is out of symnc. ... Please enable NTP on all cluster nodes." According to this post, https://community.dynatrace.com/t5/Open-Q-A/time-sync/m-p/112956/highlight/true#M8618, "It's important to have NTP setup correctly on those hosts. Otherwise OneAgents will be synced with the wrong time and will produce PurePaths that might be off (in the future)."

But how do I set up NTP? I thought NTP comes with Casssandra. I checked NTP service on cluster nodes and it appeared as not installed.

++
$ nptstat
bash: nptstat: command not found...
$ systemctl status ntpd
Unit ntpd.service could not be found.
++

Any suggestions?

[y_buccellato · Accepted solution]

NTP doesn't depend on Dynatrace itself.

You or the Admin of the system that you are working on should set up NTP on Virtual Machine itself and the configuration of the NTP should reference a central NTP serves that keeps track of the current time.

You should definitely speak with one of your Org. Admins and he'll probably have the solution for you

[dtdeer]

y_buccellato,

Thanks for the quick response. I have two more questions:

1. Should I set up NTP service on ActiveGate nodes too?

2. Between ntpd and chronyd, which one works better with Dynatrace, or which one is a better choice generally?

Regards,

[Julius_Loman · Accepted solution]

1. Not necessarily, but I highly recommend having NTP on each host. Even hosts not related to Dynatrace.
2. Both will work fine, probably Chrony would be a preferred choice in recent Linux distributions.

[dtdeer]

Julius_Loman,

Thank you. I made sure chronyd running on all cluster hosts and activegates. 

Regards