Timezone mismatch causes Poller Cluster Health report unhealthy despite working

[duhow]

The problem

If MySQL / MariaDB uses a different timezone than the pollers or the main LibreNMS application does, Poller Cluster Health appears in red, as Last CheckIn time "exceeds" the allowed time (UTC -> UTC+2 = 2h exceed last checkin) , despite pollers are actually working and scanning devices properly.

Since MySQL uses NOW(), it will default to Database Timezone.

librenms/LibreNMS/service.py

Lines 808 to 811 in eb65361

	self._db.query(
	"INSERT INTO poller_cluster(node_id, poller_name, poller_version, last_report, master) "
	'values("{0}", "{1}", "{2}", NOW(), {3}) '
	'ON DUPLICATE KEY UPDATE poller_version="{2}", last_report=NOW(), master={3}; '.format(

Both application and database should be comparing against UTC in order to solve this problem.
Use UTC_TIMESTAMP() function instead of NOW().

Output of ./validate.php

N/A

What was the last working version of LibreNMS?

No response

Anything in the logs that might be useful for us?

No response