False positives on down device detection

[phoeneous]

We've recently implemented Cacti and I'm trying to put together a best practice method for simple ping monitoring. All of our devices sporadically alert me with the following alert. What causes this?

Code: Select all

Host is DOWN
Manage global status : down (uptime 163126263 -> 163126263)
Cacti host id : 4

The host is not going down, losing connectivity, or rebooting.

Each device is setup to use Ping and SNMP (ICMP) for downed device detection. I thought maybe it was polling to quickly so I changed the following values:

Ping Timeout Value = 800
Ping Retry Count = 2

[BSOD2600]

What cacti version and poller?
You using the thold plugin?

Conversely, if the uptime of a device isn't incrementing between polling intervals, something is wrong with the device's uptime implementation.

[phoeneous]

What cacti version and poller?
You using the thold plugin?

Conversely, if the uptime of a device isn't incrementing between polling intervals, something is wrong with the device's uptime implementation.

Version 0.8.7d. Poller is cmd.php. The thold plugin 0.4.1 is enabled but we havent set any threshold templates.

What data source would trigger this alert?

[BSOD2600]

Any sort of alerting would be the work of thold, since cacti doesn't do that out of the box.

Also, cacti 0.8.7e is out.