Devices take a significant amount of time to be identified as "Down" in Cacti.

[HeshanD]

Hi Guys,

Recently we found out we had a planned ISP outage and we haven't got any alert for cacti. So today I did a test and noticed a delay of approximately 8 minutes to identify the device is down and an additional 13 minutes to recognize the system as back online.

These are the alerts

For Down
======
System Error : **** Secondary (69.160.***.**) is DOWN
Reason: ICMP: Ping timed out

Average system response: 89.22 ms
System availability: 99.91 %
Total Checks Since Clear: 3282
Total Failed Checks: 3
Last Date Checked DOWN : 2024-01-10 01:07:59
Device Previously UP for: 10m
NOTES:
Device NOTES:

For Up
=====
System **** Secondary (69.160.***.**) status: UP

Current ping response: 62.17 ms
Average system response: 89.19 ms
System availability: 99.91 %
Total Checks Since Clear: 3,285
Total Failed Checks: 3
Last Date Checked UP: 2024-01-10 01:07:59
Device Previously DOWN for: 20m 27s

Snmp Info:
Name -
Location -
Uptime - ( ms)
System - SNMP not in use

NOTE:
Device Notes:



How can we enhance these times? Our clients require explanations for any downtime exceeding a minute, and Cacti currently falls short in identifying such incidents within that timeframe.

[HeshanD]

Hi Guys Any update for this?

[macan]

If I understand correctly, have a look at configuration->Settings-> Device defaults -> Up/Down settings. There are default values. Each device has own settings. Try to edit device and there is Host Failure Count