We are setting up DFM for just the "Unresponsive" events initially with email-notify being sent for both Active and Clear states.
As per the documentation DFM uses a combination of SNMP and ICMP polling to check the device availability with ICMP interval being dependent on the "Reachability" interval defined in the polling parameters.
However I have also found out that ICMP polling interval is also dependent on the "Retry" parameter ?
E.g. With Reachability set to default of 240sec, the ICMP polling will be
Retry = 1, ICMP = 220
Retry = 2, ICMP = 200,
Retry = 3, ICMP = 180,
Retry = 4, ICMP = 160
Is there a particular formula that is used to calculate ICMP Polling Interval ?
Also the Retry interval is also effecting when the Active Event notification is sent.
E.g. Assuming the first Echo Request packet doesn't get a response then if
Retry = 1, 3 ICMP Packets are sent before the notification is sent
Retry = 2, 5 ICMP packets
Retry = 3, 7 ICMP packets
Retry = 4, 9 ICMP packets
All these ICMP retry packets are being sent with an approx. interval of 9-10 sec.
How can I customize the interval for these ICMP retry packets ?
Is it possible to control, when the Email Notification is sent after all the ICMP Retry packets result in failure?
E.g. Is this scenario possible ?
ICMP Polling Interval = 180Sec
1. Device is determined to be down
2. DFM waits for X amount of seconds
3. DFM tries ICMP polling again and if fails again then Sends the notification ?
Can this parameter help ?
# How long, in seconds, an event must remain active before the
# adapter sends a notification.
eventSmoothingInterval = 0
This is from the NMSROOT\objects\smarts\conf\notifier\mail-notify.conf