Schedules for alarms notifications in Extreme Management Center (blackout windows)

[Steve_Ballantyn]

Perhaps there is a way to do this, but I couldn't find it. I get several alarms throughout the night about 5 minutes down times that are occurring with my small remote offices that are on cable modems. The ISP seems to do maintenance in the early hours of the morning, which is fine by me. But I could do without the dozens of alarms that I get in my inbox.

Is there a way to set a blackout window for alarm notifications? Or better yet, create multiple schedules so that I can attach a blackout window to some of them - but not all of them? I would still want notifications for my important core stuff, regardless of the time of day.

I am running NetSight 7.0.4.29.

[Bob_Maheu]

Hello Steve,

There is no way to setup a 'blackout' period for the alarms.

You can however increase the SNMP timeout and retries for the devices in those remote locations allowing for a longer time for a failed poll to trigger the alarm.

The default SNMP timeouts are defined in NetSight Console -> Tools - Options -> Suite Option -> Advanced SNMP Settings:

timeout 5 seconds
retries 3

To change these values on a per device basis do the following:

1. Change to the main Properties tab and select the Access radio button.
2. Select the device in question from the Device Tree.
3. Make sure that the device you want to change are highlighted and then click the Show/Hide Table editor button to the upper right (looks like a pencil with spreadsheet).
4. A new row will appear at the bottom of the window and will allow you to change values. Select the column for Timeout and enter a new values in seconds. Hit Enter. You will see all the highlighted row above now have a green ! next to the entries that will change. Do the same thing for the retries.
5. When done making changes click in the Apply button to the upper right (Yellow arrow with Green !).

If they are only going out for 5 minutes it sound like it could just be failing on the default values and gets polled correctly on the next poll cycle. In this case you can adjust the values to maybe 10 and 5 to see if that helps.