Suppress initial emails on saving alert

richardelms

Hello,

I'm just looking for some information on alerting, I've configured an application monitor template which is applied over 8 servers. Four of these servers are DR, the services on them will not be online unless we fail over. However I would like to have these monitored so we know when they go online/offline, I just don't want the initial emails stating that they're offline.

I figured it out and based on the offline components, if I apply this alert it would send out 450 emails. 1 for each service to a large number of people over 2 mailing groups. I can't even have these just sent to myself as amending the recipients and applying will prompt the alerts being sent again?

I thought I would post as you guys are always so helpful, if this is in the wrong location or answered elsewhere I apologize!

kpmarcin

Hi Richard,

I would like to suggest for you a bit different approach.

Alerting for clusters is always tricky. There are few approaches, but I am personally recommending below.

I would create groups and subgroups for each cluster in the following order:

Cluster group ------ group with "show best status"

--- Server A group ------ group with "mixed status shows warning"

--- --- Server A (health status)

--- --- Services on Server A

--- Server B group ------ group with "mixed status shows warning"

--- --- Server B (health status)

--- --- Services on Server B

This way, "Cluster group" will have "Up" (green) if at least one service has status "up".

If both services (A and will have problems, than both server groups will have problem and the "best status" for cluster group will also be warning or critical.

Also if service is down and server is down, server group will not be in warning, but in critical.

Next, you can create Alert for Cluster group, which will give you full visibility into the clustered service.

You can include information about individual components of the group inside the email message for the group Alert.

Another approach is:

-----------

Cluster group ------ group with "Show worst status"

--- Servers group ------ group with "Show worst status"

--- --- Server A (health status)

--- --- Server B (health status)

--- Services group ------ group with "Show best status"

--- --- Services on Server A

--- --- Services on Server B

In this scenario, if everything is working - you will have full visibility into health of individual servers in cluster and you will see the status of your clustered service.

Next, you can create Alert for individual groups and/or the main group - depending on your needs.

This is much better than dealing with complex configuration of logical values inside the alert and is much cleaner for web console view.

Additionally you will not have any false positive email alerting when you define such alert.

Here you can set up the status group:

I hope the above will help you.

Kind regards,

Marcin Kazmierczak.

---

IT-Indago Ltd. - Authorized Reseller & SolarWinds Certified SCP Professional

IT-Indago – Be In Control!  |  Follow us on Facebook & LinkedIn