Can we create multiple alerting profile

[dynatrace02065]

Can we create multiple alert profiles in a scenario where the server alerting is configured for one application team, and they are already receiving notifications for any issues? The infrastructure team also wants to be alerted for infrastructure-related alerts such as CPU, memory, swap, disk utilization, uptime, and system downtime. They should be able to have their own alerting and threshold setup to receive alerts, without notifying the other application team.

[AntonPineiro]

Hi,

I would say:

1. Creating a management zone for that application.
2. Creating alerting profiles filtering by that application.
3. Filtering by severy rules and event filters those alerting profiles.

You can check here which events are possible to be triggered.

Best regards

[GilesDay · Accepted solution]

We've done this with tagging, and custom event properties. In our area, the WIndows team only supports the C:, and the app team supports other drives... (don't get me started on if this is the best way, it just happens to be the way it is happening)

The server is tagged with the SupportID-OS:XYZ, SupportID-App:123;

The C: has a custom event property saying Event-Category:OS; Event-Recipient:TICKET/EMAIL/NOC

The 😧 has a custom event property saying Event-Category:App; Event-Recipient:TICKET/EMAIL/CommandCenter

We then have multiple alert profiles

OS-TICKET has all tags Event category:OS, Event-Recipient:TICKET

OS-EMAIL-group1  has all tags Event category:OS, Event-Recipient:EMAIL, supportID-OS:123 (because I cant use a dynamic lookup in the Email notification profile)

OS-EMAIL-group2...   has all tags Event category:OS, Event-Recipient:EMAIL, supportID-OS:456(because I cant use a dynamic lookup in the Email notification profile)

OS-CommandCenter:has all tags Event category:OS, Event-Recipient:CommandCenter,

App-Ticket similar to above, except Event Category is App

App-Email

App-CommandCenter

Then we have notification profiles

for the Ticket/COmmandCenter it is a custom webhook, for email it is the email one.

In the Ticket Notification profiles, we pull in the tag from the server for the SupportID-App or OS depending on the profile, for the CommandCenter we hard code the command center ID.

Its not pretty, but it seems to work with our complicated environment. If you want me to expand on something, let me know.

[AntonPineiro]

Hi,

I understand you have created metric events for C and D drives, and adding your event property, right?

Just to be sure if you have tagged drives from API without using metric events.

Best regards

[dynatrace02065]

So basically, we need to setup additional alerting profile for the same host in host group. As host can be part of only one host group. and with this we can have multiple alerting profiles with different thresholds for the same hosts.

• When CPU usage exceeds 70%, Team 1 will get an alert for hostA.
• When CPU usage exceeds 90%, Team 2 will get an alert hostA

[AntonPineiro · Accepted solution]

Hi,

Since same host can be for differents teams base on CPU, you can follow @GilesDay answer.

Create two metric events with specific CPU usage + property. Filtering alerting profiles base on that properties.

Best regards