Best Practice for alerting delay

[leonatay]

Hi community, I noticed that there is default delay of 30 minutes for resource and slowdown alert. Is there a background on why is this the default and the best practice?

Thank you.

-Leona

[Eric_Yu · Accepted solution]

Hi Leona,

It's based off severity levels for alerts, here's the order (from highest to lowest):

• Monitoring unavailable
• Availability
• Error
• Slowdown
• Resource
• Custom
• Info

Slowdown and Resource alerts being the lowest in severity (excluding custom and info).

Those 30min that are set as the default delay are meant to avoid over notification, but you should try to adjust your alerting profiles according to your environment and your needs.

[rich20509]

Thanks for sharing these insights mate as I found it very much useful and informative. 

[PierreGutierrez · Accepted solution]

Hi @leonatay 
By default, dynatrace always provides you with alternatives based on good practices. I couldn't tell you exactly why it recommends 30 minutes, but I can share the following with you:

The Event category "Resource" and "Slowdown" are considered almost the last in severity level for dynatrace. (except "custom") https://docs.dynatrace.com/docs/shortlink/event-types 

Something important to keep in mind, the image shows us the time in which "The alerting profile will be assigned" to the Problem, it is different than "at what time the Problem is alerted".

I hope it's helpful 💪

[Mizső]

Hi @leonatay,

I think there is not any golden rule. It depends on the importance of the "attached" application. I have never used the default, my "basic" set is this, I always fine tune this set based on the app profile, behaviour and customer needs:

Monitoring available -> 0 mins

Availability -> 0 mins

Error -> 10 mins

Slowdown -> 15 mins

Resource -> 15 mins

Custom - > 0 mins (important custom alerts)

I hope it helps.

Best regards,

Mizső