One of the biggest challenges we have faced so far with our Dynatrace deployment is attempting to figure out the best approach to a tenant or tenants. We have a mixed bag of Google Cloud Platform, AWS, and a ton of on-prem. Some of our cloud platforms have dependencies that reside on-prem while others are 100% cloud.
We are not huge by any means. We are talking under a thousand servers all of which include virtual and physical. We would like to have a single dashboard for our Operations Center but the minute we break into the idea of more than one tenant, this of course becomes a huge challenge. Also, because tenants are not aware of one another, backtrace becomes restrained when dealing with more than one tenant.
I am very curious to know what route others on here have taken and why?
If you have a mixed bag of cloud and on-prem, do you separate them in terms of tenants?
What has been your deciding factors?
Solved! Go to Solution.
Since you clearly want to have a single dashboard and if you want to have it in Dynatrace, you need to have a single tenant. Management zones will probably help you a lot if you need to slice your environment (for example for maintaining access to different areas for different groups of responsibility).
I think having multiple tenants fits only in cases if they monitor completely different environments (if you monitor environments of different customers). If you have at least something in common in those environments, I'd go for a single tenant definitely.
Maintaining a large tenant depends on the complexity. I'd recommend to establish a "best practice" and automate where possible (e.g. establish correct automatic tagging, process group identification) preferably during deployment, not afterwards. But this really depends on your particular environment and complexity. I'd just recommend having a simple spare tenant for testing purposes (such as rules testing).
Just my 2 cents.
Hi Larry! Long time no see. The situation you're in is nothing exotic. Most of our mid-sized to large accounts face the same questions for varying reasons and in many of those cases, it is perfectly valid to split your IT landscape into multiple Dynatrace environments.
To still provide a single pane of glass, we're working on cross-environment dashboards (see https://www.dynatrace.com/news/blog/reporting-at-scale-leveraging-cross-environment-dashboards-eap) and other cross-environment integrations (e.g. service flows). The former is already available in an early adopter version and should go GA in Q2 this year.
Do you see any requirements in particular in your situation, that we might have missed?