This message was deleted.
# general
s
This message was deleted.
s
This is the great opportunity of Platform Engineering - they can embed cost control into the platform, as with GRC and security. For example, a test environment may be created on demand with a push-button, but also be automatically cleaned after a week.
You can send automated notifications somewhere like Slack about decommissioning, possibly even providing an "extend" option.
k
Definitely time limiting resources (and having the ability to extend) is critical in managing costs. In our particular case, a test env lives about 4 hours unless extended (and our internally developed slack bot will allow the user to extend the time by responding to the warning). The other thing we do is have a looser and separately cost tracked “experimental” cloud account so people can try things without compliance roadblocks, i.e. to try out new AWS service xyz before figuring out how to integrate into the platform.
Also, any self service tooling should really be good at tagging resources in the cloud so things can be grouped to understand costs