[whip113]

I have 500k ids today. Now you want me to make those ids dynamic and correlate how much Splunk data with Vault audit data? And what is my pattern matching regex going to look like when the ids I'm trying to match on are randomly generated? And how do I pick out anomalous behavior from the noise I just intentionally created because Terraform can't stop leaking my secrets? And what about the performance? How does Vault scale to generate that many identities? And how do I audit my authorizations since on Vault I'd just see what groups the IDs were added to but not the groups those groups belonged to? What about replicating my authentication backends? Active Directory replication takes minutes to replicate a password in some environments, it's going to take longer to replicate a new identity and its group memberships. And while I can revoke an identity after some time, that doesn't mean existing authenticated sessions are terminated, it just means subsequent authentication with the same secret will fail.

[FLUX-YOU]

Maybe this feature isn't a good fit for your organization?

But based on what you've just asked, I'd definitely never create anything like what you're working with because it sounds like ANY addition to your infrastructure is problematic.

[infamousjoeg]

THIS! EVERYTHING THIS!