[Fleet] Improve error message for expired certificate authority

[kpollich]

Currently, providing an expired certificate authority for Fleet Server, Logstash, etc will report a generic Unknown Certificate Authority error. It'd be ideal if we could detect the expiration case explicitly and provide a more specific error message.

[elasticmachine]

Pinging @elastic/fleet (Team:Fleet)

[jen-huang]

@kpollich Should we move this to fleet-server repo too?

[kpollich]

I think this one should stay in Kibana because we report the error message in the UI if I'm understanding the issue correctly.

[jen-huang]

@kpollich My understanding of the SDH is that the error was reported in the CLI when attempting to set up the Fleet Server.

There wouldn't be a way of auditing the certs on the Kibana side since the params point to directories on the host machine.

[kpollich]

Ah I misunderstood the initial SDH then. I'll transfer this now 😄

[kaanyalti]

Created this PR that shows that the error message mentioned in this issue is most probably not due to an expired CA.

https://github.com/elastic/fleet-server/pull/3423/files

[kpollich]

Thanks Kaan, that PR demonstrates the specific error cases well. Appreciate you creating it and investigating here 🙏

The error messages in question come from Go's HTTP internals, and as such we don't really have a way to replace or improve them. I'm going to close this as not planned as a result.