What happened:
When setting up Teleport, you can get a warning message that the host identification has changed. While many online suggestions suggest removing the server from known hosts. This is a bit of a hack and doesn't really solve the underlying issue.
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@ WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED! @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
IT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY!
Someone could be eavesdropping on you right now (man-in-the-middle attack)!
Problem.
To solve this, users of the TSH/ OpenSSH client needs to add the authentication server to
Login to Authentication Server.
Run tctl auth export --type=host This will explore the cert authority, that'll need to be copied into ~/.ssh/known_hosts
BUG Note: that the cert-authority host defaults to the node name. This should match to the domain for the host. e.g. for teleport.practice.io, it should be *.practice.io If I was just using a test AWS ELB, I would have to use something like practice-proxy-fdsfds34.elb.us-west-2.amazonaws.com I would have to use *.elb.us-west-2.amazonaws.com
What happened: When setting up Teleport, you can get a warning message that the host identification has changed. While many online suggestions suggest removing the server from known hosts. This is a bit of a hack and doesn't really solve the underlying issue.
Problem.
To solve this, users of the TSH/ OpenSSH client needs to add the authentication server to
tctl auth export --type=host
This will explore the cert authority, that'll need to be copied into~/.ssh/known_hosts
BUG Note: that the cert-authority host defaults to the node name. This should match to the domain for the host. e.g. for
teleport.practice.io
, it should be*.practice.io
If I was just using a test AWS ELB, I would have to use something likepractice-proxy-fdsfds34.elb.us-west-2.amazonaws.com
I would have to use*.elb.us-west-2.amazonaws.com
^ above default to
nodename
, should be a wildcard on the proxy domain.What you expected to happen:
We should make it clear how the certificate authority works, maybe pull in the DNS and information about SSH
principals
.We could also provide the @cert-authority via the UI for end users to not encounter the TOFU problem.