2

I originally posted this on Information Security but am hoping to attract an answer here since this is OSX specific.

So I'm fairly new to more secure forms of key management, I've been used to storing my keys inside key files on my computer. Recently I wanted to try and see if I could setup SSH authentication to my webserver using a key stored on my NitroKeyPro making my keychain more portable and secure in the process.

I followed this guide http://xmodulo.com/linux-security-with-nitrokey-usb-smart-card.html pretty much step by step but noticed that in the end, I did not need my NitroKeyPro to be inserted into my computer at all for the authentication to succeed.

I have a feeling that upon exporting my key it somehow got added to my local key storage making the NitroKey redundant but I am not knowledgable enough about the exact workings to be sure.

Would anyone be able to help me ensure that I can only SSH into my web server while my NitroKey is inserted into my computer?

Notes:

  • OS: OSX El Capitan 10.11.4
  • NitroKeyPro
  • Even while the NitroKey is inserted into my computer it does NOT ask me to enter a pin when I attempt to SSH.
  • OpenSC 0.15.0
  • gpg 2.0.28

I tried removing from ~/.ssh the following:

id.rsa
private_key.pem

after attempting to SSH to my web server again I get:

Permission denied (publickey,gssapi-keyex,gssapi-with-mic).

I assume this indicates that the SSH session cannot find my key to authenticate with, I checked if my computer was detecting the NitroKey by running:

gpg --card-status

and received card information like I would expect.

  • Is there nobody that has any experience with OSX and smartcards that can shed some light on this? – MSB May 3 '16 at 8:29

You must log in to answer this question.

Browse other questions tagged .