Re: OpenSSH: hardening hostkeys permissions

2023-03-02 Thread Colin Walters
On Thu, Dec 8, 2022, at 9:51 AM, Daniel P. Berrangé wrote: > I think the "Upgrade/compatibility impact" section ought to call out the > possible risk with config mgmt tools like puppet/ansible, that might be > managing SSH host keys and their permissions/ownership So that was done with: > The

Re: OpenSSH: hardening hostkeys permissions

2022-12-08 Thread Dmitry Belyavskiy
On Thu, Dec 8, 2022 at 3:51 PM Daniel P. Berrangé wrote: > On Thu, Dec 08, 2022 at 03:41:32PM +0100, Dmitry Belyavskiy wrote: > > Dear Daniel, > > Thanks for your feedback! > > > > On Wed, Dec 7, 2022 at 2:55 PM Daniel P. Berrangé > > wrote: > > > > > On Wed, Dec 07, 2022 at 01:48:48PM +0100, Dm

Re: OpenSSH: hardening hostkeys permissions

2022-12-08 Thread Chris Adams
Once upon a time, Dmitry Belyavskiy said: > Drafted here, to be published: > https://fedoraproject.org/wiki/Changes/SSHKeySignSuidBit I guess the original idea was to reduce the setuid footprint (which is a good goal). I though host-based auth was deprecated at this point anyway - it's not enabl

Re: OpenSSH: hardening hostkeys permissions

2022-12-08 Thread Daniel P . Berrangé
On Thu, Dec 08, 2022 at 03:41:32PM +0100, Dmitry Belyavskiy wrote: > Dear Daniel, > Thanks for your feedback! > > On Wed, Dec 7, 2022 at 2:55 PM Daniel P. Berrangé > wrote: > > > On Wed, Dec 07, 2022 at 01:48:48PM +0100, Dmitry Belyavskiy wrote: > > > The problem we expect is that after revertin

Re: OpenSSH: hardening hostkeys permissions

2022-12-08 Thread Dmitry Belyavskiy
Dear Daniel, Thanks for your feedback! On Wed, Dec 7, 2022 at 2:55 PM Daniel P. Berrangé wrote: > On Wed, Dec 07, 2022 at 01:48:48PM +0100, Dmitry Belyavskiy wrote: > > The problem we expect is that after reverting the patch we can lose the > > remote access to the hosts because sshd will reject

Re: OpenSSH: hardening hostkeys permissions

2022-12-07 Thread Daniel P . Berrangé
On Wed, Dec 07, 2022 at 01:48:48PM +0100, Dmitry Belyavskiy wrote: > The problem we expect is that after reverting the patch we can lose the > remote access to the hosts because sshd will reject starting because of > group reading permissions. This should be covered by the upgrade scriptlet, > thou