Re: OpenSSH: which public keys are required/recommended?
From: Dimitri Maziuk (dima@127.0.0.1)Date: 03/28/02
- Previous message: Simon Tatham: "Re: OpenSSH: which public keys are required/recommended?"
- In reply to: Richard Silverman: "Re: OpenSSH: which public keys are required/recommended?"
- Next in thread: Dimitri Maziuk: "Re: OpenSSH: which public keys are required/recommended?"
- Next in thread: Richard Silverman: "Re: OpenSSH: which public keys are required/recommended?"
- Reply: Dimitri Maziuk: "Re: OpenSSH: which public keys are required/recommended?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
From: Dimitri Maziuk <dima@127.0.0.1> Date: Thu, 28 Mar 2002 01:17:51 +0000 (UTC)
begin 666 Richard Silverman:
>>>>>> "DM" == Dimitri Maziuk <dima@127.0.0.1> writes:
>
> DM> What I noticed after upgrading to OpenSSH 3.1 is that if a host
> DM> has DSA key but no RSA key in authorized_keys2, ssh will barf. So
> DM> there is a reason to prefer RSA keys -- they seem to work better.
> DM> I'm not sure if it is a bug or a feature.
>
> An authorized_keys file contains not host keys, but rather user keys.
> Perhaps you meant the known_hosts file?
Yes. Sorry, brane fart.
...And it would be more helpful if
> you would give an explicit error message, rather just saying it, uh --
> "barfs."
See below.
> I will make a guess, and say that perhaps you're seeing the client
> complain about not being able to confirm the server identity. If the
> server offers both DSA and RSA host keys, and you have only the DSA one,
> this will happen, because by default the client selects the RSA one.
Yep, that's what happened, with the usual "authenticity of host cannot
be established" message.
> Using "ssh -o HostKeyAlgorithms=ssh-dss ..." would get around this issue.
Well, I just generated RSA host keys for affected hosts.
<Curious>
We have a bunch of admin scripts that run via ssh from a central server
(cron jobs). I wonder what would've happened if I didn't test the upgrade
& generate missing RSA keys. Would those cron jobs just sit there waiting
for "yes or no" until cron run queue overflowed (or server's process table,
whichever comes first) a few days later?
</Curious>
Dima
--
Riding roughshod over some little used trifle like the English language is not a
big deal to an important technology innovator like Microsoft. They did just that
by naming a major project dot-Net (".Net"). Before that, a period followed by a
capital letter was used to mark a sentence boundary. --T. Gottfried, RISKS 21.91
- Next message: Peter Smith: "Multicast packet tunneling"
- Previous message: Simon Tatham: "Re: OpenSSH: which public keys are required/recommended?"
- In reply to: Richard Silverman: "Re: OpenSSH: which public keys are required/recommended?"
- Next in thread: Dimitri Maziuk: "Re: OpenSSH: which public keys are required/recommended?"
- Next in thread: Richard Silverman: "Re: OpenSSH: which public keys are required/recommended?"
- Reply: Dimitri Maziuk: "Re: OpenSSH: which public keys are required/recommended?"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Relevant Pages
|
|