Stephen John Smoogen wrote:
> On Thu, Sep 17, 2009 at 8:21 AM, Jean-Michel Barbet
> <[log in to unmask]> wrote:
>> Hello,
>>
>> I am puzzled by this one :
>>
>> Performing a ssh-keyscan from an SL4.5 i368 server to retrieve the
>> RSA key of a newly installed server in SL5.3 x86_64 does not return
>> the right RSA public key. It works between SL4.5 servers.
> 
> What key does it return? How do they compare? Does the 5.3 box get the
> correct key? Can you check to see if a 4.8 system does betteR?

Hi Stephen,

Sorry for the delay. the returned key was completely different than
any other public keys from the same machine (I mean rsa1,dsa) and
it looked shorter... Executing ssh-keyscan from another SL5.3 box
was also returning bad keys.

I found that the origin of this issue was that I copied on the freshly
installed SL5.3 box a RSA key pair generated on a SL4.5 machine and this
seems to be the problem. If I regenerate all keys on the SL5.3 box,
ssh-keyscan retrieves the correct keys. So the issues is solved but
I cannot explain what I observed. A failure would be better than being
able to fetch a bad key without warning.

Thanks Stephen.

JM


-- 
------------------------------------------------------------------------
Jean-michel BARBET                    | Tel: +33 (0)2 51 85 84 86
Laboratoire SUBATECH Nantes France    | Fax: +33 (0)2 51 85 84 79
CNRS-IN2P3/Ecole des Mines/Universite | E-Mail: [log in to unmask]
------------------------------------------------------------------------