Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Russ Allbery
Peter Jordan writes: > hmmm, although i have set supported enctypes > supported_enctypes = aes256-cts:normal > and restarted kdc nothing seens to have changed. > > After calling "kinit" klist -5e show me: > Etype (skey, tkt): Triple DES cbc mode with HMAC/sha1, Triple DES cbc > mode with HM

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Peter Jordan
Russ Allbery, Fri Jul 10 2009 19:24:52 GMT+0200 (CEST): > Peter Jordan writes: >> Russ Allbery, Fri Jul 10 2009 16:31:14 GMT+0200 (CEST): > > >> But for new installations a change is not a bad idea? > > Yeah, for new installations it's generally best to start the master key > at the strongest s

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Russ Allbery
Peter Jordan writes: > Russ Allbery, Fri Jul 10 2009 16:31:14 GMT+0200 (CEST): >> Yes. The master key isn't used on the network and changing it is >> very difficult in lenny. > But for new installations a change is not a bad idea? Yeah, for new installations it's generally best to start the ma

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Peter Jordan
Russ Allbery, Fri Jul 10 2009 16:31:14 GMT+0200 (CEST): > Peter Jordan writes: > >> Let the option >> master_key_type = des3-hmac-sha1 >> as it is? > > Yes. The master key isn't used on the network and changing it is very > difficult in lenny. But for new installations a change is not a b

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Bastian Blank
On Fri, Jul 10, 2009 at 07:31:33AM -0700, Russ Allbery wrote: > Peter Jordan writes: > > We use NFSv4. > I think the current version may have that same problem. Urgs, yes. Bastian -- There is an order of things in this universe. -- Apollo, "Who Mourns for Adonais?" stardate 346

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Russ Allbery
"Boyd Stephen Smith Jr." writes: > Russ Allbery wrote: >> But yes, you don't want to get Kerberos tickets on an insecure system. > I thought tickets only lasted for a small period of time, and could be > expired early if need be so that you could use them on insecure > machines. True, you can g

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Boyd Stephen Smith Jr.
In <87ws6gppyi@windlord.stanford.edu>, Russ Allbery wrote: >Peter Jordan writes: >> Russ Allbery, Fri Jul 10 2009 00:56:57 GMT+0200 (CEST): >>> Not without applying custom patches that are rather a hack. You can, >>> however, do PKINIT, which lets you use smart cards that can do X.509 >>> aut

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Russ Allbery
Peter Jordan writes: > Russ Allbery, Fri Jul 10 2009 00:56:57 GMT+0200 (CEST): >> Not without applying custom patches that are rather a hack. You can, >> however, do PKINIT, which lets you use smart cards that can do X.509 >> authentication (some of which are quite inexpensive these days). >> We

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Russ Allbery
Peter Jordan writes: > Russ Allbery, Fri Jul 10 2009 00:55:42 GMT+0200 (CEST): >> However, if you also have AFS, which I recall that you do, you can't >> turn it off at that level. You have to leave DES as a supported >> enctype since the AFS service key at present still has to be DES >> (althou

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Russ Allbery
Peter Jordan writes: > Let the option > master_key_type = des3-hmac-sha1 > as it is? Yes. The master key isn't used on the network and changing it is very difficult in lenny. > No change in /etc/krb5.conf required? Correct. Clients will negotiate the strongest available encryption key

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Peter Jordan
Russ Allbery, Fri Jul 10 2009 00:56:57 GMT+0200 (CEST): > Peter Jordan writes: > >> btw is it possible to use any kind of one time password mechanism with >> mit kdc? > > Not without applying custom patches that are rather a hack. You can, > however, do PKINIT, which lets you use smart cards th

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Peter Jordan
Russ Allbery, Fri Jul 10 2009 00:55:42 GMT+0200 (CEST): > Peter Jordan writes: >> Russ Allbery, Thu Jul 09 2009 21:51:50 GMT+0200 (CEST): > > > > However, if you also have AFS, which I recall that you do, you can't > turn it off at that level. You have to leave DES as a supported enctype > sin

Re: HEAD's UP: possible 0day SSH exploit in the wild

2009-07-10 Thread Peter Jordan
Russ Allbery, Fri Jul 10 2009 00:55:42 GMT+0200 (CEST): > Peter Jordan writes: >> Russ Allbery, Thu Jul 09 2009 21:51:50 GMT+0200 (CEST): > >>> Ensuring that you use AES enctypes for all keys (disable DES and >>> ideally also 3DES) > >> How? > > In /etc/krb5kdc/kdc.conf, set the supported_encty