I recently compiled 4.7p1 in anticipation of upgrading our current
(3.6.1p2) running version -- by no means ahead of schedule, granted.
However, I seem to be unable to get hostbased authenticaton to work.
This should have been a simple question of checking sshd_config,
ensuring correct host keys are present, and checking file permissions.
Failing that, verbose output from a test client and daemon should
be able to tell me exactly where the problem lies. Not so fast, though:
sshd-4.7p1 has the following to say when invoked with 'sshd -Dddd -p2222 -f
<sshd_config provided below>' and ssh 3.6.1p2 is called with 'ssh -vvv
serverhost -p2222':
<snip daemon setup and publickey, which I don't try for here:>
debug2: input_userauth_request: try method hostbased
debug1: userauth_hostbased: cuser root chost clienthost. pkalg ssh-dss slen 55
debug3: mm_key_allowed entering
debug3: mm_request_send entering: type 20
debug3: monitor_read: checking request 20
debug3: mm_answer_keyallowed entering
debug3: mm_answer_keyallowed: key_from_blob: 0x14267f60
debug2: userauth_hostbased: chost clienthost. resolvedname clienthost.uio.no
ipaddr <IPv4addr>
debug2: auth_rhosts2: clientuser root hostname clienthost. ipaddr clienthost.
debug1: temporarily_use_uid: 0/0 (e=0/0)
debug1: restore_uid: 0/0
debug1: temporarily_use_uid: 0/0 (e=0/0)
debug1: restore_uid: 0/0
Failed hostbased for root from <IPv4addr> port 712 ssh2
debug3: mm_answer_keyallowed: key 0x14267f60 is disallowed
debug3: mm_request_send entering: type 21
debug3: mm_request_receive entering
debug3: mm_key_allowed: waiting for MONITOR_ANS_KEYALLOWED
debug3: mm_request_receive_expect entering: type 21
debug3: mm_request_receive entering
debug2: userauth_hostbased: authenticated 0
debug1: userauth-request for user root service ssh-connection method hostbased
debug1: attempt 3 failures 3
debug2: input_userauth_request: try method hostbased
debug1: userauth_hostbased: cuser root chost clienthost. pkalg ssh-rsa slen 143
debug3: mm_key_allowed entering
debug3: mm_request_send entering: type 20
debug3: monitor_read: checking request 20
debug3: mm_answer_keyallowed entering
debug3: mm_answer_keyallowed: key_from_blob: 0x142679d0
debug2: userauth_hostbased: chost clienthost. resolvedname clienthost.uio.no
ipaddr <IPv4addr>
debug2: auth_rhosts2: clientuser root hostname clienthost. ipaddr clienthost.
debug1: temporarily_use_uid: 0/0 (e=0/0)
debug1: restore_uid: 0/0
debug1: temporarily_use_uid: 0/0 (e=0/0)
debug1: restore_uid: 0/0
Failed hostbased for root from <IPv4addr> port 712 ssh2
debug3: mm_answer_keyallowed: key 0x142679d0 is disallowed
debug3: mm_request_send entering: type 21
debug3: mm_request_receive entering
debug3: mm_key_allowed: waiting for MONITOR_ANS_KEYALLOWED
debug3: mm_request_receive_expect entering: type 21
debug3: mm_request_receive entering
debug2: userauth_hostbased: authenticated 0
Connection closed by <IPv4addr>
debug1: do_cleanup
ssh on the clienthost should have its say, too:
debug2: we sent a publickey packet, wait for reply
debug1: Authentications that can continue: publickey,password,hostbased
debug2: we did not send a packet, disable method
debug3: authmethod_lookup hostbased
debug3: remaining preferred: keyboard-interactive,password
debug3: authmethod_is_enabled hostbased
debug1: Next authentication method: hostbased
debug2: userauth_hostbased: chost clienthost.
debug2: we sent a hostbased packet, wait for reply
debug1: Authentications that can continue: publickey,password,hostbased
debug2: userauth_hostbased: chost clienthost.
debug2: we sent a hostbased packet, wait for reply
debug1: Authentications that can continue: publickey,password,hostbased
debug1: No more client hostkeys for hostbased authentication.
debug2: we did not send a packet, disable method
debug3: authmethod_lookup password
debug3: remaining preferred: ,password
debug3: authmethod_is_enabled password
debug1: Next authentication method: password
root@serverhost's password:
My sshd_config looks like this:
# $OpenBSD: sshd_config,v 1.75 2007/03/19 01:01:29 djm Exp $
# This is the sshd server system-wide configuration file. See
# sshd_config(5) for more information.
# This sshd was compiled with PATH=/usr/bin:/bin:/usr/sbin:/sbin:/local/bin
# The strategy used for options in the default sshd_config shipped with
# OpenSSH is to specify options with their default value where
# possible, but leave them commented. Uncommented options change a
# default value.
Port 22
#AddressFamily any
#ListenAddress 0.0.0.0
#ListenAddress ::
# Disable legacy (protocol version 1) support in the server for new
# installations. In future the default will change to require explicit
# activation of protocol 1
Protocol 2
# HostKey for protocol version 1
#HostKey /etc/ssh/ssh_host_key
# HostKeys for protocol version 2
HostKey /etc/ssh/ssh_host_rsa_key
HostKey /etc/ssh/ssh_host_dsa_key
#(COMMENT: I included this on the go because I suspected DNS mismatch between
#clienthost and clienthost.domain.name was the cause of the problem.)
HostbasedUsesNameFromPacketOnly yes
# Lifetime and size of ephemeral version 1 server key
#KeyRegenerationInterval 1h
#ServerKeyBits 768
# Logging
# obsoletes QuietMode and FascistLogging
#SyslogFacility AUTH
#LogLevel INFO
# Authentication:
#LoginGraceTime 2m
PermitRootLogin yes
#StrictModes yes
MaxAuthTries 10
#RSAAuthentication yes
#PubkeyAuthentication yes
#AuthorizedKeysFile .ssh/authorized_keys
#(COMMENT: I dont't want protocol version 1):
# For this to work you will also need host keys in /etc/ssh/ssh_known_hosts
RhostsRSAAuthentication no
# similar for protocol version 2
#(If I had a dime for each time I checked this one...)
HostbasedAuthentication yes
# Change to yes if you don't trust ~/.ssh/known_hosts for
# RhostsRSAAuthentication and HostbasedAuthentication
#IgnoreUserKnownHosts no
# Don't read the user's ~/.rhosts and ~/.shosts files
#(COMMENT: I trust my users to do this)
IgnoreRhosts no
# To disable tunneled clear text passwords, change to no here!
PasswordAuthentication yes
#PermitEmptyPasswords no
# Change to no to disable s/key passwords
#(COMMENT: I believe this should be disabled if hostbased is to work)
ChallengeResponseAuthentication no
# Kerberos options
#KerberosAuthentication no
#KerberosOrLocalPasswd yes
#KerberosTicketCleanup yes
#KerberosGetAFSToken no
# GSSAPI options
#GSSAPIAuthentication no
#GSSAPICleanupCredentials yes
# Set this to 'yes' to enable PAM authentication, account processing,
# and session processing. If this is enabled, PAM authentication will
# be allowed through the ChallengeResponseAuthentication and
# PasswordAuthentication. Depending on your PAM configuration,
# PAM authentication via ChallengeResponseAuthentication may bypass
# the setting of "PermitRootLogin without-password".
# If you just want the PAM account and session checks to run without
# PAM authentication, then enable this but set PasswordAuthentication
# and ChallengeResponseAuthentication to 'no'.
UsePAM yes
#COMMENT: (I wish to enable PAM, but I can renege on this for the moment.
#In the long run, it will have to work. Disabling it doesn't seem to make
#a difference right now anyhow.)
XAuthLocation /usr/bin/xauth
#AllowTcpForwarding yes
#GatewayPorts no
X11Forwarding yes
X11DisplayOffset 10
X11UseLocalhost no
PrintMotd yes
#PrintLastLog yes
#TCPKeepAlive yes
#UseLogin no
#UsePrivilegeSeparation yes
#PermitUserEnvironment no
#Compression delayed
#ClientAliveInterval 0
#ClientAliveCountMax 3
#UseDNS yes
#PidFile /var/run/sshd.pid
#MaxStartups 10
#PermitTunnel no
# no default banner path
#Banner /some/path
# override default of no subsystems
Subsystem sftp /local/libexec/sftp-server
# Example of overriding settings on a per-user basis
#Match User anoncvs
# X11Forwarding no
# AllowTcpForwarding no
# ForceCommand cvs server
My /etc/ssh/ directory looks like this, for those curious about
my file permissions:
-rw-r--r-- 1 root root 15 2008-02-26 16:53 hosts.equiv
-rw------- 1 root root 132839 2007-07-13 17:01 moduli
-rw-r--r-- 1 root root 15 2008-02-26 15:02 shosts.equiv
-rw-r--r-- 1 root root 1600 2008-02-26 16:07 ssh_config
-rw-r--r-- 1 root root 1495 2008-02-25 17:50 ssh_config.old
-rw-r--r-- 1 root root 3289 2008-02-26 17:24 sshd_config
-rw-r--r-- 1 root root 3289 2008-02-26 17:23 sshd_config.current
-rw-r--r-- 1 root root 3297 2008-02-25 18:10 sshd_config.old
-rw------- 1 root root 668 2008-02-25 17:01 ssh_host_dsa_key
-rw-r--r-- 1 root root 610 2008-02-25 17:01 ssh_host_dsa_key.pub
-rw------- 1 root root 983 2008-02-25 17:01 ssh_host_key
-rw-r--r-- 1 root root 647 2008-02-25 17:01 ssh_host_key.pub
-rw------- 1 root root 1675 2008-02-25 17:01 ssh_host_rsa_key
-rw-r--r-- 1 root root 402 2008-02-25 17:01 ssh_host_rsa_key.pub
-rw-r--r-- 1 root root 230 2008-02-26 15:12 ssh_known_hosts
-rw-r--r-- 1 root root 229 2008-02-26 15:12 ssh_known_hosts2
-rw-r--r-- 1 root root 1333 2008-02-26 15:11 ssh_known_hosts2.bak
-rw-r--r-- 1 root root 474 2008-02-26 15:11 ssh_known_hosts.bak
Whereas ssh on the clienthost, naturally, is setuid root (as it should be
to work with hostbased for root to other hosts).
I've tried copying my /.rhosts into /.shosts. I've checked and double-checked
that the host keys are exactly the same, and I've also used ssh-keyscan
(does it still assume rsa1 is the default?) to ensure that they're
updated. Somewhere along the way I came over the string "shosts.equiv",
and added the line '+ clienthost.domain.name' to it, equally for hosts.equiv.
I guess I should've assumed they wouldn't do much, but there you are.
Our old OpenSSH-3.6.1p2 installation didn't use ssh-keysign; at least, it isn't
present in /usr/libexec. ssh is setuid root on the client. It works beautifully
to other serverhosts that run the same version. I have as far as possible
attempted to mimic the existing config files of ssh and sshd of 3.6.1p2 in
4.7p1, but I am wary of trusting this approach. Under any circumstance, there
aren't *that* many options to begin with in either file which regulate hostbased
a'n, is there?
Is there an updated checklist somewhere that applies to the later versions of
OpenSSH that I can use as an authoritative resource on the issue of hostbased
a'n? The various resources available seem to diverge somewhat, and I have a
distinct feeling I'm turning knobs in mid-air. Failing that, if someone sees
glaring mistakes, now's the chance to earn eternal glory. In advance thanks for
any helpful pointers.
--
Øystein Larsen <oystein.larsen@usit.uio.no>
