It was found that OpenSSH would allow an attacker to request a large number of keyboard-interactive devices when entering a password, which could allow a remote attacker to bypass the MaxAuthTries limit defined in the sshd_config file. This flaw only affects OpenSSH configurations that have the 'KbdInteractiveAuthentication' configuration option set to 'yes'. By default, this option has the same value as the 'ChallengeResponseAuthentication' option. By default, all versions of Red Hat Enterprise Linux have the 'ChallengeResponseAuthentication' option set to 'no', meaning default OpenSSH configurations are not affected by this flaw. Upstream patch: http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.bin/ssh/auth2-chall.c.diff?r1=1.42&r2=1.43&f=h CVE assignment: http://seclists.org/oss-sec/2015/q3/173 External References: https://kingcope.wordpress.com/2015/07/16/openssh-keyboard-interactive-authentication-brute-force-vulnerability-maxauthtries-bypass/
Created openssh tracking bugs for this issue: Affects: fedora-all [bug 1245971]
As noted in comment 0 above, this issue does not affect the default OpenSSH sshd configuration in Red Hat Enterprise Linux 4, 5, 6, and 7, because the default sshd_config (/etc/ssh/sshd_config) explicitly disables ChallengeResponseAuthentication using "ChallengeResponseAuthentication no". Only configurations with "ChallengeResponseAuthentication yes" are affected by this issues. Note that the lack of the ChallengeResponseAuthentication configuration directive in sshd_config does not imply that the ChallengeResponseAuthentication is disabled, as its default value is "yes". This issue affects keyboard-interactive authentication method. One simple way to check if a system may be affected by this issue is to connect to it using ssh client running in verbose mode and check if keyboard-interactive is listed among authentication methods that server offered. For example, run: $ ssh -v test.0.1 In the generated debug output, search for line as: debug1: Authentications that can continue: publickey,gssapi-keyex,gssapi-with-mic,password or debug1: Authentications that can continue: publickey,gssapi-keyex,gssapi-with-mic,password,keyboard-interactive If the list does not include "keyboard-interactive", it is not affected by this issue because of its configuration. Systems with keyboard-interactive method enabled may be affected. The default PAM configuration also mitigates impact of this flaw, as pam_unix is configured to add 2 second delay after unsuccessful authentication. This limits the number of password prompts displayed before LoginGraceTime (the default value is 120 seconds) is reached. I.e. no more than 60 password guesses per connection should be expected rather than thousands mentioned in the report for FreeBSD. This issue can be mitigated by disabling ChallengeResponseAuthentication in the sshd_config configuration file and restarting the sshd service. The impact of this issue is limited to increasing efficiency of the brute-force password guessing attacks. Given the limited impact on Red Hat Enterprise Linux and the fact that default configuration is not affected, this is not planned to be fixed immediately, but may be corrected in future OpenSSH updates in Red Hat Enterprise Linux 6 and 7. This issue is not planned to be corrected in Red Hat Enterprise Linux 5 and earlier. Mitigation: This issue can be mitigated by disabling keyboard-interactive authentication method. That can be achieved by setting "ChallengeResponseAuthentication no" in the /etc/ssh/sshd_config configuration file and restarting the sshd service. Statement: This issue does not affect the default OpenSSH sshd configuration in Red Hat Enterprise Linux 4, 5, 6 and 7.
openssh-6.9p1-3.fc22 has been pushed to the Fedora 22 stable repository. If problems still persist, please make note of it in this bug report.
This issue was fixed upstream in OpenSSH 7.0: http://www.openssh.com/txt/release-7.0
openssh-6.6.1p1-16.fc21 has been pushed to the Fedora 21 stable repository. If problems still persist, please make note of it in this bug report.
This issue has been addressed in the following products: Red Hat Enterprise Linux 7 Via RHSA-2015:2088 https://rhn.redhat.com/errata/RHSA-2015-2088.html
Nice that this was fixed in RHEL7. What is the status of the fix for RHEL6 - and why not fix for RHEL5?
This issue has been addressed in the following products: Red Hat Enterprise Linux 6 Via RHSA-2016:0466 https://rhn.redhat.com/errata/RHSA-2016-0466.html