The Apache Pulsar C++ Client does not verify peer TLS certificates when making HTTPS calls for the OAuth2.0 Client Credential Flow, even when tlsAllowInsecureConnection is disabled via configuration. This vulnerability allows an attacker to perform a man in the middle attack and intercept and/or modify the GET request that is sent to the ClientCredentialFlow 'issuer url'. The intercepted credentials can be used to acquire authentication data from the OAuth2.0 server to then authenticate with an Apache Pulsar cluster. An attacker can only take advantage of this vulnerability by taking control of a machine 'between' the client and the server. The attacker must then actively manipulate traffic to perform the attack. The Apache Pulsar Python Client wraps the C++ client, so it is also vulnerable in the same way. This issue affects Apache Pulsar C++ Client and Python Client versions 2.7.0 to 2.7.4; 2.8.0 to 2.8.3; 2.9.0 to 2.9.2; 2.10.0 to 2.10.1; 2.6.4 and earlier. Any users running affected versions of the C++ Client or the Python Client should rotate vulnerable OAuth2.0 credentials, including client_id and client_secret. 2.7 C++ and Python Client users should upgrade to 2.7.5 and rotate vulnerable OAuth2.0 credentials. 2.8 C++ and Python Client users should upgrade to 2.8.4 and rotate vulnerable OAuth2.0 credentials. 2.9 C++ and Python Client users should upgrade to 2.9.3 and rotate vulnerable OAuth2.0 credentials. 2.10 C++ and Python Client users should upgrade to 2.10.2 and rotate vulnerable OAuth2.0 credentials. 3.0 C++ users are unaffected and 3.0 Python Client users will be unaffected when it is released. Any users running the C++ and Python Client for 2.6 or less should upgrade to one of the above patched versions.

In Apache APISIX before 2.13.0, when decoding JSON with duplicate keys, lua-cjson will choose the last occurred value as the result. By passing a JSON with a duplicate key, the attacker can bypass the body_schema validation in the request-validation plugin. For example, `{"string_payload":"bad","string_payload":"good"}` can be used to hide the "bad" input. Systems satisfy three conditions below are affected by this attack: 1. use body_schema validation in the request-validation plugin 2. upstream application uses a special JSON library that chooses the first occurred value, like jsoniter or gojay 3. upstream application does not validate the input anymore. The fix in APISIX is to re-encode the validated JSON input back into the request body at the side of APISIX. Improper Input Validation vulnerability in __COMPONENT__ of Apache APISIX allows an attacker to __IMPACT__. This issue affects Apache APISIX Apache APISIX version 2.12.1 and prior versions.

The project received a report that all versions of Apache OpenOffice through 4.1.8 can open non-http(s) hyperlinks. The problem has existed since about 2006 and the issue is also in 4.1.9. If the link is specifically crafted this could lead to untrusted code execution. It is always best practice to be careful opening documents from unknown and unverified sources. The mitigation in Apache OpenOffice 4.1.10 (unreleased) assures that a security warning is displayed giving the user the option of continuing to open the hyperlink.

Apache Struts 2 2.3.20 through 2.3.28.1 mishandles token validation, which allows remote attackers to conduct cross-site request forgery (CSRF) attacks via unspecified vectors.

Apache JSPWiki user preferences form is vulnerable to CSRF attacks, which can lead to account takeover. Apache JSPWiki users should upgrade to 2.11.2 or later.

Schema parsing in the parquet-avro module of Apache Parquet 1.15.0 and previous versions allows bad actors to execute arbitrary code. While 1.15.1 introduced a fix to restrict untrusted packages, the default setting of trusted packages still allows malicious classes from these packages to be executed. The exploit is only applicable if the client code of parquet-avro uses the "specific" or the "reflect" models deliberately for reading Parquet files. ("generic" model is not impacted) Users are recommended to upgrade to 1.15.2 or set the system property "org.apache.parquet.avro.SERIALIZABLE_PACKAGES" to an empty string on 1.15.1. Both are sufficient to fix the issue.

By design, the JDBCAppender in Log4j 1.2.x accepts an SQL statement as a configuration parameter where the values to be inserted are converters from PatternLayout. The message converter, %m, is likely to always be included. This allows attackers to manipulate the SQL by entering crafted strings into input fields or headers of an application that are logged allowing unintended SQL queries to be executed. Note this issue only affects Log4j 1.x when specifically configured to use the JDBCAppender, which is not the default. Beginning in version 2.0-beta8, the JDBCAppender was re-introduced with proper support for parameterized SQL queries and further customization over the columns written to in logs. Apache Log4j 1.2 reached end of life in August 2015. Users should upgrade to Log4j 2 as it addresses numerous other issues from the previous versions.

Improper Authentication vulnerability in TLS origin validation of Apache Traffic Server allows an attacker to create a man in the middle attack. This issue affects Apache Traffic Server 8.0.0 to 8.1.0.

An Improper Input Validation vulnerability in DataImportHandler of Apache Solr allows an attacker to provide a Windows UNC path resulting in an SMB network call being made from the Solr host to another host on the network. If the attacker has wider access to the network, this may lead to SMB attacks, which may result in: * The exfiltration of sensitive data such as OS user hashes (NTLM/LM hashes), * In case of misconfigured systems, SMB Relay Attacks which can lead to user impersonation on SMB Shares or, in a worse-case scenario, Remote Code Execution This issue affects all Apache Solr versions prior to 8.11.1. This issue only affects Windows.

Apache Airflow providers-google's `ComputeEngineSSHHook` disables SSH host-key verification by default, exposing SSH traffic between an Airflow worker and a Compute Engine VM to in-path network attackers who can intercept or modify the session. Users are advised to upgrade to `apache-airflow-providers-google` 22.0.0 or later.

The Apache Xerces C++ XML parser on versions 3.0.0 before 3.2.5 contains a use-after-free error triggered during the scanning of external DTDs. Users are recommended to upgrade to version 3.2.5 which fixes the issue, or mitigate the issue by disabling DTD processing. This can be accomplished via the DOM using a standard parser feature, or via SAX using the XERCES_DISABLE_DTD environment variable. This issue has been disclosed before as CVE-2018-1311, but unfortunately that advisory incorrectly stated the issue would be fixed in version 3.2.3 or 3.2.4.

Connect2id Nimbus JOSE+JWT before v7.9 can throw various uncaught exceptions while parsing a JWT, which could result in an application crash (potential information disclosure) or a potential authentication bypass.

Improperly Controlled Modification of Dynamically-Determined Object Attributes vulnerability in Apache Camel Camel-Coap component. Apache Camel's camel-coap component is vulnerable to Camel message header injection, leading to remote code execution when routes forward CoAP requests to header-sensitive producers (e.g. camel-exec) The camel-coap component maps incoming CoAP request URI query parameters directly into Camel Exchange In message headers without applying any HeaderFilterStrategy.   Specifically, CamelCoapResource.handleRequest() iterates over OptionSet.getUriQuery() and calls camelExchange.getIn().setHeader(...) for every query parameter. CoAPEndpoint extends DefaultEndpoint rather than DefaultHeaderFilterStrategyEndpoint, and CoAPComponent does not implement HeaderFilterStrategyComponent; the component contains no references to HeaderFilterStrategy at all. As a result, an unauthenticated attacker who can send a single CoAP UDP packet to a Camel route consuming from coap:// can inject arbitrary Camel internal headers (those prefixed with Camel*) into the Exchange. When the route delivers the message to a header-sensitive producer such as camel-exec, camel-sql, camel-bean, camel-file, or template components (camel-freemarker, camel-velocity), the injected headers can alter the producer's behavior. In the case of camel-exec, the CamelExecCommandExecutable and CamelExecCommandArgs headers override the executable and arguments configured on the endpoint, resulting in arbitrary OS command execution under the privileges of the Camel process. The producer's output is written back to the Exchange body and returned in the CoAP response payload by CamelCoapResource, giving the attacker an interactive RCE channel without any need for out-of-band exfiltration.                                                                                                                                                                         Exploitation prerequisites are minimal: a single unauthenticated UDP datagram to the CoAP port (default 5683). CoAP (RFC 7252) has no built-in authentication, and DTLS is optional and disabled by default. Because the protocol is UDP-based, HTTP-layer WAF/IDS controls do not apply. This issue affects Apache Camel: from 4.14.0 through 4.14.5, from 4.18.0 before 4.18.1, 4.19.0. Users are recommended to upgrade to version 4.18.1 or 4.19.0, fixing the issue.

Improper authentication is possible in Apache Traffic Control versions 3.0.0 and 3.0.1 if LDAP is enabled for login in the Traffic Ops API component. Given a username for a user that can be authenticated via LDAP, it is possible to improperly authenticate as that user without that user's correct password.

A carefully crafted package/compressed file that, when unzipped/uncompressed yields the same file (a quine), causes a StackOverflowError in Apache Tika's RecursiveParserWrapper in versions 1.7-1.21. Apache Tika users should upgrade to 1.22 or later.

A carefully crafted or corrupt zip file can cause an OOM in Apache Tika's RecursiveParserWrapper in versions 1.7-1.21. Users should upgrade to 1.22 or later.

The code which checks HMAC in form submissions used String.equals() for comparisons, which results in a timing side channel for the comparison of the HMAC signatures. This could lead to remote code execution if an attacker is able to determine the correct signature for their payload. The comparison should be done with a constant time algorithm instead.

A number of HTTP endpoints in the Airflow webserver (both RBAC and classic) did not have adequate protection and were vulnerable to cross-site request forgery attacks.

It is possible to configure Apache CXF to use the com.sun.net.ssl implementation via 'System.setProperty("java.protocol.handler.pkgs", "com.sun.net.ssl.internal.www.protocol");'. When this system property is set, CXF uses some reflection to try to make the HostnameVerifier work with the old com.sun.net.ssl.HostnameVerifier interface. However, the default HostnameVerifier implementation in CXF does not implement the method in this interface, and an exception is thrown. However, in Apache CXF prior to 3.2.5 and 3.1.16 the exception is caught in the reflection code and not properly propagated. What this means is that if you are using the com.sun.net.ssl stack with CXF, an error with TLS hostname verification will not be thrown, leaving a CXF client subject to man-in-the-middle attacks.

In Apache jUDDI 3.2 through 3.3.4, if using the WADL2Java or WSDL2Java classes, which parse a local or remote XML document and then mediates the data structures into UDDI data structures, there are little protections present against entity expansion and DTD type of attacks. Mitigation is to use 3.3.5.

In Apache Karaf version prior to 3.0.9, 4.0.9, 4.1.1, when the webconsole feature is installed in Karaf, it is available at .../system/console and requires authentication to access it. One part of the console is a Gogo shell/console that gives access to the command line console of Karaf via a Web browser, and when navigated to it is available at .../system/console/gogo. Trying to go directly to that URL does require authentication. And optional bundle that some applications use is the Pax Web Extender Whiteboard, it is part of the pax-war feature and perhaps others. When it is installed, the Gogo console becomes available at another URL .../gogo/, and that URL is not secured giving access to the Karaf console to unauthenticated users. A mitigation for the issue is to manually stop/uninstall Gogo plugin bundle that is installed with the webconsole feature, although of course this removes the console from the .../system/console application, not only from the unauthenticated endpoint. One could also stop/uninstall the Pax Web Extender Whiteboard, but other components/applications may require it and so their functionality would be reduced/compromised.

Apache Struts versions 2.3 to 2.3.34 and 2.5 to 2.5.16 suffer from possible Remote Code Execution when alwaysSelectFullNamespace is true (either by user or a plugin like Convention Plugin) and then: results are used with no namespace and in same time, its upper package have no or wildcard namespace and similar to results, same possibility when using url tag which doesn't have value and action set and in same time, its upper package have no or wildcard namespace.

Deserialization of Untrusted Data Vulnerability in Apache Software Foundation Apache InLong.This issue affects Apache InLong: from 1.4.0 through 1.6.0. Attackers would bypass the 'autoDeserialize' option filtering by adding blanks. Users are advised to upgrade to Apache InLong's 1.7.0 or cherry-pick  https://github.com/apache/inlong/pull/7674 https://github.com/apache/inlong/pull/7674 to solve it.

The fix for CVE-2026-41635 was not applied to the 2.1.X and 2.2.X branches. Here was the original issue description: Apache MINA's AbstractIoBuffer.resolveClass() contains two branches, one of them (for static classes or primitive types) does not check the class at all, bypassing the classname allowlist and allowing arbitrary code to be executed. The fix checks if the class is present in the accepted class filter before calling Class.forName(). Affected versions are Apache MINA 2.1.0 <= 2.1.11, and 2.2.0 <= 2.2.6. The problem is resolved in Apache MINA 2.1.12, and 2.2.7 by applying the classname allowlist earlier. Affected are applications using Apache MINA that call IoBuffer.getObject(). Applications using Apache MINA are advised to upgrade.

The Dubbo Provider will check the incoming request and the corresponding serialization type of this request meet the configuration set by the server. But there's an exception that the attacker can use to skip the security check (when enabled) and reaching a deserialization operation with native java serialization. Apache Dubbo 2.7.13, 3.0.2 fixed this issue by quickly fail when any unrecognized request was found.

In Apache Log4j 2.x before 2.8.2, when using the TCP socket server or UDP socket server to receive serialized log events from another application, a specially crafted binary payload can be sent that, when deserialized, can execute arbitrary code.

A deserialization vulnerability existed when decode a malicious package.This issue affects Apache Dubbo: from 3.1.0 through 3.1.10, from 3.2.0 through 3.2.4. Users are recommended to upgrade to the latest version, which fixes the issue.

When handler-router component is enabled in servicecomb-java-chassis, authenticated user may inject some data and cause arbitrary code execution. The problem happens in versions between 2.0.0 ~ 2.1.3 and fixed in Apache ServiceComb-Java-Chassis 2.1.5

In Apache Dubbo, users may choose to use the Hessian protocol. The Hessian protocol is implemented on top of HTTP and passes the body of a POST request directly to a HessianSkeleton: New HessianSkeleton are created without any configuration of the serialization factory and therefore without applying the dubbo properties for applying allowed or blocked type lists. In addition, the generic service is always exposed and therefore attackers do not need to figure out a valid service/method name pair. This is fixed in 2.7.13, 2.6.10.1

A deserialization vulnerability existed in dubbo hessian-lite 3.2.12 and its earlier versions, which could lead to malicious code execution. This issue affects Apache Dubbo 2.7.x version 2.7.17 and prior versions; Apache Dubbo 3.0.x version 3.0.11 and prior versions; Apache Dubbo 3.1.x version 3.1.0 and prior versions.

In Apache Geode before v1.4.0, the TcpServer within the Geode locator opens a network port that deserializes data. If an unprivileged user gains access to the Geode locator, they may be able to cause remote code execution if certain classes are present on the classpath.

In Apache Linkis <=1.3.1, because the parameters are not effectively filtered, the attacker uses the MySQL data source and malicious parameters to configure a new data source to trigger a deserialization vulnerability, eventually leading to remote code execution. Versions of Apache Linkis <= 1.3.0 will be affected. We recommend users upgrade the version of Linkis to version 1.3.2.

Apache Dubbo is a java based, open source RPC framework. Versions prior to 2.6.10 and 2.7.10 are vulnerable to pre-auth remote code execution via arbitrary bean manipulation in the Telnet handler. The Dubbo main service port can be used to access a Telnet Handler which offers some basic methods to collect information about the providers and methods exposed by the service and it can even allow to shutdown the service. This endpoint is unprotected. Additionally, a provider method can be invoked using the `invoke` handler. This handler uses a safe version of FastJson to process the call arguments. However, the resulting list is later processed with `PojoUtils.realize` which can be used to instantiate arbitrary classes and invoke its setters. Even though FastJson is properly protected with a default blocklist, `PojoUtils.realize` is not, and an attacker can leverage that to achieve remote code execution. Versions 2.6.10 and 2.7.10 contain fixes for this issue.

The camel-hessian component in Apache Camel 2.x before 2.19.4 and 2.20.x before 2.20.1 is vulnerable to Java object de-serialisation vulnerability. De-serializing untrusted data can lead to security flaws.

Apache Dubbo prior to 2.6.9 and 2.7.9 by default supports generic calls to arbitrary methods exposed by provider interfaces. These invocations are handled by the GenericFilter which will find the service and method specified in the first arguments of the invocation and use the Java Reflection API to make the final call. The signature for the $invoke or $invokeAsync methods is Ljava/lang/String;[Ljava/lang/String;[Ljava/lang/Object; where the first argument is the name of the method to invoke, the second one is an array with the parameter types for the method being invoked and the third one is an array with the actual call arguments. In addition, the caller also needs to set an RPC attachment specifying that the call is a generic call and how to decode the arguments. The possible values are: - true - raw.return - nativejava - bean - protobuf-json An attacker can control this RPC attachment and set it to nativejava to force the java deserialization of the byte array located in the third argument.

The camel-castor component in Apache Camel 2.x before 2.19.4 and 2.20.x before 2.20.1 is vulnerable to Java object de-serialisation vulnerability. De-serializing untrusted data can lead to security flaws.

A bug in Apache Airflow's XCom PATCH endpoint `PATCH /api/v2/xcomEntries/{key}` allowed an authenticated UI/API user with XCom write permission on a Dag to set XCom entries under reserved key names (e.g. `return_value`) that the matching POST endpoint already validated against `FORBIDDEN_XCOM_KEYS`. The endpoint also accepted serialized payload shapes the triggerer's deserializer treats as code; combined, this allowed RCE on the triggerer when the affected task next deferred. Affects deployments where untrusted users have XCom write permission on Dags that defer to the triggerer. This is a fix-bypass of CVE-2026-33858: PR #64148 added the `FORBIDDEN_XCOM_KEYS` validator only on the POST/set path; the PATCH path was not covered. Users who already upgraded for CVE-2026-33858 should additionally upgrade to `apache-airflow` 3.2.2 or later to cover the PATCH-path bypass.

In Apache Storm versions 1.1.0 to 1.2.2, when the user is using the storm-kafka-client or storm-kafka modules, it is possible to cause the Storm UI daemon to deserialize user provided bytes into a Java class.

Apache OFBiz has unsafe deserialization prior to 17.12.07 version An unauthenticated user can perform an RCE attack

A critical unauthenticated remote code execution vulnerability was found all recent versions of Apache Tapestry. The affected versions include 5.4.5, 5.5.0, 5.6.2 and 5.7.0. The vulnerability I have found is a bypass of the fix for CVE-2019-0195. Recap: Before the fix of CVE-2019-0195 it was possible to download arbitrary class files from the classpath by providing a crafted asset file URL. An attacker was able to download the file `AppModule.class` by requesting the URL `http://localhost:8080/assets/something/services/AppModule.class` which contains a HMAC secret key. The fix for that bug was a blacklist filter that checks if the URL ends with `.class`, `.properties` or `.xml`. Bypass: Unfortunately, the blacklist solution can simply be bypassed by appending a `/` at the end of the URL: `http://localhost:8080/assets/something/services/AppModule.class/` The slash is stripped after the blacklist check and the file `AppModule.class` is loaded into the response. This class usually contains the HMAC secret key which is used to sign serialized Java objects. With the knowledge of that key an attacker can sign a Java gadget chain that leads to RCE (e.g. CommonsBeanUtils1 from ysoserial). Solution for this vulnerability: * For Apache Tapestry 5.4.0 to 5.6.1, upgrade to 5.6.2 or later. * For Apache Tapestry 5.7.0, upgrade to 5.7.1 or later.

The fix for CVE-2026-41409 was not applied to the 2.1.X and 2.2.X branches. Here was the original issue description: The fix for CVE-2024-52046 in Apache MINA AbstractIoBuffer.getObject() was incomplete. The classname allowlist of classes allowed to be deserialized was applied too late after a static initializer in a class to be read might already have been executed. Affected versions are Apache MINA 2.1.0 <= 2.1.11, and 2.2.0 <= 2.2.6. The problem is resolved in Apache MINA 2.1.12, and 2.2.7 by applying the classname allowlist earlier. Affected are applications using Apache MINA that call IoBuffer.getObject(). Applications using Apache MINA are advised to upgrade The fix for CVE-2024-52046 in Apache MINA AbstractIoBuffer.getObject() was incomplete. The classname allowlist of classes allowed to be deserialized was applied too late after a static initializer in a class to be read might already have been executed. Affected versions are Apache MINA 2.1.0 <= 2.1.110, and 2.2.0 <= 2.2.6. The problem is resolved in Apache MINA 2.1.12, and 2.2.7 by applying the classname allowlist earlier. Affected are applications using Apache MINA that call IoBuffer.getObject(). Applications using Apache MINA are advised to upgrade

Apache Brooklyn uses the SnakeYAML library for parsing YAML inputs. SnakeYAML allows the use of YAML tags to indicate that SnakeYAML should unmarshal data to a Java type. In the default configuration in Brooklyn before 0.10.0, SnakeYAML will allow unmarshalling to any Java type available on the classpath. This could provide an authenticated user with a means to cause the JVM running Brooklyn to load and run Java code without detection by Brooklyn. Such code would have the privileges of the Java process running Brooklyn, including the ability to open files and network connections, and execute system commands. There is known to be a proof-of-concept exploit using this vulnerability.

Apache OpenMeetings before 3.1.2 is vulnerable to Remote Code Execution via RMI deserialization attack.

Arbitrary Class Instantiation via Model Manifest in Apache OpenNLP ExtensionLoader Versions Affected: before 1.9.5, before 2.5.9, before 3.0.0-M3 Description:  The ExtensionLoader.instantiateExtension(Class, String) method loads a class by its fully-qualified name via Class.forName() and invokes its no-arg constructor, with the class name sourced from the manifest.properties entry of a model archive. The existing isAssignableFrom check correctly rejects classes that are not subtypes of the expected extension interface (BaseToolFactory for factory=, ArtifactSerializer for serializer-class-*), but the check runs after Class.forName() has already loaded and initialized the named class. Class.forName() with default initialization semantics executes the target class's static initializer before returning, so an attacker who can supply a crafted model archive can cause the static initializer of any class on the classpath to run during model loading, regardless of whether that class passes the subsequent type check. Exploitation requires a class with attacker-useful side effects in its static initializer (for example, JNDI lookup, outbound network I/O, or filesystem access) to be present on the classpath, so this is not a drop-in remote code execution; however, the attack surface grows as third-party model distribution becomes more common (community model repositories, Hugging Face-style sharing), where users routinely load model files from origins they do not control. A secondary, narrower vector affects deployments that ship legitimate BaseToolFactory or ArtifactSerializer subclasses with side-effecting no-arg constructors: a malicious manifest can name such a class and force its constructor to run during model load. Mitigation:  * 2.x users should upgrade to 2.5.9. * 3.x users should upgrade to 3.0.0-M3. Note: The fix introduces a package-prefix allowlist that is consulted before Class.forName() is invoked, so the static initializer of a disallowed class is never executed. Classes under the opennlp. prefix remain permitted by default. Deployments that load models referencing factories or serializers outside opennlp.* must opt those packages in, either programmatically via ExtensionLoader.registerAllowedPackage(String) before the first model load, or by setting the OPENNLP_EXT_ALLOWED_PACKAGES system property to a comma-separated list of allowed package prefixes. Users who cannot upgrade immediately should ensure that all model files are sourced from trusted origins and should audit their classpath for classes with side-effecting static initializers or constructors, particularly any that perform JNDI lookups, network requests, or filesystem operations during class initialization.

When an application with unsupported Codehaus versions of Groovy from 1.7.0 to 2.4.3, Apache Groovy 2.4.4 to 2.4.7 on classpath uses standard Java serialization mechanisms, e.g. to communicate between servers or to store local data, it was possible for an attacker to bake a special serialized object that will execute code directly when deserialized. All applications which rely on serialization and do not isolate the code which deserializes objects were subject to this vulnerability.

Deserialization of Untrusted Data vulnerability of Apache ShardingSphere-UI allows an attacker to inject outer link resources. This issue affects Apache ShardingSphere-UI Apache ShardingSphere-UI version 4.1.1 and later versions; Apache ShardingSphere-UI versions prior to 5.0.0.

The DiskFileItem class in Apache Wicket 6.x before 6.25.0 and 1.5.x before 1.5.17 allows remote attackers to cause a denial of service (infinite loop) and write to, move, and delete files with the permissions of DiskFileItem, and if running on a Java VM before 1.3.1, execute arbitrary code via a crafted serialized Java object.

Apache OFBiz has unsafe deserialization prior to 17.12.06. An unauthenticated attacker can use this vulnerability to successfully take over Apache OFBiz.

Apache Tika before 1.14 allows Java code execution for serialized objects embedded in MATLAB files. The issue exists because Tika invokes JMatIO to do native deserialization.

The Apache XML-RPC (aka ws-xmlrpc) library 3.1.3, as used in Apache Archiva, allows remote attackers to execute arbitrary code via a crafted serialized Java object in an <ex:serializable> element.