Apache CloudStack before 4.5.2 does not properly preserve VNC passwords when migrating KVM virtual machines, which allows remote attackers to gain access by connecting to the VNC server.

In Apache HTTP Server 2.4.32-2.4.39, when mod_remoteip was configured to use a trusted intermediary proxy server using the "PROXY" protocol, a specially crafted PROXY header could trigger a stack buffer overflow or NULL pointer deference. This vulnerability could only be triggered by a trusted proxy and not by untrusted HTTP clients.

Impala sessions use a 16 byte secret to verify that the session is not being hijacked by another user. However, these secrets appear in the Impala logs, therefore Impala users with access to the logs can use another authenticated user's sessions with specially constructed requests. This means the attacker is able to execute statements for which they don't have the necessary privileges otherwise. Impala deployments with Apache Sentry or Apache Ranger authorization enabled may be vulnerable to privilege escalation if an authenticated attacker is able to hijack a session or query from another authenticated user with privileges not assigned to the attacker. Impala deployments with audit logging enabled may be vulnerable to incorrect audit logging as a user could undertake actions that were logged under the name of a different authenticated user. Constructing an attack requires a high degree of technical sophistication and access to the Impala system as an authenticated user. Mitigation: If an Impala deployment uses Apache Sentry, Apache Ranger or audit logging, then users should upgrade to a version of Impala with the fix for IMPALA-10600. The Impala 4.0 release includes this fix. This hides session secrets from the logs to eliminate the risk of any attack using this mechanism. In lieu of an upgrade, restricting access to logs that expose secrets will reduce the risk of an attack. Restricting access to the Impala deployment to trusted users will also reduce the risk of an attack. Log redaction techniques can be used to redact secrets from the logs.

In Apache DolphinScheduler before 1.3.6 versions, authorized users can use SQL injection in the data source center. (Only applicable to MySQL data source with internal login account password)

Apache Solr's Kerberos plugin can be configured to use delegation tokens, which allows an application to reuse the authentication of an end-user or another application. There are two issues with this functionality (when using SecurityAwareZkACLProvider type of ACL provider e.g. SaslZkACLProvider). Firstly, access to the security configuration can be leaked to users other than the solr super user. Secondly, malicious users can exploit this leaked configuration for privilege escalation to further expose/modify private data and/or disrupt operations in the Solr cluster. The vulnerability is fixed from Apache Solr 6.6.1 onwards.

In Apache Hadoop 2.8.0, 3.0.0-alpha1, and 3.0.0-alpha2, the LinuxContainerExecutor runs docker commands as root with insufficient input validation. When the docker feature is enabled, authenticated users can run commands as root.

Several REST service endpoints of Apache Archiva are not protected against Cross Site Request Forgery (CSRF) attacks. A malicious site opened in the same browser as the archiva site, may send an HTML response that performs arbitrary actions on archiva services, with the same rights as the active archiva session (e.g. administrator rights).

In Apache Geode before v1.4.0, the Geode server stores application objects in serialized form. Certain cluster operations and API invocations cause these objects to be deserialized. A user with DATA:WRITE access to the cluster may be able to cause remote code execution if certain classes are present on the classpath.

authz.c in the mod_dav_svn module for the Apache HTTP Server, as distributed in Apache Subversion 1.5.x before 1.5.8 and 1.6.x before 1.6.13, when SVNPathAuthz short_circuit is enabled, does not properly handle a named repository as a rule scope, which allows remote authenticated users to bypass intended access restrictions via svn commands.

The optional ShellUserGroupProvider in Apache NiFi 1.10.0 to 1.16.2 and Apache NiFi Registry 0.6.0 to 1.16.2 does not neutralize arguments for group resolution commands, allowing injection of operating system commands on Linux and macOS platforms. The ShellUserGroupProvider is not included in the default configuration. Command injection requires ShellUserGroupProvider to be one of the enabled User Group Providers in the Authorizers configuration. Command injection also requires an authenticated user with elevated privileges. Apache NiFi requires an authenticated user with authorization to modify access policies in order to execute the command. Apache NiFi Registry requires an authenticated user with authorization to read user groups in order to execute the command. The resolution removes command formatting based on user-provided arguments.

Apache Qpid AMQP 0-x JMS client before 6.0.4 and JMS (AMQP 1.0) before 0.10.0 does not restrict the use of classes available on the classpath, which might allow remote authenticated users with permission to send messages to deserialize arbitrary objects and execute arbitrary code by leveraging a crafted serialized object in a JMS ObjectMessage that is handled by the getObject function.

JMSSink in all versions of Log4j 1.x is vulnerable to deserialization of untrusted data when the attacker has write access to the Log4j configuration or if the configuration references an LDAP service the attacker has access to. The attacker can provide a TopicConnectionFactoryBindingName configuration causing JMSSink to perform JNDI requests that result in remote code execution in a similar fashion to CVE-2021-4104. Note this issue only affects Log4j 1.x when specifically configured to use JMSSink, which is not the default. 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.

The getObject method of the javax.jms.ObjectMessage class in the (1) JMS Core client, (2) Artemis broker, and (3) Artemis REST component in Apache ActiveMQ Artemis before 1.4.0 might allow remote authenticated users with permission to send messages to the Artemis broker to deserialize arbitrary objects and execute arbitrary code by leveraging gadget classes being present on the Artemis classpath.

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

Apache Guacamole 1.2.0 and 1.3.0 do not properly validate responses received from a SAML identity provider. If SAML support is enabled, this may allow a malicious user to assume the identity of another Guacamole user.

Apache Superset up to and including 1.3.0 when configured with ENABLE_TEMPLATE_PROCESSING on (disabled by default) allowed SQL injection when a malicious authenticated user sends an http request with a custom URL.

In Apache HTTP Server 2.4 releases 2.4.37 and 2.4.38, a bug in mod_ssl when using per-location client certificate verification with TLSv1.3 allowed a client to bypass configured access control restrictions.

In Apache HTTP Server 2.4 release 2.4.38 and prior, a race condition in mod_auth_digest when running in a threaded server could allow a user with valid credentials to authenticate using another username, bypassing configured access control restrictions.

In Apache CouchDB, a malicious user with permission to create documents in a database is able to attach a HTML attachment to a document. If a CouchDB admin opens that attachment in a browser, e.g. via the CouchDB admin interface Fauxton, any JavaScript code embedded in that HTML attachment will be executed within the security context of that admin. A similar route is available with the already deprecated _show and _list functionality. This privilege escalation vulnerability allows an attacker to add or remove data in any database or make configuration changes. This issue affected Apache CouchDB prior to 3.1.2

In all previously released Apache HBase 2.x versions (2.0.0-2.0.4, 2.1.0-2.1.3), authorization was incorrectly applied to users of the HBase REST server. Requests sent to the HBase REST server were executed with the permissions of the REST server itself, not with the permissions of the end-user. This issue is only relevant when HBase is configured with Kerberos authentication, HBase authorization is enabled, and the REST server is configured with SPNEGO authentication. This issue does not extend beyond the HBase REST server.

JMSAppender in Log4j 1.2 is vulnerable to deserialization of untrusted data when the attacker has write access to the Log4j configuration. The attacker can provide TopicBindingName and TopicConnectionFactoryBindingName configurations causing JMSAppender to perform JNDI requests that result in remote code execution in a similar fashion to CVE-2021-44228. Note this issue only affects Log4j 1.2 when specifically configured to use JMSAppender, which is not the default. 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.

Apache Airflow, in versions prior to 2.7.0, contains a security vulnerability that can be exploited by an authenticated user possessing Connection edit privileges. This vulnerability allows the user to access connection information and exploit the test connection feature by sending many requests, leading to a denial of service (DoS) condition on the server. Furthermore, malicious actors can leverage this vulnerability to establish harmful connections with the server. Users of Apache Airflow are strongly advised to upgrade to version 2.7.0 or newer to mitigate the risk associated with this vulnerability. Additionally, administrators are encouraged to review and adjust user permissions to restrict access to sensitive functionalities, reducing the attack surface.

Apache Ambari before 2.1, as used in IBM Infosphere BigInsights 4.x before 4.1, includes cleartext passwords on a Configs screen, which allows physically proximate attackers to obtain sensitive information by reading password fields.

The svn_repos_trace_node_locations function in Apache Subversion before 1.7.21 and 1.8.x before 1.8.14, when path-based authorization is used, allows remote authenticated users to obtain sensitive path information by reading the history of a node that has been moved from a hidden path.

The JK Connector (aka mod_jk) 1.2.0 through 1.2.26 in Apache Tomcat allows remote attackers to obtain sensitive information via an arbitrary request from an HTTP client, in opportunistic circumstances involving (1) a request from a different client that included a Content-Length header but no POST data or (2) a rapid series of requests, related to noncompliance with the AJP protocol's requirements for requests containing Content-Length headers.

The doRead method in Apache Tomcat 4.1.32 through 4.1.34 and 5.5.10 through 5.5.20 does not return a -1 to indicate when a certain error condition has occurred, which can cause Tomcat to send POST content from one request to a different request.

XWork 2.2.1 in Apache Struts 2.2.1, and OpenSymphony XWork in OpenSymphony WebWork, allows remote attackers to obtain potentially sensitive information about internal Java class paths via vectors involving an s:submit element and a nonexistent method, a different vulnerability than CVE-2011-1772.3.

LibreOffice before 4.4.5 and Apache OpenOffice before 4.1.2 uses the stored LinkUpdateMode configuration information in OpenDocument Format files and templates when handling links, which might allow remote attackers to obtain sensitive information via a crafted document, which embeds data from local files into (1) Calc or (2) Writer.

Apache HTTP Server, when running on Linux with a document root on a Windows share mounted using smbfs, allows remote attackers to obtain unprocessed content such as source files for .php programs via a trailing "\" (backslash), which is not handled by the intended AddType directive.

Apache Subversion SVN authz protected copyfrom paths regression Subversion servers reveal 'copyfrom' paths that should be hidden according to configured path-based authorization (authz) rules. When a node has been copied from a protected location, users with access to the copy can see the 'copyfrom' path of the original. This also reveals the fact that the node was copied. Only the 'copyfrom' path is revealed; not its contents. Both httpd and svnserve servers are vulnerable.

Apache CloudStack before 4.5.2 might allow remote authenticated administrators to obtain sensitive password information for root accounts of virtual machines via unspecified vectors related to API calls.

In Apache Incubator Superset before 0.32, a user can view database names that he has no access to on a dropdown list in SQLLab

The Apache Storm Logviewer daemon exposes HTTP-accessible endpoints to read/search log files on hosts running Storm. In Apache Storm versions 0.9.1-incubating to 1.2.2, it is possible to read files off the host's file system that were not intended to be accessible via these endpoints.

Apache Tomcat 6.0.0 to 6.0.13, 5.5.0 to 5.5.24, 5.0.0 to 5.0.30, 4.1.0 to 4.1.36, and 3.3 to 3.3.2 treats single quotes ("'") as delimiters in cookies, which might cause sensitive information such as session IDs to be leaked and allow remote attackers to conduct session hijacking attacks.

In Apache OFBiz 16.11.01 to 16.11.04, the OFBiz HTTP engine (org.apache.ofbiz.service.engine.HttpEngine.java) handles requests for HTTP services via the /webtools/control/httpService endpoint. Both POST and GET requests to the httpService endpoint may contain three parameters: serviceName, serviceMode, and serviceContext. The exploitation occurs by having DOCTYPEs pointing to external references that trigger a payload that returns secret information from the host.

In Apache Spark 2.1.0 to 2.1.2, 2.2.0 to 2.2.1, and 2.3.0, it's possible for a malicious user to construct a URL pointing to a Spark cluster's UI's job and stage info pages, and if a user can be tricked into accessing the URL, can be used to cause script to execute and expose information from the user's view of the Spark UI. While some browsers like recent versions of Chrome and Safari are able to block this type of attack, current versions of Firefox (and possibly others) do not.

Apache Axis 1.0 allows remote attackers to obtain sensitive information by requesting a non-existent WSDL file, which reveals the installation path in the resulting exception message.

In Eclipse Jetty 9.4.32 to 9.4.38, 10.0.0.beta2 to 10.0.1, and 11.0.0.beta2 to 11.0.1, if a user uses a webapps directory that is a symlink, the contents of the webapps directory is deployed as a static webapp, inadvertently serving the webapps themselves and anything else that might be in that directory.

mod_authz_svn in Apache Subversion 1.7.x before 1.7.21 and 1.8.x before 1.8.14, when using Apache httpd 2.4.x, does not properly restrict anonymous access, which allows remote anonymous users to read hidden files via the path name.

Apache Ambari 2.x before 2.4.0 includes KDC administrator passwords on the kadmin command line, which allows local users to obtain sensitive information via a process listing.

The AJP connector in Apache Tomcat 4.0.1 through 4.0.6 and 4.1.0 through 4.1.36, as used in Hitachi Cosminexus Application Server and standalone, does not properly handle when a connection is broken before request body data is sent in a POST request, which can lead to an information leak when "unsuitable request body data" is used for a different request, possibly related to Java Servlet pages.

When updating a Process Group via the API in NiFi versions 1.3.0 to 1.9.2, the response to the request includes all of its contents (at the top most level, not recursively). The response included details about processors and controller services which the user may not have had read access to.

Apache Tomcat 5.5.x before 5.5.34, 6.x before 6.0.33, and 7.x before 7.0.17, when the MemoryUserDatabase is used, creates log entries containing passwords upon encountering errors in JMX user creation, which allows local users to obtain sensitive information by reading a log file.

Apache Tomcat 6.x before 6.0.45, 7.x before 7.0.68, 8.x before 8.0.31, and 9.x before 9.0.0.M2 does not place org.apache.catalina.manager.StatusManagerServlet on the org/apache/catalina/core/RestrictedServlets.properties list, which allows remote authenticated users to bypass intended SecurityManager restrictions and read arbitrary HTTP requests, and consequently discover session ID values, via a crafted web application.

Apache Tika server (aka tika-server) in Apache Tika 1.9 might allow remote attackers to read arbitrary files via the HTTP fileUrl header.

Apache Directory LDAP API before 1.0.0-M31 allows attackers to conduct timing attacks via unspecified vectors.

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.

An administrator with user search entitlements in Apache Syncope 1.2.x before 1.2.11, 2.0.x before 2.0.8, and unsupported releases 1.0.x and 1.1.x which may be also affected, can recover sensitive security values using the fiql and orderby parameters.

The clustered setup of Apache MXNet allows users to specify which IP address and port the scheduler will listen on via the DMLC_PS_ROOT_URI and DMLC_PS_ROOT_PORT env variables. In versions older than 1.0.0, however, the MXNet framework will listen on 0.0.0.0 rather than user specified DMLC_PS_ROOT_URI once a scheduler node is initialized. This exposes the instance running MXNet to any attackers reachable via the interface they didn't expect to be listening on. For example: If a user wants to run a clustered setup locally, they may specify to run on 127.0.0.1. But since MXNet will listen on 0.0.0.0, it makes the port accessible on all network interfaces.

In Apache Hadoop 3.0.0-alpha1 to 3.0.0, 2.9.0, 2.8.0 to 2.8.3, and 2.5.0 to 2.7.5, HDFS exposes extended attribute key/value pairs during listXAttrs, verifying only path-level search access to the directory rather than path-level read permission to the referent.