Improper Limitation of a Pathname to a Restricted Directory in Logstash Leading to Arbitrary File Write

Improper Limitation of a Pathname to a Restricted Directory (CWE-22) in Logstash can lead to arbitrary file write and potentially remote code execution via Relative Path Traversal (CAPEC-139). The archive extraction utilities used by Logstash do not properly validate file paths within compressed archives. An attacker who can serve a specially crafted archive to Logstash through a compromised or attacker-controlled update endpoint can write arbitrary files to the host filesystem with the privileges of the Logstash process. In certain configurations where automatic pipeline reloading is enabled, this can be escalated to remote code execution.

Logstash Improper Certificate Validation in TCP output

Improper certificate validation in Logstash's TCP output could lead to a man-in-the-middle (MitM) attack in “client” mode, as hostname verification in TCP output was not being performed when the ssl_verification_mode => full was set.

Logstash 1.5.x before 1.5.3 and 1.4.x before 1.4.4 allows remote attackers to read communications between Logstash Forwarder agent and Logstash server.

Logstash 1.4.x before 1.4.5 and 1.5.x before 1.5.4 with Lumberjack output or the Logstash forwarder does not validate SSL/TLS certificates from the Logstash server, which might allow attackers to obtain sensitive information via a man-in-the-middle attack.

Logstash prior to version 2.3.4, Elasticsearch Output plugin would log to file HTTP authorization headers which could contain sensitive information.

Logstash prior to version 2.1.2, the CSV output can be attacked via engineered input that will create malicious formulas in the CSV data.

Logstash versions prior to 2.3.3, when using the Netflow Codec plugin, a remote attacker crafting malicious Netflow v5, Netflow v9 or IPFIX packets could perform a denial of service attack on the Logstash instance. The errors resulting from these crafted inputs are not handled by the codec and can cause the Logstash process to exit.

Elasticsearch Logstash 1.0.14 through 1.4.x before 1.4.2 allows remote attackers to execute arbitrary commands via a crafted event in (1) zabbix.rb or (2) nagios_nsca.rb in outputs/.

Directory traversal vulnerability in the file output plugin in Elasticsearch Logstash before 1.4.3 allows remote attackers to write to arbitrary files via vectors related to dynamic field references in the path option.

Logstash Insertion of Sensitive Information into Log File

An issue was identified by Elastic whereby sensitive information is recorded in Logstash logs under specific circumstances. The prerequisites for the manifestation of this issue are: * Logstash is configured to log in JSON format https://www.elastic.co/guide/en/logstash/current/running-logstash-command-line.html , which is not the default logging format. * Sensitive data is stored in the Logstash keystore and referenced as a variable in Logstash configuration.

In Logstash versions after 6.4.0 and before 6.8.15 and 7.12.0 a TLS certificate validation flaw was found in the monitoring feature. When specifying a trusted server CA certificate Logstash would not properly verify the certificate returned by the monitoring server. This could result in a man in the middle style attack against the Logstash monitoring data.

Logstash versions before 7.4.1 and 6.8.4 contain a denial of service flaw in the Logstash Beats input plugin. An unauthenticated user who is able to connect to the port the Logstash beats input could send a specially crafted network packet that would cause Logstash to stop responding.

A sensitive data disclosure flaw was found in the way Logstash versions before 5.6.15 and 6.6.1 logs malformed URLs. If a malformed URL is specified as part of the Logstash configuration, the credentials for the URL could be inadvertently logged as part of the error message.

X-Pack Machine Learning versions before 6.2.4 and 5.6.9 had a cross-site scripting (XSS) vulnerability. If an attacker is able to inject data into an index that has a ML job running against it, then when another user views the results of the ML job it could allow the attacker to obtain sensitive information from or perform destructive actions on behalf of that other ML user.

X-Pack Machine Learning versions before 6.2.4 and 5.6.9 had a cross-site scripting (XSS) vulnerability. Users with manage_ml permissions could create jobs containing malicious data as part of their configuration that could allow the attacker to obtain sensitive information from or perform destructive actions on behalf of other ML users viewing the results of the jobs.

When logging warnings regarding deprecated settings, Logstash before 5.6.6 and 6.x before 6.1.2 could inadvertently log sensitive information.