opentelemetry-js is the OpenTelemetry JavaScript Client. Prior to 0.217.0, a single malformed HTTP request crashes any Node.js process running the OpenTelemetry JS Prometheus exporter. The metrics endpoint (default 0.0.0.0:9464) has no error handling around URL parsing, so a request with an invalid URI causes an uncaught TypeError that terminates the process. This vulnerability is fixed in 0.217.0.

Prometheus is an open-source monitoring system and time series database. From 2.49.0 to before 3.5.3 and 3.11.3, in the Prometheus server's legacy web UI (enabled via the command-line flag --enable-feature=old-ui), the histogram heatmap chart view does not escape le label values when inserting them into the HTML for use as axis tick mark labels. An attacker who can inject crafted metrics can execute JavaScript in the browser of any Prometheus user who views the metric in the heatmap chart UI. This vulnerability is fixed in 3.5.3 and 3.11.3.

prometheus

Prometheus is an open-source monitoring system and time series database. Prior to versions 3.5.3 and 3.11.3, the remote read endpoint (/api/v1/read) does not validate the declared decoded length in a snappy-compressed request body before allocating memory. An unauthenticated attacker can send a small payload that causes a huge heap allocation per request. Under concurrent load this can exhaust available memory and crash the Prometheus process. This issue has been patched in versions 3.5.3 and 3.11.3.

prometheus

Prometheus is an open-source monitoring system and time series database. Prior to versions 3.5.3 and 3.11.3, the client_secret field in the Azure AD remote write OAuth configuration (storage/remote/azuread) was typed as string instead of Secret. Prometheus redacts fields of type Secret when serving the configuration via the /-/config HTTP API endpoint. Because the field was a plain string, the Azure OAuth client secret was exposed in plaintext to any user or process with access to that endpoint. This issue has been patched in versions 3.5.3 and 3.11.3.

prometheus

Improper Certificate Validation via Global SSL Context Downgrade in Apache Storm Prometheus Reporter Versions Affected: from 2.6.3 to 2.8.6 Description:  In production deployments where an administrator enables storm.daemon.metrics.reporter.plugin.prometheus.skip_tls_validation (by default it is disabled) intending to affect only the Prometheus reporter, the undocumented global side effect creates an attack surface across every TLS-protected communication channel in the Storm daemon. The PrometheusPreparableReporter class implements an INSECURE_TRUST_MANAGER that accepts all SSL certificates without validation, with empty checkClientTrusted and checkServerTrusted methods. Most critically, when the storm.daemon.metrics.reporter.plugin.prometheus.skip_tls_validation configuration option is enabled (default = disabled) for HTTPS Prometheus PushGateway connections, the INSECURE_CONNECTION_FACTORY calls SSLContext.setDefault(sslContext), which globally replaces the JVM's default SSL context rather than applying the insecure context only to the Prometheus connection. This payload flows through storm.yaml configuration → PrometheusPreparableReporter.prepare() → INSECURE_CONNECTION_FACTORY → SSLContext.setDefault(), resulting in a JVM-wide TLS security downgrade. All subsequent HTTPS connections in the process - including ZooKeeper, Thrift, Netty, and UI connections - silently trust all certificates, including self-signed, expired, and attacker-generated ones, enabling man-in-the-middle interception of cluster state, topology submissions, tuple data, and administrative credentials. Mitigation: 2.x users should upgrade to 2.8.7 if the Prometheus Metrics Reporter is used. Prometheus Metrics Reporter Users who cannot upgrade immediately should remove the storm.daemon.metrics.reporter.plugin.prometheus.skip_tls_validation: true setting from their storm.yaml configuration and instead configure a proper truststore containing the PushGateway's certificate.

apache

Prometheus is an open-source monitoring system and time series database. Versions 3.0 through 3.5.1 and 3.6.0 through 3.11.1 have stored cross-site scripting vulnerabilities in multiple components of the Prometheus web UI where metric names and label values are injected into innerHTML without escaping. In both the Mantine UI and old React UI, chart tooltips on the Graph page render metric names containing HTML/JavaScript without sanitization. In the old React UI, the Metric Explorer fuzzy search results use dangerouslySetInnerHTML without escaping, and heatmap cell tooltips interpolate le label values without sanitization. With Prometheus v3.x defaulting to UTF-8 metric and label name validation, characters like <, >, and " are now valid in metric names and labels. An attacker who can inject metrics via a compromised scrape target, remote write, or OTLP receiver endpoint can execute arbitrary JavaScript in the browser of any Prometheus user who views the metric in the Graph UI, potentially enabling configuration exfiltration, data deletion, or Prometheus shutdown depending on enabled flags. This issue has been fixed in versions 3.5.2 and 3.11.2. If developers are unable to immediately update, the following workarounds are recommended: ensure that the remote write receiver (--web.enable-remote-write-receiver) and the OTLP receiver (--web.enable-otlp-receiver) are not exposed to untrusted sources; verify that all scrape targets are trusted and not under attacker control; avoid enabling admin or mutating API endpoints (e.g., --web.enable-admin-api or --web.enable-lifecycle) in environments where untrusted data may be ingested; and refrain from clicking untrusted links, particularly those containing functions such as label_replace, as they may generate poisoned label names and values.

prometheus

An Exposure of Sensitive Information to an Unauthorized Actor vulnerability exists in Apache DolphinScheduler. This vulnerability may allow unauthorized actors to access sensitive information, including database credentials. This issue affects Apache DolphinScheduler versions 3.1.*. Users are recommended to upgrade to: * version ≥ 3.2.0 if using 3.1.x As a temporary workaround, users who cannot upgrade immediately may restrict the exposed management endpoints by setting the following environment variable: ``` MANAGEMENT_ENDPOINTS_WEB_EXPOSURE_INCLUDE=health,metrics,prometheus ``` Alternatively, add the following configuration to the application.yaml file: ``` management:    endpoints:      web:         exposure:           include: health,metrics,prometheus ``` This issue has been reported as CVE-2023-48796: https://cveprocess.apache.org/cve5/CVE-2023-48796

apache

Memory Allocation with Excessive Size Value (CWE-789) in the Prometheus remote_write HTTP handler in Metricbeat can lead Denial of Service via Excessive Allocation (CAPEC-130).

openITCOCKPIT is an open source monitoring tool built for different monitoring engines like Nagios, Naemon and Prometheus. openITCOCKPIT Community Edition 5.3.1 and earlier contains an unsafe PHP deserialization pattern in the processing of changelog entries. Serialized changelog data derived from attacker-influenced application state is unserialized without restricting allowed classes. Although no current application endpoint was found to introduce PHP objects into this data path, the presence of an unrestricted unserialize() call constitutes a latent PHP object injection vulnerability. If future code changes, plugins, or refactors introduce object values into this path, the vulnerability could become immediately exploitable with severe impact, including potential remote code execution.

it-novum

openITCOCKPIT is an open source monitoring tool built for different monitoring engines like Nagios, Naemon and Prometheus. Versions 5.3.1 and below contain an unsafe deserialization sink in the Gearman worker implementation. The worker function registered as oitc_gearman calls PHP's unserialize() on job payloads without enforcing class restrictions or validating data origin. While the intended deployment assumes only trusted internal components enqueue Gearman jobs, this trust boundary is not enforced in application code. In environments where the Gearman service or worker is exposed to untrusted systems, an attacker may submit crafted serialized payloads to trigger PHP Object Injection in the worker process. This vulnerability is exploitable when Gearman listens on non-local interfaces, network access to TCP/4730 is unrestricted, or untrusted systems can enqueue jobs. Default, correctly hardened deployments may not be immediately exploitable, but the unsafe sink remains present in code regardless of deployment configuration. Enforcing this trust boundary in code would significantly reduce risk and prevent exploitation in misconfigured environments. This issue has been fixed in version 5.4.0.

it-novum

Scraparr is a Prometheus Exporter for various components of the *arr Suite. From 3.0.0-beta to before 3.0.2, when the Readarr integration was enabled, the exporter exposed the configured Readarr API key as the alias metric label value. Users were affected only if all of the following conditions are met, Readarr scraping feature was enabled and no alias configured, the exporter’s /metrics endpoint was accessible to external or unauthorized users, and the Readarr instance is externally accessible. If the /metrics endpoint was publicly accessible, the Readarr API key could have been disclosed via exported metrics data. This vulnerability is fixed in 3.0.2.

thecfu

Improper Validation of Array Index (CWE-129) exists in Metricbeat can allow an attacker to cause a Denial of Service through Input Data Manipulation (CAPEC-153) via specially crafted, malformed payloads sent to the Graphite server metricset or Zookeeper server metricset. Additionally, Improper Input Validation (CWE-20) exists in the Prometheus helper module that can allow an attacker to cause a Denial of Service through Input Data Manipulation (CAPEC-153) via specially crafted, malformed metric data.

elastic

This vulnerability in Grafana's datasource proxy API allows authorization checks to be bypassed by adding an extra slash character in the URL path. Users with minimal permissions could gain unauthorized read access to GET endpoints in Alertmanager and Prometheus datasources. The issue primarily affects datasources that implement route-specific permissions, including Alertmanager and certain Prometheus-based datasources.

An issue was discovered in Fluent Bit 3.1.9. When the Prometheus Remote Write input plugin is running and listening on an IP address and port, one can send a packet with Content-Length: 0 and it crashes the server. Improper handling of the case when Content-Length is 0 allows a user (with access to the endpoint) to perform a remote Denial of service attack. The crash happens because of a NULL pointer dereference when 0 (from the Content-Length) is passed to the function cfl_sds_len, which in turn tries to cast a NULL pointer into struct cfl_sds. This is related to process_payload_metrics_ng() at prom_rw_prot.c.

treasuredata

A flaw was found in openshift-gitops-operator-container. The openshift.io/cluster-monitoring label is applied to all namespaces that deploy an ArgoCD CR instance, allowing the namespace to create a rogue PrometheusRule. This issue can have adverse effects on the platform monitoring stack, as the rule is rolled out cluster-wide when the label is applied.

Envoy Gateway is an open source project for managing Envoy Proxy as a standalone or Kubernetes-based application gateway. A user with access to the Kubernetes cluster can use a path traversal attack to execute Envoy Admin interface commands on proxies managed by any version of Envoy Gateway prior to 1.2.6. The admin interface can be used to terminate the Envoy process and extract the Envoy configuration (possibly containing confidential data). Version 1.2.6 fixes the issue. As a workaround, the `EnvoyProxy` API can be used to apply a bootstrap config patch that restricts access strictly to the prometheus stats endpoint. Find below an example of such a bootstrap patch.

envoyproxy

Exposure of Sensitive Information to an Unauthorized Actor vulnerability was discovered in Open Design Alliance CDE inWEB SDK before 2025.3. Installing CDE Server with default settings allows unauthorized users to visit prometheus metrics page. This can allow attackers to understand more things about the target application which may help in further investigation and exploitation.

authentik is an open-source identity provider. Due to the usage of a non-constant time comparison for the /-/metrics/ endpoint it was possible to brute-force the SECRET_KEY, which is used to authenticate the endpoint. The /-/metrics/ endpoint returns Prometheus metrics and is not intended to be accessed directly, as the Go proxy running in the authentik server container fetches data from this endpoint and serves it on a separate port (9300 by default), which can be scraped by Prometheus without being exposed publicly. authentik 2024.8.5 and 2024.10.3 fix this issue. Since the /-/metrics/ endpoint is not intended to be accessed publicly, requests to the endpoint can be blocked by the reverse proxy/load balancer used in conjunction with authentik.

goauthentik

RabbitMQ is a feature rich, multi-protocol messaging and streaming broker. In affected versions queue deletion via the HTTP API was not verifying the `configure` permission of the user. Users who had all of the following: 1. Valid credentials, 2. Some permissions for the target virtual host & 3. HTTP API access. could delete queues it had no (deletion) permissions for. This issue has been addressed in version 3.12.11 of the open source rabbitMQ release and in versions 1.5.2, 3.13.0, and 4.0.0 of the tanzu release. Users are advised to upgrade. Users unable to upgrade may disable management plugin and use, for example, Prometheus and Grafana for monitoring.

The O-RAN E2T I-Release Prometheus metric Increment function can crash in sctpThread.cpp for message.peerInfo->sctpParams->e2tCounters[IN_SUCC][MSG_COUNTER][ProcedureCode_id_RICsubscription]->Increment().

The O-RAN E2T I-Release Prometheus metric Increment function can crash in sctpThread.cpp for message.peerInfo->counters[IN_INITI][MSG_COUNTER][ProcedureCode_id_E2setup]->Increment().

The O-RAN E2T I-Release buildPrometheusList function can have a NULL pointer dereference because peerInfo can be NULL.