安全¶

OpenTelemetry Collector 默认以安全的方式运行，但它是配置驱动的。本文档介绍了收集器的重要安全方面和注意事项。本文档适用于最终用户和组件开发人员。它假定至少对 Collector 体系结构和功能有基本的了解。

Note: Please review the configuration documentation prior to this security document.

TL;DR¶

终端用户¶

Configuration
SHOULD only enable the minimum required components
SHOULD ensure sensitive configuration information is stored securely
Permissions
SHOULD not run Collector as root/admin user
MAY require privileged access for some components
Receivers/Exporters
SHOULD use encryption and authentication
SHOULD limit exposure of servers to authorized users
MAY pose a security risk if configuration parameters are modified improperly
Processors
SHOULD configure obfuscation/scrubbing of sensitive metadata
SHOULD configure recommended processors
Extensions
SHOULD NOT expose sensitive health or telemetry data

For more information about securing the OpenTelemetry Collector, see this blog post.

组件开发人员¶

Configuration
MUST come from the central configuration file
SHOULD use configuration helpers
Permissions
SHOULD minimize privileged access
MUST document what required privileged access and why
Receivers/Exporters
MUST default to encrypted connections
SHOULD leverage helper functions
Extensions
SHOULD NOT expose sensitive health or telemetry data by default

配置¶

The Collector binary does not contain an embedded or default configuration and MUST NOT start without a configuration file being specified. The configuration file passed to the Collector MUST be validated prior to be loaded. If an invalid configuration is detected, the Collector MUST fail to start as a protective mechanism.

Note: Issue #886 proposes adding a default configuration to the binary.

The configuration drives the Collector's behavior and care should be taken to ensure the configuration only enables the minimum set of capabilities and as such exposes the minimum set of required ports. In addition, any incoming or outgoing communication SHOULD leverage TLS and authentication.

The Collector keeps the configuration in memory, but where the configuration is loaded from at start time depends on the packaging used. For example, in Kubernetes secrets and configmaps CAN be leveraged. In comparison, the Docker image embeds the configuration in the container where is it not stored in an encrypted manner by default.

The configuration MAY contain sensitive information including:

Authentication information such as API tokens
TLS certificates including private keys

Sensitive information SHOULD be stored securely such as on an encrypted filesystem or secret store. Environment variables CAN be used to handle sensitive and non-sensitive data as the Collector MUST support environment variable expansion.

For more information on environment variable expansion, see this documentation.

When defining Go structs for configuration data that may contain sensitive information, use the configopaque package to define fields with the configopaque.String type. This ensures that the data is masked when serialized to prevent accidental exposure.

For more information, see the configopaque documentation.

Component developers MUST get configuration information from the Collector's configuration file. Component developers SHOULD leverage configuration helper functions.

More information about configuration is provided in the following sections.

权限¶

The Collector supports running as a custom user and SHOULD NOT be run as a root/admin user. For the majority of use-cases, the Collector SHOULD NOT require privileged access to function. Some components MAY require privileged access and care should be taken before enabling these components. Collector components MAY require external permissions including network access or RBAC.

Component developers SHOULD minimize privileged access requirements and MUST document what requires privileged access and why.

More information about permissions is provided in the following sections.

接收器和出口器¶

Receivers and Exporters can be either push or pull-based. In either case, the connection established SHOULD be over a secure and authenticated channel. Unused receivers and exporters SHOULD be disabled to minimize the attack vector of the Collector.

Receivers and Exporters MAY expose buffer, queue, payload, and/or worker settings via configuration parameters. If these settings are available, end-users should proceed with caution before modifying the default values. Improperly setting these values may expose the Collector to additional attack vectors including resource exhaustion.

It is possible that a receiver MAY require the Collector run in a privileged mode in order to operate, which could be a security concern, but today this is not the case.

Component developers MUST default to encrypted connections (via the insecure: false configuration setting) and SHOULD leverage gRPC and http helper functions.

防止拒绝服务攻击¶

Users SHOULD bind receivers' servers to addresses that limit connections to authorized users. For example, if the OTLP receiver OTLP/gRPC server only has local clients, the endpoint setting SHOULD be bound to localhost:

receivers:
  otlp:
    protocols:
      grpc:
        endpoint: localhost:4317

Generally, localhost-like addresses should be preferred over the 0.0.0.0 address. For more information, see CWE-1327.

处理器¶

Processors sit between receivers and exporters. They are responsible for processing the data in some way. From a security perspective, they are useful in a couple ways.

清除敏感数据¶

It is common for a Collector to be used to scrub sensitive data before exporting it to a backend. This is especially important when sending the data to a third-party backend. The Collector SHOULD be configured to obfuscate or scrub sensitive data before exporting.

Note: Issue #2466 proposes adding default obfuscation or scrubbing of known sensitive metadata.

保护资源利用¶

In addition, processors offer safeguards around resource utilization. The batch and especially memory_limiter processor help ensure that the Collector is resource efficient and does not out of memory when overloaded. At least these two processors SHOULD be enabled on every defined pipeline.

For more information on recommended processors and order, see this documentation.

扩展¶

While receivers, processors, and exporters handle telemetry data directly, extensions typical serve different needs.

健康和遥测¶

The initial extensions provided health check information, Collector metrics and traces, and the ability to generate and collect profiling data. When enabled with their default settings, all of these extensions except the health check extension are only accessibly locally to the Collector. Care should be taken when configuring these extensions for remote access as sensitive information may be exposed as a result.

Component developers SHOULD NOT expose health or telemetry data outside the Collector by default.

转发¶

A forwarding extension is typically used when some telemetry data not natively supported by the Collector needs to be collected. For example, the http_forwarder extension can receive and forward HTTP payloads. Forwarding extensions are similar to receivers and exporters so the same security considerations apply.

观察者¶

An observer is capable of performing service discovery of endpoints. Other components of the collector such as receivers MAY subscribe to these extensions to be notified of endpoints coming or going. Observers MAY require certain permissions in order to perform service discovery. For example, the k8s_observer requires certain RBAC permissions in Kubernetes, while the host_observer requires the Collector to run in privileged mode.

子流程¶

Extensions may also be used to run subprocesses. This can be useful when collection mechanisms that cannot natively be run by the Collector (e.g. FluentBit). Subprocesses expose a completely separate attack vector that would depend on the subprocess itself. In general, care should be taken before running any subprocesses alongside the Collector.