A processor changes telemetry inside a pipeline, such as filtering, annotating, or transforming records. A connector bridges pipelines by forwarding, replicating, condensing, or correlating data between them. In practice, processors refine data within one flow, while connectors let teams sequence pipelines and move telemetry across paths to support more advanced observability designs.
How processors and connectors differ in the Collector pipeline model
The simplest way to separate them is by where they act. A processor works inside a pipeline and changes the telemetry already flowing through it. A connector sits between pipelines and moves telemetry from one path to another, which makes it useful when you want one stream to feed multiple downstream uses or when one signal should be turned into another.
That distinction matters because the Collector is not just a pass-through agent. It is a pipeline builder, so the component you choose changes whether you are refining data in place or shaping how data travels across multiple flows. The difference affects how you design routing, reduction, correlation, and reuse of telemetry.
For observability teams, the practical test is whether the component should preserve the pipeline shape or alter it. If the task is to filter noise, enrich records, redact fields, or transform attributes before export, a processor is the right fit. If the task is to replicate traces into metrics, route data into another pipeline, or combine streams for downstream analysis, a connector is doing the heavier architectural work. The Observability, telemetry and SRE for security teams guide shows how these Collector patterns support security use cases without turning the pipeline into a single monolithic flow.
Why the distinction matters in real Collector designs
Processors are usually the tool for local hygiene and policy enforcement. They change the contents of telemetry without changing the destination topology, so they are best when the same stream needs filtering, sampling, redaction, or enrichment before it leaves the pipeline. Connectors are more structural: they let you bridge one pipeline into another, which is useful when a single signal needs to drive multiple outcomes or when one telemetry type should be repurposed for another stage.
That structural role is why connectors are often associated with more advanced designs. They help teams decouple collection, reduction, and distribution, rather than forcing every downstream use case to sit behind the same exporter. In practice, that makes connectors more about pipeline composition, while processors are more about record-level manipulation.
When the data path starts to look like a fan-out, a feedback loop, or a pipeline chain, the connector is usually the relevant component. When the question is how to change the contents of a record before it is exported, the processor is usually the relevant component. The Security Data Pipeline Guide is a useful reference for the broader pattern of normalisation, routing, and reduction that surrounds these Collector choices.
Choosing the right component for telemetry hygiene versus telemetry flow
A good rule is to ask whether the change is about data shape or data path. Data shape changes, such as filtering, masking, and attribute rewriting, belong with processors. Data path changes, such as forwarding, replicating, condensing, or correlating telemetry between pipelines, belong with connectors.
This is why the two are complementary rather than interchangeable. A processor can make telemetry safer, smaller, or more uniform. A connector can make telemetry more reusable, more composable, and more useful across multiple observability stages. The collector design becomes clearer when you keep those responsibilities separate instead of trying to use one component type for every job.
Teams that are building security-sensitive telemetry flows should also be careful not to confuse structural routing with protection. A connector can move data into another pipeline, but that does not automatically make the data safer or cleaner. The JWT and Session Tokens in Code and Logs Guide is a reminder that if sensitive values may appear in telemetry, the processor layer is where redaction and suppression decisions usually belong.
Risk and Threat Considerations
The main risk is misplacing control at the wrong layer. If teams rely on a connector to solve a hygiene problem, sensitive or noisy telemetry can be moved too far before it is filtered. If they rely on a processor to solve a topology problem, they can create brittle pipelines that are hard to reuse, correlate, or scale across observability paths.
Failure mechanism: The Collector design breaks down when a component is used for a job it was not meant to do, especially when record-level changes and pipeline-level routing are blended together. That can leave sensitive data flowing farther than intended or make the telemetry architecture difficult to reason about under load.
Impact: The result is usually poorer observability design, higher operational complexity, and a greater chance that telemetry is either over-retained, over-duplicated, or handled inconsistently across pipelines.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
CIS Controls v8, NIST SP 800-53 Rev 5 and NIST CSF 2.0 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| CIS Controls v8 | CIS-13 — Network Monitoring and Defense | Collector pipelines shape telemetry movement and visibility. |
| Recommendation — Use telemetry routing and filtering to improve detection coverage and reduce noise. | ||
| NIST SP 800-53 Rev 5 | AU-6 — Audit Record Review, Analysis, and Reporting | Telemetry processors and connectors affect how audit data is refined and forwarded. |
| AU-12 — Audit Record Generation | Collector design governs how telemetry is produced, transformed, and routed for analysis. | |
| SI-4 — System Monitoring | The distinction informs how monitoring telemetry is enriched, reduced, and distributed. | |
| Recommendation — Review and normalize telemetry so audit data remains usable across pipelines. Configure collection paths so required audit records are captured and delivered reliably. Route monitoring telemetry through the right pipeline stage for analysis and alerting. | ||
| NIST CSF 2.0 | DE.CM-01 — Monitoring for Anomalies and Events | Telemetry pipeline design supports continuous monitoring and event visibility. |
| Recommendation — Ensure telemetry flows support continuous anomaly and event monitoring. | ||
Practitioner Guidance
What to verify: Check whether the requirement is to change telemetry contents or to change telemetry flow. If the requirement mentions filtering, enrichment, masking, or transformation, start with a processor. If it mentions fan-out, bridging, correlation, or reuse across pipelines, start with a connector.
Common mistake: Do not treat connectors as a generic replacement for processors. The collector becomes harder to operate when routing logic, reduction logic, and content-transformation logic are mixed together without a clear boundary.
What good looks like: A clean design has processors handling record hygiene close to the source and connectors handling cross-pipeline movement where architectural reuse is needed. That separation makes the pipeline easier to test, easier to document, and easier to troubleshoot when telemetry volume or topology changes.
Practitioner takeaway: If the question is “what should this telemetry become?”, think processor. If the question is “where should this telemetry go next?”, think connector.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Free weekly newsletter
Subscribe to the NHI & AI Identity Journal
The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.
Bonus 33% off our NHI Course when you subscribe.
Reviewed and updated by the NHIMG editorial team on November 2, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org