Certifications guide

OTCA: The OpenTelemetry Certified Associate Exam, Domain by Domain

OTCA is a 90-minute, multiple-choice exam on OpenTelemetry — the API and SDK, the Collector, and the observability ideas underneath. Its weighting is the most lopsided of the CNCF associate exams, and that is the whole study plan.

read
7 min read
read
last updated
August 2026
last updated
Overview

What OTCA actually certifies

The OpenTelemetry Certified Associate is the CNCF and Linux Foundation credential for engineers who instrument systems with OpenTelemetry. It is an online, proctored, multiple-choice exam of 90 minutes, priced at $250 for the exam alone, with no prerequisites, one free retake, a 12-month eligibility window and a certification valid for two years. As with PCA, the Linux Foundation publishes neither a question count nor a pass mark.

Four domains, and they are nothing like evenly weighted: The OpenTelemetry API and SDK 46 %, The OpenTelemetry Collector 26 %, Fundamentals of Observability 18 %, Maintaining and Debugging Observability Pipelines 10 %. Nearly half the exam sits in one domain, and almost three quarters of it is API, SDK and Collector combined — this is a hands-on-the-library exam wearing a multiple-choice format.

Gnoseed is not affiliated with CNCF or the Linux Foundation and neither sells nor proctors this exam; it offers free spaced-repetition practice on the published curriculum. Verify the figures above against the official exam page before you book.

Breakdown

The exam at a glance

One exam, four domains, and a distribution that should decide your revision order before you read a single page.

OTCA

OpenTelemetry Certified Associate

AssociateMultiple choice · online proctored · 90 min · $250Prereq: None

<strong>The API and SDK (46 %)</strong> covers the data model, the three signals, SDK pipelines, configuration, context propagation, composability and agents — everything you touch when instrumenting an application. <strong>The Collector (26 %)</strong> is configuration, deployment, scaling, pipelines and transforming data. <strong>Fundamentals of Observability (18 %)</strong> is telemetry data, semantic conventions, instrumentation and analysis. <strong>Maintaining and Debugging Pipelines (10 %)</strong> is context propagation again, plus debugging, error handling and schema management. Note what is absent: no vendor backend, no dashboard product, and no Prometheus.

Choose this if…

  • You are rolling OpenTelemetry out and want the gaps in your model found
  • You want a vendor-neutral instrumentation credential, not a backend one
  • You already know tracing concepts and need the OTel specifics on top
Where the marks are

Nearly half the exam is one domain

Rank the four domains and the picture is stark: API and SDK is 46 %, more than the bottom two combined and nearly double the Collector. If your revision is evenly spread across four domains you are spending a quarter of your time on a domain worth a tenth of the paper.

Inside that domain the competencies are the ones you meet writing instrumentation: the data model, the three signals, SDK pipelines, configuration, context propagation, composability and extension, and agents. Two of them reward precision. The metric instruments are a genuine multiple-choice trap — a Counter is monotonic and a queue length that falls needs an UpDownCounter, while a Gauge reports the value when it is read. And logs are not instrumented the way traces and metrics are: the Logs Bridge API exists for logging-library authors to write appenders, so as an application developer you bridge the library you already use rather than calling it.

The Collector at 26 % is the other half of the practical work: agent versus gateway deployment, the memory limiter that has to sit first in the pipeline to apply backpressure, the exporter queue that decides whether a backend outage costs you data, and the debug exporter you reach for when telemetry arrives but never lands.

What this deck does and does not cover

Two of the five modules are not on the blueprint

The Observability track was written as a general observability course, not as exam prep, so it is worth being explicit about the overlap. Three modules map onto OTCA directly: OpenTelemetry (the API, SDK, Collector and schema material), Distributed Tracing (the data model and context propagation that the 46 % domain is largely made of), and Three Pillars & Signals (the Fundamentals domain).

The other two — Logs & Loki and Grafana — are not on the OTCA blueprint at all. Loki and Grafana are Grafana-stack products, and OTCA is deliberately vendor-neutral: it tests how you produce and route telemetry, not where you store or draw it. Study them because they are how observability actually gets done, not because the exam asks.

The Prometheus track is linked here for one narrow reason: OTCA's Fundamentals domain covers telemetry data and analysis, and metrics are where most engineers' intuitions are weakest. If counters, gauges and histograms are already second nature to you, skip it.

OTCA or PCA

They look adjacent and test opposite things

Both are 90-minute, $250, multiple-choice CNCF associate exams about observability, and they are frequently confused. The split is clean: PCA is about a backend — scraping, storing and querying metrics with PromQL, which is 28 % of that paper — while OTCA is about production, the libraries and pipeline that emit telemetry before any backend sees it. Neither is a prerequisite for the other and there is no ordering between them.

Pick by what you do. If you write the code that emits spans and metrics, or you run the Collector fleet, OTCA describes your job. If you run the Prometheus that scrapes those services and you live in PromQL and Alertmanager, PCA does. The one genuine overlap is conceptual: both exams devote a domain to observability fundamentals, so preparation there counts twice.

FAQ

Common questions

What are the OTCA exam domains and their weights? +

Four domains: The OpenTelemetry API and SDK 46 %, The OpenTelemetry Collector 26 %, Fundamentals of Observability 18 %, and Maintaining and Debugging Observability Pipelines 10 %. CNCF open-sources the full curriculum, so the competencies under each are public.

Is the OTCA exam hands-on? +

No. It is an online, proctored, multiple-choice exam of 90 minutes with no terminal — unlike the CKA. The content is practical even though the format is not, because three quarters of it is about the API, SDK and Collector you would otherwise be configuring.

Are there prerequisites for OTCA, and how much does it cost? +

There are no prerequisites. The exam alone is $250, includes one free retake and a 12-month eligibility window, and the certification is valid for two years. Bundles with a Linux Foundation subscription are sold at a higher price point.

Should I take OTCA or PCA first? +

Neither is a prerequisite for the other, so pick by your work rather than by order. OTCA is about producing telemetry — instrumentation libraries, SDK configuration and the Collector. PCA is about a backend that stores and queries metrics, with PromQL at 28 % of the paper.

Do I need to know Grafana or Loki for OTCA? +

No. OTCA is vendor-neutral and its blueprint names no dashboard or storage product. Both are useful in practice and appear in this track, but nothing on the exam depends on them.

How many questions are on the OTCA exam? +

The Linux Foundation does not publish a question count or a passing score for OTCA, the same as for PCA. Plan around the domain weights instead — they are published, and they are unusually lopsided.

Start where the marks are

Plant your first seed today. Nearly half of OTCA is the API and SDK — ten minutes a day on the data model, signals and context propagation is the highest-value revision you can do.

Start learning free