5 exercises — each question asks whether a given expression is a natural English collocation used by native speakers.
How to approach True or False collocations
Ask yourself: would a native speaker say this naturally, or does it sound translated?
Check the verb: many errors come from using a generic verb (do/make) instead of the precise one
Some expressions sound almost right — the difference is often one word
0 / 5 completed
1 / 5
True or False: 'trigger a pipeline' is a natural DevOps collocation — used when an event causes a CI/CD pipeline to start.
TRUE — "trigger a pipeline" is core CI/CD vocabulary.
"Trigger" implies an automatic cause-and-effect chain — one event causes something else to happen. This is exactly how CI/CD pipelines work: an event (push, PR, schedule) triggers the pipeline to run.
trigger a pipeline ✅ — "A push to main triggers the deployment pipeline."
trigger a build ✅ — "Opening a PR triggers a build in CI."
trigger a deployment ✅ — "Merging to main triggers a production deployment."
trigger an alert ✅ — "High error rates trigger an alert in PagerDuty."
trigger a webhook ✅ — "The webhook triggers on every push event."
"Start a pipeline" ✅ is also natural but implies manual initiation. "Trigger" specifically implies automation — an event causes the pipeline to start without human intervention.
2 / 5
True or False: 'do a rollback' is the natural English phrase when reverting a deployment to a previous state.
FALSE — "do a rollback" sounds non-native in professional DevOps English.
The correct forms depend on whether you use the noun or the verb:
Verb form (preferred in conversation):
roll back ✅ — "We need to roll back immediately — the error rate spiked."
"Roll back to the previous version." ✅
Noun form with a precise action verb:
trigger a rollback ✅ — automated: "The health check failed and triggered a rollback."
execute a rollback ✅ — technical: "Execute the rollback procedure from the runbook."
perform a rollback ✅ — formal: "We performed a rollback at 02:00 UTC."
initiate a rollback ✅ — start the process
"The rollback was successful." ✅ — noun in a result statement.
"Do a rollback" ❌ and "make a rollback" ❌ both mark non-native usage. In post-incident reports and SRE runbooks, precise language matters. Using "do" or "make" instead of the correct verb is a signal that the writer is translating from another language.
3 / 5
True or False: 'monitor the deployment' is a natural English collocation in DevOps and site reliability engineering.
TRUE — "monitor" is a precise and essential verb in DevOps and SRE vocabulary.
"Monitor" means to continuously track metrics, logs, and health indicators — not just passively watch, but actively measure against thresholds.
monitor a deployment ✅ — "Monitor the deployment for error rate spikes and latency increases."
monitor the system ✅ — ongoing operational practice
monitor error rates ✅ — a key deployment health signal
monitor uptime ✅ — availability tracking
The monitoring stack:
Prometheus — metrics collection
Grafana — dashboards and visualisation
Datadog / New Relic — full-stack monitoring platforms
alerting — notifications when metrics exceed thresholds
In SRE vocabulary, observability = logs + metrics + traces. "Monitor the deployment" means watching for errors, latency spikes, or unexpected behaviour during and after a release. "Watch the deployment" ❌ sounds casual and non-technical.
4 / 5
True or False: 'make a container' is the natural way to describe creating a Docker container in English.
FALSE — "make a container" is not natural Docker/container vocabulary.
Docker has a well-defined lifecycle with specific verbs for each stage. Understanding these collocations signals real Docker experience:
Image operations:
build an image ✅ — docker build: "Build the image from the Dockerfile."
pull an image ✅ — docker pull: download from a registry
push an image ✅ — docker push: upload to a registry
Container operations:
run a container ✅ — docker run: "Run the container in detached mode."
create a container ✅ — docker create: create without starting
spin up a container ✅ — informal: "Spin up a Redis container for local development."
The key distinction: you build an image, then run a container from that image. "Make a container" ❌ does not appear in Docker documentation, Kubernetes docs, or any real DevOps context. There is no docker make command.
5 / 5
True or False: 'scale out' is a real technical collocation — and it means something different from 'scale up'.
TRUE — "scale out" and "scale up" are both real collocations with distinct meanings.
These two terms describe fundamentally different scaling strategies:
scale out ✅ = horizontal scaling — add more instances/nodes: "Scale out from 3 to 10 web server instances."
scale up ✅ = vertical scaling — add more CPU/RAM to existing instances: "Scale up the database server from 8 to 32 vCPUs."
Additional scaling collocations:
scale in ✅ — remove instances (reduce horizontal capacity)
scale down ✅ — reduce vertical resources
auto-scaling ✅ — automatic scaling based on load metrics
elastic scaling ✅ — cloud-native dynamic scaling
When to use each strategy:
Scale out for stateless services — web servers, microservices, API gateways (easy to add more copies)
Scale up for stateful resources — databases, caches (harder to distribute)
Both "scale out" (informal verb phrase) and "scale horizontally" (technical descriptor) are used by native speakers — "scale out" is more natural in conversation while "horizontal scaling" appears in architectural documentation.