5 exercises on the vocabulary of debugging, deploying, and incident response. These collocations appear daily in stand-ups, Slack threads, and incident channels.
0 / 5 completed
1 / 5
We ___ a hotfix to production after the P0 incident.
Deploy — the technical verb for putting code on a server:
Deploy is the precise technical term for the act of copying, configuring, and activating software on a target environment. In an incident context, "deploy a hotfix to production" is the most natural and professional collocation.
Deploy vs. Release vs. Launch:
deploy = technical act of putting code onto a server/environment. You deploy to staging, to production, to a Kubernetes cluster. Very common in incident language: "We deployed the hotfix at 14:32."
release = making a versioned artifact available. Often involves a version number: "We released v1.4.2." Implies a more formal, planned event.
launch = introducing something new to the public, often a product or major feature: "We launched the new checkout flow." Used more in product/marketing contexts than in hotfix language.
Incident collocations:
deploy a hotfix / a patch / a fix
roll back the deployment
promote to production (from staging)
push to production (informal)
2 / 5
Can you ___ this issue? I've been trying to reproduce it for an hour.
Debug — the standard developer collocation for finding the root cause of a code problem:
Debug an issue is the most natural collocation when asking a developer to find and fix a problem in code. "Debug" carries the specific meaning of stepping through code, reading logs, and identifying the source of the fault.
The developer vocabulary for investigation:
debug an issue / a bug / a problem → actively trace and investigate the code
investigate an issue → broader term, often used in incident contexts and post-mortems
reproduce a bug → confirm the bug happens consistently before debugging
diagnose a problem → identify what is wrong (used in both code and ops)
triage a bug → assess severity and priority before assigning
"Look at" is used in very informal conversation ("Can you look at this?") but is vague. "Examine" is formal and rarely used by developers in this context. "Investigate" is natural and correct, but in the specific context of a coding issue, "debug" is more precise and idiomatic.
3 / 5
The build pipeline ___ all unit tests before deploying.
Run — the dominant verb for executing automated tests and processes:
Run is the standard, idiomatic choice when describing what a pipeline or CI system does with tests. It is the most natural collocation in developer English.
Pipeline action vocabulary:
The pipeline runs tests → most natural in speech and writing
The pipeline executes tests → grammatically correct, used in formal docs and enterprise tooling (TestRail, etc.)
The pipeline builds the artefact → for the build stage
The pipeline deploys to staging → for the deploy stage
What CI pipelines do (in order):
run linting
run unit tests
run integration tests
build the Docker image or artefact
push the image to a registry
deploy to an environment
"Does" and "performs" are not natural with tests in this context. "Execute" is correct but registers as more formal than the everyday "run".
4 / 5
The on-call engineer ___ the alert and started investigating.
Acknowledge — the professional term for responding to an alert or incident:
Acknowledge an alert is the standard vocabulary in incident management and monitoring. When an on-call engineer "acknowledges" an alert, they signal to the system and the team that they have seen it and are taking ownership of the investigation. This is a specific, formal action — not just "seeing" the alert.
Incident and alert vocabulary:
acknowledge an alert / an incident → formal handover: "I've got this, I'm investigating"
triage an incident → assess severity and decide next steps
escalate an incident → bring in more senior help or wake up another team
resolve an incident → fix it and close it
declare an incident → officially open an incident for a P0/P1 situation
write / publish a post-mortem → document what happened after resolution
Monitoring platform language: PagerDuty, Opsgenie, and similar tools have an explicit "Acknowledge" button — this is why the vocabulary is so standardised. You "ack" an alert (informal short form), meaning you've taken it.
5 / 5
After fixing the bug, always ___ the fix in staging before pushing to production.
Verify, confirm, and check — all valid, with slight register differences:
All three verbs collocate naturally with "fix" in this context, and you will hear all of them in real teams. The choice depends on register:
Verify a fix: The most formal and precise term. "Verify" implies a systematic check — often used in QA, incident post-mortems, and runbooks. "Please verify the fix in staging before promoting to production."
Confirm a fix: Slightly less formal, conveys checking that the expected behaviour is now present. Often used in tickets and handoffs: "Can you confirm the fix resolves the reported issue?"
Check a fix: Most informal, most common in spoken conversation. "Just check the fix works in staging first." Natural in team Slack messages and stand-ups.
When all three appear in practice:
Runbook / documentation: "Verify the fix by navigating to the affected endpoint"
Jira comment: "Confirmed fix in staging — ready for prod"
Slack message: "Can someone check the fix on staging before we deploy?"
Related collocations:validate a fix, test the fix, sign off on the fix.