5 exercises — the same verb with a different particle means something completely different. Roll back vs. roll out vs. roll over. Master the particle patterns used in deployments, Git workflows, and code reviews.
Particle patterns in IT phrasal verbs
UP → creation, start, increase: spin up · scale up · set up · stand up · build up
DOWN → stop, remove, decrease, decompose: tear down · scale down · break down · shut down · drill down
BACK → revert, return to previous state: roll back · fall back · revert back · push back
OUT → release, remove, make visible: roll out · phase out · factor out · filter out · call out
INTO → investigate, enter deeper: look into · dig into · break into · tap into
OFF → separate, start from, shut: branch off · kick off · sign off · switch off
0 / 5 completed
1 / 5
A release manager writes in the deployment runbook: "If the error rate exceeds 5% within the first 30 minutes, immediately roll _____ the release to the last stable tag." Which particle is correct?
Roll back = revert to a previous version — go backwards to a known good state.
The particle back signals direction: movement backward in time or to a prior state. This is the critical particle to know in deployment contexts.
The "roll" family in IT — each particle changes the meaning completely:
roll back → revert to a previous version: "Roll back the deployment if errors spike."
roll out → release to users gradually or fully: "Roll out the feature to 10% of users first."
roll up → combine/aggregate into a summary: "Roll up the metrics into a weekly dashboard." OR sleeve: not used technically for deployments.
roll over → transfer/carry over (budgets, IP addresses, subscriptions): "Roll over unused API credits to next month." Also: a loop in protocols — not a deployment term.
Memory trick: When something goes wrong, you go back → roll back. When you release to the world, it goes out → roll out.
2 / 5
A DevOps engineer is writing infrastructure-as-code documentation: "After the load test, _____ the environment to free up resources. We'll spin it _____ again before the next test cycle."
Tear down = shut down and completely remove an environment or resource. More than just stopping — it destroys the infrastructure so it no longer exists and incurs no cost.
Spin up = start / launch / create a resource (server, container, environment). "Spin up" uses the particle up because you're building something up from nothing.
The "up" particle family in IT deployments:
spin up → create and start: "Spin up a container"
scale up → increase capacity: "Scale up from 2 to 8 replicas"
set up → configure and make ready: "Set up the pipeline"
stand up → deploy and make live (architectural): "Stand up a new microservice"
The "down" particle family:
tear down → destroy/remove entirely: "Tear down the test environment"
scale down → decrease capacity: "Scale down overnight to save cost"
shut down → stop: "Shut down the process gracefully"
spin down → stop and deallocate: "Spin down the idle worker"
Pattern: UP = creation, starting, increasing. DOWN = removal, stopping, decreasing.
3 / 5
A developer writes a Git workflow guide. Choose the sentence where all particles are correct:
(1) "Create a feature branch by branching _____ from main." (2) "When the feature is ready, open a PR to merge it _____ main." (3) "After the PR is approved and merged, clean _____ the feature branch."
Git workflow particles:
(1) Branch off (from) = create a new branch that diverges from the source. "Off" signals separation/departure from the base. "Branch off from main" means you're starting a new path away from the main line.
(2) Merge into = incorporate changes into a target branch. "Into" signals you're putting one thing inside another — your feature branch changes go into main. Alternative: "merge in" (also used: "merge in the PR"). "Merge in" and "merge into" are both used; "merge in" treats the branch as the object, "merge into [target]" names the destination.
(3) Clean up = remove, tidy, eliminate unnecessary things. "Clean up stale branches." "Clean up the codebase." The particle "up" with "clean" means to complete the action of removing mess — common in both technical and domestic contexts.
Other Git phrasal verbs with particles:
push to / push up to → send commits to remote: "Push up to origin"
pull in / pull from → fetch and merge from remote: "Pull in the latest changes"
check out → switch to a branch: "Check out feature/login"
squash into → combine commits: "Squash the 3 commits into one before merging"
4 / 5
A senior engineer uses three phrasal verbs in a code review comment: "Can you look into the memory leak? I think we need to refactor out this helper and break down the function into smaller units." Match each phrasal verb to its meaning.
Three "investigation + refactoring" phrasal verbs, each with a different particle:
Look into = investigate, examine in detail. "Into" signals depth — you're going into the problem. "Look into the memory leak" = investigate what's causing it. Compare: look at (observe casually) vs. look into (investigate seriously).
Refactor out = remove something from its current location by restructuring it. "Out" signals removal or extraction. "Refactor out the helper" means take it away from where it is — either delete it or move it somewhere cleaner. Compare: pull out (extract), factor out (extract to a common place).
Break down = divide a complex thing into smaller, understandable parts. "Down" signals decomposition — works like dissection. "Break down the function" = split a large function into focused smaller ones. Note: break down also means "fail to work" (the server broke down), so context matters.
Particle "into" = going deeper or inside: look into · dig into · get into · go into (detail) · tap into Particle "out" = removal or making visible: refactor out · factor out · pull out · call out · point out · filter out Particle "down" = decomposing: break down · drill down · narrow down · whittle down · boil down
5 / 5
A tech lead is reviewing a PR comment. Which sentence uses the particle correctly?
Roll out = release, deploy, or distribute gradually or widely. "Out" signals going from a central point outward — to users, to servers, to the world. "Roll out the feature to 10% of users." "Roll out the change across all regions."
Why the other options are wrong:
Option A — "scale over": ❌ Not a standard IT phrasal verb. You scale up (increase), down (decrease), or out (add more instances horizontally). "Scale out" = add more nodes/instances. "Scale up" = make each node more powerful.
Option B — "break it out": The phrase "break out" exists (escape; or break out into a separate unit), but "break it out into three smaller functions" is slightly unusual. The standard phrasing is "break it down into three smaller functions" (decompose) or "extract into" (refactoring term). "Break it out" is used more for moving something to a separate location, not just dividing it.
Option D — "phase up": ❌ Not standard. You phase in (gradually introduce) or phase out (gradually retire). There is no "phase up."
Scale family summary: • scale up → more powerful resources (vertical scaling) • scale out / scale horizontally → more instances • scale down → reduce resources • scale back → reduce scope or ambition: "We scaled back the feature set for v1"