Skip to main content

pVisor case catalog

The case catalog is the executable index for pvisor run. It starts with a minimal host Run and progresses through staged changes, provider selection, network policy, Gateway capture, and prepared RunSpecs.

Use the catalog when you need a reproducible command or a regression target. Each case documents its prerequisites, command, expected result, and machine check. The checks are folded into the page so the same examples can be read by people and executed by scripts/run-pvisor-cases.py.

Choose a case group

GoalCases
First command and Run identityA01–A03
Limits and runtime controlsA04, B01–B04
Review, apply, and drop changesC01–C05
Host and VM executionD01–E06
Native OCI containersF01–F04
Network and Gateway behaviorG01–H03
RunConfig and RunSpec inputsI01–I03
Complete combinationsJ01–J03

Run the catalog

Prepare a workspace and list the available cases:

mkdir -p /tmp/pvisor-cases/workspace
cd /tmp/pvisor-cases/workspace
python3 scripts/run-pvisor-cases.py --list

Run selected cases or produce a report:

python3 scripts/run-pvisor-cases.py \
--pvisor target/release/pvisor \
--case A01,C01 \
--keep
python3 scripts/run-pvisor-cases.py \
--pvisor target/release/pvisor \
--report target/pvisor-case-report.md

The runner creates an isolated workspace per case. Replace placeholder paths and images with local resources when a case requires a Linux rootfs, OCI runtime, VM image, or Agent executable. Missing prerequisites are reported as failures so that an environment problem cannot look like a passing case.

The complete command-by-command catalog is maintained in the localized version of this page; the executable assertions below each example are language-neutral.