Files
fips/testing/ci-cleanup.sh
T
Johnathan Corgan cdda660f10 Stop local CI writing the shared mutable test image tag
The run built per-run images and then retagged them to fips-test:latest as
a compatibility bridge for the consumers that had never been migrated. The
bridge was the defect: while it existed, two concurrent runs shared one
binding, so a suite could start containers from the other run's binaries
and the verdict was recorded against a commit whose code never ran. It
fails silently by construction — the run is green either way, and nothing
compares a running container's binary against the commit under test.

Every consumer now reads the run's image, so the retag is deleted rather
than kept. That is the point of deleting it: a consumer that was missed
fails loudly instead of quietly resolving whichever run wrote the tag last.

Two consequences handled here. The cleanup script ran ip(8) inside
fips-test:latest to reap simulation interfaces, and nothing writes that tag
any more, so it now takes the caller's image, then the run's, then any
surviving test image — the last of which is what keeps an unscoped --reap
working, since that path execs before the run identity is exported. And a
guard checks statically that nothing names the shared tag, because on a
host with a hand-built copy lying around a reintroduction would run green;
it is break-checked against a reintroduced compose consumer and a
reintroduced default. Both runners gate on it, as they do the other guards.
2026-07-26 15:47:22 +00:00

302 lines
14 KiB
Bash
Executable File

#!/bin/bash
# Reap FIPS CI resources: containers, networks, volumes, images, veth pairs.
#
# Force-removes everything created by ci-local.sh that is still around —
# whether a run finished cleanly, was preempted (SIGTERM/SIGKILL), OOM-killed,
# or crashed. Two complementary selectors make this robust no matter how a
# prior run died:
#
# 1. The CI label com.corganlabs.fips-ci=1 (attached to every direct
# `docker run`/network/volume ci-local drives). Every run additionally
# stamps com.corganlabs.fips-ci.run=<run-id> on the same resources.
# 2. The compose project-name prefix fipsci_ (every compose project ci-local
# starts is named fipsci_<run-id>_<suite>, so its containers/networks/
# volumes all carry com.docker.compose.project=fipsci_... and are named
# with that prefix).
#
# The generic CI label is shared by every run on the host, so an unscoped label
# sweep would tear down a CONCURRENT run's resources. Pass --run-id to narrow
# the label sweep to one run; a run's own teardown always does. Without it the
# label sweep stays broad, which is what a manual "reap everything" wants.
#
# Host-namespace veth interfaces are the one non-docker resource reaped here.
# The chaos simulation creates each pair in the host namespace and then moves
# the two ends into container namespaces, so a run killed in between leaves the
# pair behind, and nothing else on the box removes it. Their names carry a
# short token derived from the scenario's name suffix, so --veth-suffixes takes
# the suffixes a run used and reaps only the tokens those could produce. Only a
# reap with neither --run-id nor --veth-suffixes matches every simulation veth
# name, in step with the broad label sweep. --project-prefix does not scope the
# veth sweep at all: a host interface carries no compose project.
#
# These interfaces are also the one resource reaped WITHOUT regard to the CI
# label — they cannot carry one — so an unscoped reap takes the host ends of a
# bare `chaos.sh` simulation's veth pairs too, even though it leaves that
# simulation's unlabelled containers running. Scope with --veth-suffixes (or
# just don't reap) while a bare simulation is up.
#
# Usage:
# ci-cleanup.sh Reap ALL fips-ci resources (any run)
# ci-cleanup.sh --project-prefix P Restrict the compose-project sweep to
# names starting with P (scopes the reap
# to a single run). Does NOT scope the
# host veth sweep — see --veth-suffixes
# ci-cleanup.sh --run-id ID Restrict the label sweep to the run
# labelled ID (leaves other runs alone)
# ci-cleanup.sh --label L Override the CI label (default above)
# ci-cleanup.sh --images "a b,c" Also `docker rmi -f` these image tags
# (space- or comma-separated)
# ci-cleanup.sh --veth-suffixes "a b" Restrict the host veth sweep to the
# name suffixes a single run used
# (space- or comma-separated). Without
# it the sweep reaches every simulation
# veth name on the host, including a
# bare `chaos.sh` run's live ones
#
# Safe to run when there is nothing to reap, and safe to run repeatedly.
# Also reachable as `ci-local.sh --reap`.
set -uo pipefail
SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd)"
LABEL="com.corganlabs.fips-ci=1"
RUN_LABEL_KEY="com.corganlabs.fips-ci.run"
PROJECT_PREFIX="fipsci_" # broad default: every CI run
RUN_ID="" # broad default: every CI run
IMAGES=""
VETH_SUFFIXES="" # empty AND no --run-id: every simulation veth name
# ip(8) runs inside this image, the same way the simulation creates the
# interfaces, so the reap works wherever the simulation does. Any fips test
# image will do; it is wanted only for its iproute2.
#
# Empty here and resolved after the argument loop, because the resolution has
# to consider --veth-image. The old default of fips-test:latest is no longer
# safe on its own: ci-local.sh does not write that tag, so on a host that has
# only ever run the harness it need not exist at all, and the reap this script
# advertises as the remedy for orphaned interfaces would be a permanent no-op.
VETH_IMAGE=""
while [[ $# -gt 0 ]]; do
case "$1" in
--label) LABEL="$2"; shift 2 ;;
--project-prefix) PROJECT_PREFIX="$2"; shift 2 ;;
--run-id) RUN_ID="$2"; shift 2 ;;
--images) IMAGES="$2"; shift 2 ;;
--veth-suffixes) VETH_SUFFIXES="$2"; shift 2 ;;
--veth-image) VETH_IMAGE="$2"; shift 2 ;;
-h|--help) sed -n '2,/^set /{ /^set /d; s/^# \?//; p }' "$0"; exit 0 ;;
*) echo "Unknown option: $1" >&2; exit 2 ;;
esac
done
# Resolve the image to run ip(8) in: the caller's choice, then the run's own
# image, then any surviving fips test image. That last fallback is what keeps
# an unscoped `ci-local.sh --reap` working — it execs this script from inside
# its own argument loop, before the run identity is exported, so it can pass
# neither. The empty case is handled at the point of use, which already warns
# and skips rather than failing the sweep.
if [[ -z "$VETH_IMAGE" ]]; then
VETH_IMAGE="${FIPS_TEST_IMAGE:-}"
fi
if [[ -z "$VETH_IMAGE" ]] && command -v docker >/dev/null 2>&1; then
VETH_IMAGE="$(timeout 10 docker image ls --format '{{.Repository}}:{{.Tag}}' fips-test 2>/dev/null | head -n1)"
fi
[[ -z "$VETH_IMAGE" ]] && VETH_IMAGE="fips-test:latest"
if ! command -v docker >/dev/null 2>&1; then
# No docker, nothing to reap.
exit 0
fi
if ! docker info >/dev/null 2>&1; then
# Daemon unreachable; treat as nothing to reap rather than wedging a caller.
exit 0
fi
# Each docker mutation is wrapped in `timeout` so a stuck daemon/resource can
# never wedge a caller (ci-local's signal trap relies on this being bounded).
TMO=30
# Selector for the label sweep. With --run-id it matches only the named run, so
# a concurrent run's resources are left alone; without it, every CI run.
if [[ -n "$RUN_ID" ]]; then
SWEEP_LABEL="${RUN_LABEL_KEY}=${RUN_ID}"
else
SWEEP_LABEL="$LABEL"
fi
# Distinct compose project names (read off container labels) that start with
# the configured prefix.
ci_projects() {
docker ps -a --format '{{.Label "com.docker.compose.project"}}' 2>/dev/null \
| grep -E "^${PROJECT_PREFIX}" | sort -u
}
reap_containers() {
# By CI label.
docker ps -aq --filter "label=${SWEEP_LABEL}" 2>/dev/null \
| xargs -r timeout "$TMO" docker rm -f >/dev/null 2>&1 || true
# By compose project (carried even when container_name is explicit).
local p
for p in $(ci_projects); do
docker ps -aq --filter "label=com.docker.compose.project=${p}" 2>/dev/null \
| xargs -r timeout "$TMO" docker rm -f >/dev/null 2>&1 || true
done
}
reap_networks() {
docker network ls -q --filter "label=${SWEEP_LABEL}" 2>/dev/null \
| xargs -r timeout "$TMO" docker network rm >/dev/null 2>&1 || true
# Compose networks are named <project>_<net> → match by name prefix so
# orphaned networks (whose containers are already gone) are still caught.
docker network ls --format '{{.Name}}' 2>/dev/null | grep -E "^${PROJECT_PREFIX}" \
| xargs -r timeout "$TMO" docker network rm >/dev/null 2>&1 || true
}
reap_volumes() {
docker volume ls -q --filter "label=${SWEEP_LABEL}" 2>/dev/null \
| xargs -r timeout "$TMO" docker volume rm >/dev/null 2>&1 || true
docker volume ls --format '{{.Name}}' 2>/dev/null | grep -E "^${PROJECT_PREFIX}" \
| xargs -r timeout "$TMO" docker volume rm >/dev/null 2>&1 || true
}
# Every failure below leaves interfaces behind rather than widening the sweep,
# so each one is silent by construction. Say so on stderr instead, or a reap
# that reclaimed nothing looks exactly like a reap that had nothing to reclaim.
veth_warn() { echo "ci-cleanup: host veth sweep skipped: $*" >&2; }
# One node id, exactly as chaos/sim/topology.py renders it (f"n{i+1:02d}"):
# zero-padded to two digits, and never zero-padded beyond that. Spelling it out
# keeps shapes the simulation cannot emit (vh01020a) out of the sweep.
VETH_NODE_ID='(0[0-9]|[1-9][0-9]+)'
# Regex matching the host veth names to remove. With --veth-suffixes it covers
# only the tokens those suffixes hash to, so a concurrent run's interfaces —
# which carry a different token — cannot match. The token derivation is read
# from the simulation itself rather than repeated here, so widening it cannot
# leave this matching the old width. Empty output means "reap nothing".
veth_pattern() {
# vh{token}{NN}{MM}{a,b}: the token is 4 hex or wholly absent — never a
# part of one — and the two node ids follow. Anchored and shaped this
# tightly so the unscoped sweep cannot reach an interface the simulation
# never made. Broad only for an unscoped reap: once --run-id names a
# single run, no missing or empty suffix list may widen this back out to
# every run.
if [[ -z "$RUN_ID" && -z "$VETH_SUFFIXES" ]]; then
printf '^vh([0-9a-f]{4})?%s%s[ab]$' "$VETH_NODE_ID" "$VETH_NODE_ID"
return 0
fi
if [[ -z "$VETH_SUFFIXES" ]]; then
veth_warn "--run-id given with no --veth-suffixes"
return 0
fi
local sfx=() tok alt="" out rc
read -ra sfx <<< "${VETH_SUFFIXES//,/ }"
if [[ ${#sfx[@]} -eq 0 ]]; then
veth_warn "--veth-suffixes is empty"
return 0
fi
if ! command -v python3 >/dev/null 2>&1; then
veth_warn "python3 not found, cannot derive interface tokens"
return 0
fi
# SCRIPT_DIR comes from $0, so invoking this script through a symlink or a
# copy on $PATH points PYTHONPATH at a tree with no simulation in it. Say
# which tree was tried when the derivation fails, rather than no-opping.
out="$(PYTHONPATH="$SCRIPT_DIR/chaos" python3 -m sim.naming "${sfx[@]}" 2>&1)"
rc=$?
if [[ $rc -ne 0 ]]; then
veth_warn "token derivation failed under $SCRIPT_DIR/chaos: $out"
return 0
fi
for tok in $out; do
[[ "$tok" =~ ^[0-9a-f]{4}$ ]] && alt="${alt:+$alt|}$tok"
done
if [[ -z "$alt" ]]; then
veth_warn "no interface tokens derived from: ${sfx[*]}"
return 0
fi
printf '^vh(%s)%s%s[ab]$' "$alt" "$VETH_NODE_ID" "$VETH_NODE_ID"
}
# ip(8) in a privileged --net=host container, matching how the simulation
# creates these interfaces (see chaos/sim/veth.py).
veth_ip() {
timeout "$TMO" docker run --rm --privileged --net=host \
--entrypoint ip "$VETH_IMAGE" "$@" 2>/dev/null
}
# The same, taking `ip -batch` commands on stdin so any number of deletes costs
# one container. -force keeps ip going past an interface that vanished under us.
veth_ip_batch() {
timeout "$TMO" docker run --rm -i --privileged --net=host \
--entrypoint ip "$VETH_IMAGE" -force -batch - >/dev/null 2>&1 || true
}
reap_veths() {
local pattern
pattern="$(veth_pattern)"
[[ -z "$pattern" ]] && return 0
# Without the image there is no way to run ip(8). Orphans can outlive it —
# `docker image prune -a`, a build host that prunes between runs, or a run
# whose per-run image was already reaped all remove it while interfaces are
# still up — so this is a real skip, not "nothing was ever run here".
if ! docker image inspect "$VETH_IMAGE" >/dev/null 2>&1; then
veth_warn "image $VETH_IMAGE not present, cannot run ip(8)"
return 0
fi
# Both ends of a pair match, but deleting either removes both, so collapse
# each pair to one delete and issue the lot in a single container. The
# whole script runs under a caller-imposed timeout, and a container spawn
# per name would put a large orphan set at risk of exhausting it. `sort -u`
# orders `...a` before `...b`, so the surviving end of a pair the
# simulation was killed part-way through moving is still the one picked.
local names
names="$(veth_ip -o link show \
| sed -E 's/^[0-9]+: ([^:@]+).*/\1/' \
| grep -E "$pattern" | sort -u \
| awk '{ k = substr($0, 1, length($0) - 1)
if (!(k in seen)) { seen[k] = 1; print } }')"
[[ -z "$names" ]] && return 0
sed 's/^/link delete /' <<< "$names" | veth_ip_batch
}
reap_images() {
[[ -z "$IMAGES" ]] && return 0
local imgs
read -ra imgs <<< "${IMAGES//,/ }"
[[ ${#imgs[@]} -eq 0 ]] && return 0
timeout "$TMO" docker rmi -f "${imgs[@]}" >/dev/null 2>&1 || true
}
# Per-run build contexts left in the working tree. ci-local.sh removes its own
# from the EXIT trap, but the CI worker sends SIGKILL after SIGTERM and a
# SIGKILL runs no trap, so a preempted run can leave an 18 MB directory behind
# with nothing else that would ever notice it.
#
# Scoped mode takes only the named run's. Broad mode cannot tell a live run's
# context from an abandoned one by name, so it goes by age instead: a run lasts
# well under an hour, and a day is far outside that.
reap_build_contexts() {
local dir
if [[ -n "$RUN_ID" ]]; then
dir="$SCRIPT_DIR/docker-$RUN_ID"
[[ -d "$dir" ]] && rm -rf "$dir"
return 0
fi
while IFS= read -r dir; do
[[ -n "$dir" ]] && rm -rf "$dir"
done < <(find "$SCRIPT_DIR" -maxdepth 1 -type d -name 'docker-*' -mtime +0 2>/dev/null)
return 0
}
# Order matters: containers reference networks/volumes, so drop them first, and
# the veth sweep needs an image to run ip(8) in, so it precedes the image reap.
reap_containers
reap_networks
reap_volumes
reap_veths
reap_images
reap_build_contexts
exit 0