gistfile1.txt

#!/bin/bash
# Test script to verify check-storage disk-usage produces identical output
# before and after the HealthCheckRuntime migration.
#
# Produces a single file at /tmp/hc_migration_<label>.txt that you can
# diff in any tool.
#
# Usage:
#   git checkout main && pip install -e .
#   ./scripts/test_check_migration.sh before
#
#   git checkout feature/health-check-runtime && pip install -e .
#   ./scripts/test_check_migration.sh after
#
#   diff /tmp/hc_migration_before.txt /tmp/hc_migration_after.txt
set -euo pipefail
LABEL="${1:?Usage: $0 <before|after>}"
OUTFILE="/tmp/hc_migration_${LABEL}.txt"
CLUSTER="my_cluster"
: > "$OUTFILE"
run_scenario() {
    local name="$1"
    shift
    local log_dir
    log_dir=$(mktemp -d)
    local tmp_stdout
    tmp_stdout=$(mktemp)
    local tmp_stderr
    tmp_stderr=$(mktemp)
    set +e
    health_checks "$@" --log-folder="$log_dir" > "$tmp_stdout" 2> "$tmp_stderr"
    local ec=$?
    set -e
    # Extract Nagios output line (non-JSON lines)
    local output
    output=$(grep -v '^\[' "$tmp_stdout" || true)
    # Extract and stabilize telemetry JSON (strip volatile fields)
    local telemetry
    telemetry=$(grep '^\[' "$tmp_stdout" | python3 -c "
import json, sys
for line in sys.stdin:
    records = json.loads(line)
    for r in records:
        for k in ('timestamp', 'duration_ms', 'start_time', 'end_time', 'duration_secs'):
            r.pop(k, None)
    print(json.dumps(records, indent=2, sort_keys=True))
" 2>/dev/null || true)
    # Stabilize log file (strip timestamps)
    local logfile
    logfile=$(find "$log_dir" -name '*.log' -exec cat {} \; 2>/dev/null \
        | sed 's/\[[^]]*\]/[TS]/g' || true)
    local stderr
    stderr=$(cat "$tmp_stderr")
    # Write everything to the single output file
    {
        echo "========================================"
        echo "SCENARIO: $name"
        echo "========================================"
        echo "EXIT_CODE: $ec"
        echo "--- OUTPUT ---"
        echo "$output"
        echo "--- TELEMETRY ---"
        echo "$telemetry"
        echo "--- STDERR ---"
        echo "$stderr"
        echo ""
    } >> "$OUTFILE"
    rm -f "$tmp_stdout" "$tmp_stderr"
    rm -rf "$log_dir"
    echo "  [$ec] $name: $output"
}
echo "=== Running scenarios ==="
echo "=== Writing to: $OUTFILE ==="
echo ""
# --- OK path ---
run_scenario "01_ok" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- WARN path ---
run_scenario "02_warn" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=99
# --- CRITICAL path ---
run_scenario "03_critical" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=1
# --- Multiple volumes (loop aggregation) ---
run_scenario "04_multi_volume" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / -v /tmp \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- Verbose output ---
run_scenario "05_verbose" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99 --verbose-out
# --- Output type: nagios ---
run_scenario "06_nagios" \
    check-storage disk-usage "$CLUSTER" nagios \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- Output type: app ---
run_scenario "07_app" \
    check-storage disk-usage "$CLUSTER" app \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- Output type: epilog ---
run_scenario "08_epilog" \
    check-storage disk-usage "$CLUSTER" epilog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- Inode check ---
run_scenario "09_inode" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99 --inode-check
# --- Bad volume (error path) ---
run_scenario "10_bad_volume" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v /nonexistent_vol_12345 \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- Debug log level ---
run_scenario "11_debug" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99 --log-level=DEBUG
# --- Heterogeneous cluster ---
run_scenario "12_heterogeneous" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99 --heterogeneous-cluster-v1
# --- Killswitch ON (should exit OK with killswitch message) ---
KILLSWITCH_ON=$(mktemp --suffix=.toml)
cat > "$KILLSWITCH_ON" <<'TOML'
[HealthChecksFeatures]
disable_disk_usage = true
TOML
run_scenario "13_killswitch_on" \
    --features-config="$KILLSWITCH_ON" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=1
# --- Killswitch OFF explicitly (should behave like normal) ---
KILLSWITCH_OFF=$(mktemp --suffix=.toml)
cat > "$KILLSWITCH_OFF" <<'TOML'
[HealthChecksFeatures]
disable_disk_usage = false
TOML
run_scenario "14_killswitch_off" \
    --features-config="$KILLSWITCH_OFF" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- Killswitch ON + would-be-CRITICAL (proves killswitch overrides) ---
run_scenario "15_killswitch_overrides_critical" \
    --features-config="$KILLSWITCH_ON" \
    check-storage disk-usage "$CLUSTER" nagios \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=1
# --- Killswitch ON + verbose (killswitch message should appear in output) ---
run_scenario "16_killswitch_verbose" \
    --features-config="$KILLSWITCH_ON" \
    check-storage disk-usage "$CLUSTER" nagios \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=1 --verbose-out
# --- Corrupt features config (killswitch should default to OFF) ---
KILLSWITCH_BAD=$(mktemp --suffix=.toml)
echo "not valid toml {{{{" > "$KILLSWITCH_BAD"
run_scenario "17_killswitch_bad_config" \
    --features-config="$KILLSWITCH_BAD" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=99 --usage-critical-threshold=99
# --- WARN + verbose (verify msg appears in verbose output) ---
run_scenario "18_warn_verbose" \
    check-storage disk-usage "$CLUSTER" prolog \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=99 --verbose-out
# --- CRITICAL + nagios (verify nagios format with failure) ---
run_scenario "19_critical_nagios" \
    check-storage disk-usage "$CLUSTER" nagios \
    --sink=stdout -v / \
    --usage-warning-threshold=1 --usage-critical-threshold=1
rm -f "$KILLSWITCH_ON" "$KILLSWITCH_OFF" "$KILLSWITCH_BAD"
echo ""
echo "Done. Output written to: $OUTFILE"
echo "To compare: diff /tmp/hc_migration_before.txt /tmp/hc_migration_after.txt"