julian
ef8bd91d77
Second CI dry-run failure exposed two more issues:
1. Schema-apply runs against a fresh Postgres → fails with "Directus
isn't installed on this database. Please run 'directus bootstrap'
first." Bootstrap is what creates Directus's system tables; schema
apply requires those tables to exist. Local dev never tripped this
because bootstrap had been done in earlier sessions.
2. `node cli.js schema apply` printed an ERROR but exited 0 in the
not-installed case. schema-apply.sh trusted the exit code,
reported "schema apply complete," and the chain continued — until
the post-schema migration tried to ALTER TABLE on user tables that
never got created.
Fixes:
- entrypoint.sh: reorder steps from
pre-schema → schema-apply → post-schema → bootstrap → start
to
pre-schema → bootstrap → schema-apply → post-schema → start
Bootstrap is idempotent ("Database already initialized, skipping
install" on warm DB) so adding it earlier costs nothing on warm
boots and unblocks fresh boots.
- .gitea/workflows/build.yml: dry-run chain updated to mirror the new
entrypoint order. Bootstrap is now part of the pre-boot validation,
not skipped for speed. CI dry-run now genuinely covers the same path
the production entrypoint takes (minus the final pm2-runtime step,
which doesn't add validation value).
- scripts/schema-apply.sh: defense in depth. After the apply call
succeeds (exit 0), grep the output for ' ERROR: ' and fail loudly if
found. Catches the silent-failure pattern Directus's CLI exhibits
when bootstrap hasn't run. Error message names the likely cause
(schema-apply before bootstrap) for fast operator triage.
This is the second Phase 1 architectural correction exposed by the CI
dry-run gate. The gate is paying for itself in the very first PR it
runs against.
167 lines
6.7 KiB
Bash
Executable File
167 lines
6.7 KiB
Bash
Executable File
#!/usr/bin/env bash
|
|
# =============================================================================
|
|
# schema-apply.sh — TRM directus schema apply (image-side, boot-time)
|
|
#
|
|
# Applies the committed Directus schema snapshot to the running Postgres so
|
|
# the database schema matches what is in git. Called from entrypoint.sh as
|
|
# part of the three-step boot sequence:
|
|
# apply-db-init.sh → schema-apply.sh → directus start
|
|
#
|
|
# Usage
|
|
# Called automatically by entrypoint.sh (wired in Phase 1 task 1.7).
|
|
# Can also be run manually inside the container for debugging:
|
|
# bash scripts/schema-apply.sh
|
|
#
|
|
# Snapshot location
|
|
# /directus/snapshots/schema.yaml
|
|
# This is the image-baked path. The file is copied in by the Dockerfile.
|
|
# For the path to be overridden (e.g. in CI), set SNAPSHOT_PATH in the
|
|
# environment before calling this script.
|
|
#
|
|
# First-boot / no-snapshot behaviour
|
|
# If the snapshot file does not exist, or exists but contains only a
|
|
# .gitkeep marker (i.e. is empty or contains only whitespace), this script
|
|
# logs a skip message and exits 0. This is critical for Phase 1: tasks 1.4
|
|
# and 1.5 (collection creation) have not run yet, so there is no snapshot
|
|
# to apply. The entrypoint must not fail in this state.
|
|
#
|
|
# Dry-run preview
|
|
# Before applying, the script runs `directus schema apply --dry-run` and
|
|
# prints its output. This makes container boot logs self-explanatory:
|
|
# an operator reading logs sees exactly what is about to change.
|
|
# On a clean re-deploy where the DB already matches the snapshot, the diff
|
|
# output will show "No changes to apply" (or equivalent) and the real apply
|
|
# will be a no-op.
|
|
#
|
|
# Exit codes
|
|
# 0 Applied successfully -OR- snapshot not present / empty (skip).
|
|
# 1 directus schema apply failed (real apply, not dry-run).
|
|
# 2 directus CLI not found on PATH (image misconfiguration).
|
|
#
|
|
# Environment variables (all optional)
|
|
# SNAPSHOT_PATH Override the default /directus/snapshots/schema.yaml.
|
|
# Useful in CI where the path may differ.
|
|
# DEBUG Set to any non-empty value to enable extra verbosity.
|
|
#
|
|
# Wired into entrypoint.sh in Phase 1 task 1.7.
|
|
# =============================================================================
|
|
|
|
set -euo pipefail
|
|
|
|
# -----------------------------------------------------------------------------
|
|
# Logging helpers
|
|
# -----------------------------------------------------------------------------
|
|
|
|
log_info() {
|
|
printf '[schema-apply] %s\n' "$*"
|
|
}
|
|
|
|
log_error() {
|
|
printf '[schema-apply] ERROR: %s\n' "$*" >&2
|
|
}
|
|
|
|
# -----------------------------------------------------------------------------
|
|
# Configuration
|
|
# -----------------------------------------------------------------------------
|
|
|
|
SNAPSHOT_PATH="${SNAPSHOT_PATH:-/directus/snapshots/schema.yaml}"
|
|
|
|
# -----------------------------------------------------------------------------
|
|
# Step 0 — Verify the directus CLI is available
|
|
# -----------------------------------------------------------------------------
|
|
#
|
|
# Note: the upstream directus/directus image does NOT expose `directus` on
|
|
# PATH. The CLI is invoked as `node /directus/cli.js <subcommand>`, matching
|
|
# the upstream image's CMD (`node cli.js bootstrap && pm2-runtime ...`).
|
|
# We check that the cli.js entry script exists at the image-baked path.
|
|
|
|
readonly DIRECTUS_CLI=/directus/cli.js
|
|
|
|
if [[ ! -f "${DIRECTUS_CLI}" ]]; then
|
|
log_error "directus CLI not found at ${DIRECTUS_CLI}"
|
|
log_error "This script must run inside the directus container image."
|
|
exit 2
|
|
fi
|
|
|
|
# -----------------------------------------------------------------------------
|
|
# Step 1 — Check for snapshot existence and non-empty content
|
|
# -----------------------------------------------------------------------------
|
|
|
|
if [[ ! -f "${SNAPSHOT_PATH}" ]]; then
|
|
log_info "snapshot not found at ${SNAPSHOT_PATH} — no schema to apply, skipping"
|
|
exit 0
|
|
fi
|
|
|
|
# A file containing only a .gitkeep placeholder (empty or whitespace only)
|
|
# is treated as absent. Use `tr` to strip whitespace and check emptiness.
|
|
snapshot_content_stripped="$(tr -d '[:space:]' < "${SNAPSHOT_PATH}")"
|
|
|
|
if [[ -z "${snapshot_content_stripped}" ]]; then
|
|
log_info "snapshot at ${SNAPSHOT_PATH} is empty (placeholder only) — no schema to apply, skipping"
|
|
exit 0
|
|
fi
|
|
|
|
log_info "snapshot found at ${SNAPSHOT_PATH}"
|
|
|
|
if [[ -n "${DEBUG:-}" ]]; then
|
|
snapshot_bytes="$(wc -c < "${SNAPSHOT_PATH}" | tr -d '[:space:]')"
|
|
log_info "snapshot size: ${snapshot_bytes} bytes"
|
|
fi
|
|
|
|
# -----------------------------------------------------------------------------
|
|
# Step 2 — Dry-run preview (log the diff before applying)
|
|
# -----------------------------------------------------------------------------
|
|
|
|
log_info "--- schema diff preview (dry-run) ---"
|
|
|
|
dry_run_exit=0
|
|
# Capture output and stream it; we want each line prefixed for clarity.
|
|
while IFS= read -r line; do
|
|
log_info " ${line}"
|
|
done < <(node "${DIRECTUS_CLI}" schema apply --dry-run "${SNAPSHOT_PATH}" 2>&1) || dry_run_exit=$?
|
|
|
|
log_info "--- end diff preview ---"
|
|
|
|
# A non-zero dry-run exit is not fatal — some Directus versions exit non-zero
|
|
# when there are pending changes to report. We always proceed to the real
|
|
# apply step; only the real apply exit code is authoritative.
|
|
if [[ "${dry_run_exit}" -ne 0 ]]; then
|
|
log_info "dry-run exited ${dry_run_exit} (non-zero dry-run exit is non-fatal; proceeding to apply)"
|
|
fi
|
|
|
|
# -----------------------------------------------------------------------------
|
|
# Step 3 — Apply the snapshot
|
|
# -----------------------------------------------------------------------------
|
|
|
|
log_info "applying schema snapshot..."
|
|
|
|
apply_exit=0
|
|
apply_output=""
|
|
|
|
apply_output="$(node "${DIRECTUS_CLI}" schema apply --yes "${SNAPSHOT_PATH}" 2>&1)" || apply_exit=$?
|
|
|
|
# Always print the apply output so it appears in container logs.
|
|
while IFS= read -r line; do
|
|
log_info " ${line}"
|
|
done <<< "${apply_output}"
|
|
|
|
if [[ "${apply_exit}" -ne 0 ]]; then
|
|
log_error "directus schema apply failed (exit ${apply_exit})"
|
|
log_error "The container will not start. Fix the snapshot or the database state."
|
|
exit 1
|
|
fi
|
|
|
|
# Defense in depth: directus CLI's `schema apply` has been observed to log
|
|
# ERROR-level messages (e.g. "Directus isn't installed on this database. Please
|
|
# run \"directus bootstrap\" first.") while still exiting 0. Treat any line
|
|
# containing ' ERROR: ' (with the leading space and trailing colon — Directus's
|
|
# pino-formatted error pattern) as a fatal signal even if the CLI exited cleanly.
|
|
if grep -qE ' ERROR: ' <<< "${apply_output}"; then
|
|
log_error "directus schema apply logged ERROR-level output (CLI exited 0 but failed silently)"
|
|
log_error "Common cause: schema apply ran before directus bootstrap on a fresh DB."
|
|
log_error "Operator action: ensure entrypoint runs 'directus bootstrap' BEFORE schema-apply."
|
|
exit 1
|
|
fi
|
|
|
|
log_info "schema apply complete"
|