fix: wait for forge-core-token in prefetch instead of bailing early

On first boot (or after a wipe) hive-c0re writes forge-core-token only
after the forge container starts and the admin is provisioned. This
lags hive-c0re.service becoming active. The previous code bailed
immediately with TOKEN=placeholder if the token file was absent,
causing the runner to fail registration with 'token not found'.

Fix: merge both waits (core-token file appearance + forge API ready)
into a single 60s retry loop. The early-bail path is removed; the
script only exits cleanly if .runner is valid (writes placeholder)
or a fresh registration token is obtained.

When .runner exists but core-token is absent after 60s, we keep the
existing credentials (safe — the runner holds valid creds; next boot
will validate properly).

Closes #1224 (which tracks #1221).
This commit is contained in:
atlas 2026-06-03 22:32:53 +02:00 committed by mara
commit dacd83f278

View file

@ -26,53 +26,67 @@ let
# TOKEN= env-file populated here, never the core token itself. # TOKEN= env-file populated here, never the core token itself.
# #
# Flow: # Flow:
# 1. If forge-core-token doesn't exist yet (forge still initialising): # 1. If .runner exists: validate the runner ID against forge.
# write TOKEN=placeholder and exit — the container starts but the # Waits up to 60s for forge-core-token (hive-c0re writes it after
# runner will fail to register. systemd will restart on next boot. # forge container starts and admin is provisioned — this lags
# 2. If .runner exists: validate the runner ID against forge. # hive-c0re.service becoming active on first boot).
# 404 → purge .runner (re-registration needed). # 404 → purge .runner (re-registration needed).
# 000 (forge unreachable) → keep credentials, write placeholder. # 000 (forge unreachable) → keep credentials, write placeholder.
# other non-200 → purge .runner. # other non-200 → purge .runner.
# 3. If .runner absent or just purged: fetch a fresh registration token # Token absent after 60s → keep credentials (safe — the runner
# (retry 30s for forge to start) and write TOKEN=<real>. # holds valid creds; next boot will validate properly).
# 2. If .runner absent or just purged: wait up to 60s for both
# forge-core-token to appear AND forge API to respond, then
# fetch a fresh registration token and write TOKEN=<real>.
# Exits non-zero if both timeout — systemd logs the failure;
# the placeholder from tmpfiles means the container still starts
# but the runner will error. Operator restarts the service once
# forge is healthy.
prefetchScript = pkgs.writeShellScript "hive-ci-prefetch" '' prefetchScript = pkgs.writeShellScript "hive-ci-prefetch" ''
set -euo pipefail set -euo pipefail
TOKEN_FILE=/run/hive-ci/runner-token TOKEN_FILE=/run/hive-ci/runner-token
FORGE_URL="http://127.0.0.1:${toString forgeCfg.httpPort}" FORGE_URL="http://127.0.0.1:${toString forgeCfg.httpPort}"
RUNNER_FILE="${containerRoot}/var/lib/gitea-runner/hive/.runner" RUNNER_FILE="${containerRoot}/var/lib/gitea-runner/hive/.runner"
# Guard: if the core token hasn't been written yet (forge still
# initialising) write a placeholder and exit cleanly. The runner
# service will fail gracefully; the next boot will re-run this
# service and pick up the real token.
if [ ! -f "${coreTokenPath}" ]; then
echo "hive-ci-prefetch: core token absent, deferring registration" >&2
echo "TOKEN=placeholder" > "$TOKEN_FILE"
chmod 600 "$TOKEN_FILE"
exit 0
fi
CORE_TOKEN=$(cat ${coreTokenPath})
# Validate existing .runner credentials against forge. Purge if # Validate existing .runner credentials against forge. Purge if
# the runner was deleted (404) or the file is malformed. Keep if # the runner was deleted (404) or the file is malformed. Keep if
# forge is unreachable (000) — transient outage shouldn't wipe creds. # forge is unreachable (000) — transient outage shouldn't wipe creds.
#
# We wait up to 60s for forge-core-token to appear (hive-c0re writes
# it after provisioning the forge admin, which requires the forge
# container to start and become ready — this can lag behind
# hive-c0re.service becoming "active" on first boot). The same loop
# doubles as a wait for forge's API to become ready.
if [ -f "$RUNNER_FILE" ]; then if [ -f "$RUNNER_FILE" ]; then
RUNNER_ID=$(${pkgs.jq}/bin/jq -r '.id // empty' "$RUNNER_FILE" 2>/dev/null || true) RUNNER_ID=$(${pkgs.jq}/bin/jq -r '.id // empty' "$RUNNER_FILE" 2>/dev/null || true)
if [ -z "''${RUNNER_ID:-}" ] || [ "$RUNNER_ID" = "0" ]; then if [ -z "''${RUNNER_ID:-}" ] || [ "$RUNNER_ID" = "0" ]; then
echo "hive-ci-prefetch: .runner malformed (no id), purging for re-registration" >&2 echo "hive-ci-prefetch: .runner malformed (no id), purging for re-registration" >&2
rm -f "$RUNNER_FILE" rm -f "$RUNNER_FILE"
else else
HTTP=$(${pkgs.curl}/bin/curl -s -o /dev/null -w "%{http_code}" \ # Wait for the core token before validating (same timeout as below).
-H "Authorization: token $CORE_TOKEN" \ CORE_TOKEN=""
"$FORGE_URL/api/v1/admin/runners/$RUNNER_ID" || echo "000") for i in $(seq 1 60); do
if [ "$HTTP" = "404" ]; then if [ -f "${coreTokenPath}" ]; then
echo "hive-ci-prefetch: runner $RUNNER_ID gone from forge, purging .runner" >&2 CORE_TOKEN=$(cat ${coreTokenPath})
rm -f "$RUNNER_FILE" break
elif [ "$HTTP" = "000" ]; then fi
echo "hive-ci-prefetch: forge unreachable, keeping existing credentials" >&2 sleep 1
elif [ "$HTTP" != "200" ]; then done
echo "hive-ci-prefetch: runner validation HTTP $HTTP, purging .runner" >&2 if [ -z "''${CORE_TOKEN:-}" ]; then
rm -f "$RUNNER_FILE" echo "hive-ci-prefetch: core token absent after 60s, keeping existing .runner" >&2
else
HTTP=$(${pkgs.curl}/bin/curl -s -o /dev/null -w "%{http_code}" \
-H "Authorization: token $CORE_TOKEN" \
"$FORGE_URL/api/v1/admin/runners/$RUNNER_ID" || echo "000")
if [ "$HTTP" = "404" ]; then
echo "hive-ci-prefetch: runner $RUNNER_ID gone from forge, purging .runner" >&2
rm -f "$RUNNER_FILE"
elif [ "$HTTP" = "000" ]; then
echo "hive-ci-prefetch: forge unreachable, keeping existing credentials" >&2
elif [ "$HTTP" != "200" ]; then
echo "hive-ci-prefetch: runner validation HTTP $HTTP, purging .runner" >&2
rm -f "$RUNNER_FILE"
fi
fi fi
fi fi
fi fi
@ -85,10 +99,18 @@ let
exit 0 exit 0
fi fi
# First boot or stale creds purged — fetch a fresh registration token. # First boot or stale creds purged — wait for forge-core-token, then
# Retry up to 30s for forge to come up (containers autoStart in parallel). # fetch a fresh registration token. Single retry loop covers both:
# waiting for hive-c0re to write the token file AND for forge's API
# to become responsive (they race on first boot).
REG_TOKEN="" REG_TOKEN=""
for i in $(seq 1 30); do for i in $(seq 1 60); do
if [ ! -f "${coreTokenPath}" ]; then
echo "hive-ci-prefetch: waiting for core token (attempt $i/60)..." >&2
sleep 1
continue
fi
CORE_TOKEN=$(cat ${coreTokenPath})
REG_TOKEN=$(${pkgs.curl}/bin/curl -sf \ REG_TOKEN=$(${pkgs.curl}/bin/curl -sf \
"$FORGE_URL/api/v1/admin/runners/registration-token" \ "$FORGE_URL/api/v1/admin/runners/registration-token" \
-H "Authorization: token $CORE_TOKEN" \ -H "Authorization: token $CORE_TOKEN" \
@ -97,7 +119,7 @@ let
done done
if [ -z "''${REG_TOKEN:-}" ] || [ "$REG_TOKEN" = "null" ]; then if [ -z "''${REG_TOKEN:-}" ] || [ "$REG_TOKEN" = "null" ]; then
echo "hive-ci-prefetch: failed to fetch runner registration token from forge" >&2 echo "hive-ci-prefetch: failed to fetch runner registration token (core token absent or forge unreachable after 60s)" >&2
exit 1 exit 1
fi fi