Skip to content

armgit / mso

Public Imported from GitHub
0

Browser-based graphical shell and control plane for a Linux server you own — real terminal, file manager, live metrics and BYOK AI in one mobile-first workspace. Next.js 16 · React 19 · node-pty · Tailwind v4 · shadcn/ui. Public alpha, self-hosted.

fix(update): wait through systemd restart drain

Rahman32aa70df7da7
diff --git a/docs/CHANGELOG.md b/docs/CHANGELOG.md
index 9164468a..fc81ceae 100644
--- a/docs/CHANGELOG.md
+++ b/docs/CHANGELOG.md
@@ -13,6 +13,7 @@ this is the *what*, and it is what Settings → About shows as “What's new”.
**Fixed**
+- `update` wait through systemd restart drain
- `cli` diagnose HTTPS login and add doctor --fix
- `mcp` validate browser origins
- `security` avoid recovery guard fsync convoy
diff --git a/docs/PROGRESS.md b/docs/PROGRESS.md
index 52103fe6..5da20313 100644
--- a/docs/PROGRESS.md
+++ b/docs/PROGRESS.md
@@ -2,9 +2,9 @@
Running log of what shipped each phase. Newest at top.
-## 2026-08-31 — MSO terminal agent + provider-backed deployment setup — DONE IN SOURCE
+## 2026-08-31 — MSO terminal agent + provider-backed deployment setup — SHIPPED
-**Source milestone — ship only after the exact commit passes the repository/security gates and the live health SHA matches.** The standalone deployment-assistance pattern from `rahmanef63/si-coder-agent` is now absorbed into MSO's own architecture instead of spawning a second agent runtime. Installed Hermes and its public CLI were used as interaction references only; MSO has original terminal artwork, its own tool catalog, and its existing auth/audit model.
+**Shipped milestone.** The release passed the repository/security gates and the live health endpoint returned the release build SHA before this status was promoted. The standalone deployment-assistance pattern from `rahmanef63/si-coder-agent` is now absorbed into MSO's own architecture instead of spawning a second agent runtime. Installed Hermes and its public CLI were used as interaction references only; MSO has original terminal artwork, its own tool catalog, and its existing auth/audit model.
What changed:
@@ -20,7 +20,7 @@ What changed:
- official `mso-project-deploy` skill documents workflow start → inspect → provider doctor → bounded deployment/DNS operations → runtime/HTTPS verification → finish/cancel;
- CLI version for this interaction milestone is **1.5.0**. `docs/INFRASTRUCTURE-PROVIDERS.md`, architecture/install/MCP/ChatGPT connector docs, generated CLI reference and slice catalog are synchronized.
-Verification added specifically for this slice covers provider-store symlink/permission/secret behavior, Cloudflare one-record semantics, Hostinger exact RR-set payloads, Owner-only agent/provider routes, exact-payload terminal approval, CLI onboarding/update fixtures, MCP parity/rate limits, and documentation/toolset markers. The final full repository gate passes with **249 test files / 1,901 passing tests** (1 expected fail, 4 skipped), typecheck and dependency audit green, zero value cycles, and zero lint errors (three pre-existing max-lines warnings remain). A targeted Codex Security pass first found two Medium issues in the new infrastructure surface — whole-response buffering before its nominal 1 MiB cap and a stale full-zone Hostinger PUT. Both were fixed; the final infrastructure re-scan completed **1/1 component with 0 findings**. A separate terminal-agent scan found one Medium issue where a truncated approval preview could conceal mutation fields. The terminal now renders the complete canonical tool name + nested input, refuses approval payloads above 32 KiB, removes session-wide mutation grants, and binds the approval SHA-256 to the exact payload that the server independently re-hashes before dispatch; the final terminal-agent re-scan completed **1/1 component with 0 findings**. A parallel-suite-only lifecycle harness race was also removed by holding the fixture lock until the assertion explicitly releases it; the lifecycle file then passed 5 consecutive runs (40/40 checks). Live Dokploy/Cloudflare/Hostinger API success is intentionally not claimed until an owner supplies credentials and `mso provider doctor` checks that provider.
+Verification added specifically for this slice covers provider-store symlink/permission/secret behavior, Cloudflare one-record semantics, Hostinger exact RR-set payloads, Owner-only agent/provider routes, exact-payload terminal approval, CLI onboarding/update fixtures, MCP parity/rate limits, and documentation/toolset markers. The final full repository gate passes with **250 test files / 1,903 passing tests** (1 expected fail, 4 skipped), typecheck and dependency audit green, zero value cycles, and zero lint errors (three pre-existing max-lines warnings remain). A targeted Codex Security pass first found two Medium issues in the new infrastructure surface — whole-response buffering before its nominal 1 MiB cap and a stale full-zone Hostinger PUT. Both were fixed; the final infrastructure re-scan completed **1/1 component with 0 findings**. A separate terminal-agent scan found one Medium issue where a truncated approval preview could conceal mutation fields. The terminal now renders the complete canonical tool name + nested input, refuses approval payloads above 32 KiB, removes session-wide mutation grants, and binds the approval SHA-256 to the exact payload that the server independently re-hashes before dispatch; the final terminal-agent re-scan completed **1/1 component with 0 findings**. A parallel-suite-only lifecycle harness race was also removed by holding the fixture lock until the assertion explicitly releases it; the lifecycle file then passed 5 consecutive runs (40/40 checks). The first production handoff also exposed a release-verification false negative: the system unit can spend close to its 90 s `TimeoutStopSec` draining the old npm/Next control group, while the finalizer only waited 40 s for a new MainPID. Production still reached the gated release SHA, and the verifier now uses a bounded 120 s default restart budget (override constrained to 30–300 s) so a valid slow systemd restart is not reported as rollback. Live Dokploy/Cloudflare/Hostinger API success is intentionally not claimed until an owner supplies credentials and `mso provider doctor` checks that provider.
## 2026-08-31 — Login-origin diagnosis + `mso doctor --fix` + runtime hardening (SHIPPED)
diff --git a/scripts/self-update-restart-wait.test.ts b/scripts/self-update-restart-wait.test.ts
new file mode 100644
index 00000000..78ceb3ee
--- /dev/null
+++ b/scripts/self-update-restart-wait.test.ts
@@ -0,0 +1,19 @@
+import fs from "node:fs";
+import path from "node:path";
+import { describe, expect, it } from "vitest";
+
+const source = fs.readFileSync(path.join(process.cwd(), "scripts/self-update.sh"), "utf8");
+
+describe("self-update service restart verification budget", () => {
+ it("waits long enough for a normal systemd TimeoutStopSec + RestartSec cycle", () => {
+ expect(source).toContain('MSO_SERVICE_RESTART_WAIT_SECONDS:-120');
+ expect(source).toContain('RESTART_DEADLINE=$((SECONDS + RESTART_WAIT_SECONDS))');
+ expect(source).toContain('[ "$RESTART_WAIT_SECONDS" -ge 30 ]');
+ expect(source).toContain('[ "$RESTART_WAIT_SECONDS" -le 300 ]');
+ expect(source).not.toContain("for _ in $(seq 1 40); do\n CANDIDATE=");
+ });
+
+ it("reports timeout as an unverified release instead of claiming rollback", () => {
+ expect(source).toContain("release state is unverified — inspect the system unit before retrying");
+ });
+});
diff --git a/scripts/self-update.sh b/scripts/self-update.sh
index d32d59b1..c91c643d 100755
--- a/scripts/self-update.sh
+++ b/scripts/self-update.sh
@@ -47,18 +47,20 @@ HEAD_SHA="$(git rev-parse HEAD 2>/dev/null || true)"
[ -n "$HEAD_SHA" ] || die "could not read checkout HEAD"
CURRENT_BRANCH="$(git branch --show-current 2>/dev/null || true)"
[ "$CURRENT_BRANCH" = main ] || die "updates require checkout to be on main (current: ${CURRENT_BRANCH:-detached HEAD})"
-printf 'at %s
-' "${HEAD_SHA:0:7}"
+printf 'at %s\n' "${HEAD_SHA:0:7}"
if [ "$SHIP_FINALIZE" -eq 1 ]; then
EXPECTED_SHA="${MSO_EXPECTED_SHA:-}"
- [ "${#EXPECTED_SHA}" -eq 40 ] || die "ship finalizer requires an exact 40-character MSO_EXPECTED_SHA"
+ [ "${#EXPECTED_SHA}" -eq 40 ] \
+ || die "ship finalizer requires an exact 40-character MSO_EXPECTED_SHA"
case "$EXPECTED_SHA" in
*[!0-9a-f]*) die "ship finalizer requires an exact 40-character MSO_EXPECTED_SHA" ;;
esac
[ "$HEAD_SHA" = "$EXPECTED_SHA" ] || die "checkout HEAD changed after the release gates"
- [ "$(git rev-parse origin/main 2>/dev/null || true)" = "$EXPECTED_SHA" ] || die "origin/main no longer matches the release SHA"
- [ -z "$(git status --porcelain)" ] || die "checkout changed after verification; refusing to build uncommitted bytes"
+ [ "$(git rev-parse origin/main 2>/dev/null || true)" = "$EXPECTED_SHA" ] \
+ || die "origin/main no longer matches the release SHA"
+ [ -z "$(git status --porcelain)" ] \
+ || die "checkout changed after verification; refusing to build uncommitted bytes"
fi
if [ "$REBUILD_ONLY" -eq 0 ]; then
@@ -88,7 +90,8 @@ fi
if [ "$SHIP_FINALIZE" -eq 1 ]; then
step "using the exact commit already proven by the pre-push out-of-tree build"
- [ "$(git rev-parse HEAD)" = "$EXPECTED_SHA" ] && [ -z "$(git status --porcelain)" ] || die "checkout moved before the in-place build"
+ [ "$(git rev-parse HEAD)" = "$EXPECTED_SHA" ] && [ -z "$(git status --porcelain)" ] \
+ || die "checkout moved before the in-place build"
else
step "verifying the build out-of-tree (this does not touch the live .next)"
bash scripts/verify-build.sh >/dev/null || die "HEAD does not compile — nothing was deployed"
@@ -116,8 +119,22 @@ PID_UID="$(stat -c %u "/proc/$OLD_PID" 2>/dev/null || true)"
[ "$PID_UID" = "$SELF_UID" ] || die "mso.service MainPID $OLD_PID belongs to uid ${PID_UID:-unknown}, not $SELF_UID"
kill -TERM "$OLD_PID" || die "could not signal mso.service MainPID $OLD_PID"
+# A system service can legitimately spend close to TimeoutStopSec draining the old
+# control group before Restart=always creates the replacement. The installed MSO
+# unit currently inherits systemd's 90 s TimeoutStopSec; the old fixed 40 s loop
+# could therefore declare failure while systemd was still correctly stopping the
+# old Next/npm process, then the new service would appear seconds later. Keep a
+# bounded but realistic verification budget. The override is intentionally narrow
+# (30..300 s) so a typo cannot make the release finalizer hang forever.
+RESTART_WAIT_SECONDS="${MSO_SERVICE_RESTART_WAIT_SECONDS:-120}"
+case "$RESTART_WAIT_SECONDS" in
+ ""|*[!0-9]*) die "MSO_SERVICE_RESTART_WAIT_SECONDS must be an integer from 30 to 300" ;;
+esac
+[ "$RESTART_WAIT_SECONDS" -ge 30 ] && [ "$RESTART_WAIT_SECONDS" -le 300 ] \
+ || die "MSO_SERVICE_RESTART_WAIT_SECONDS must be an integer from 30 to 300"
+RESTART_DEADLINE=$((SECONDS + RESTART_WAIT_SECONDS))
NEW_PID=""
-for _ in $(seq 1 40); do
+while [ "$SECONDS" -lt "$RESTART_DEADLINE" ]; do
CANDIDATE="$(systemctl show -p MainPID --value mso.service 2>/dev/null || true)"
case "$CANDIDATE" in
""|*[!0-9]*) ;;
@@ -130,8 +147,9 @@ for _ in $(seq 1 40); do
esac
sleep 1
done
-[ -n "$NEW_PID" ] || die "mso.service did not come back with a new MainPID"
-printf 'restarted %s -> %s\n' "$OLD_PID" "$NEW_PID"
+[ -n "$NEW_PID" ] \
+ || die "mso.service did not return with a new MainPID within ${RESTART_WAIT_SECONDS}s; release state is unverified — inspect the system unit before retrying"
+printf 'restarted %s -> %s after %ss\n' "$OLD_PID" "$NEW_PID" "$((RESTART_WAIT_SECONDS - (RESTART_DEADLINE - SECONDS)))"
# The chunk-mismatch check CLAUDE.md warns about, verified rather than remembered —
# the same check scripts/ship.sh ends with. `active` only means npm has started, not