fix: revert kickstart to near-original baseline (Step 0 — boots clean)

Reverted install.ks.ts to near-original state from commit 64533b2.
This is the bisection baseline — 21/22 integration tests pass,
0 failed systemd services, SSH works, /boot/efi mounts.

Removed all accumulated fixes that collectively broke boot:
- ERR trap, background log streamer, bastion_log/bastion_error
- depmod rebuild, nofail on /boot/efi, SELinux autorelabel
- chcon/restorecon for /etc /var /root
- kernel-modules and dosfstools packages

Kept from current branch:
- rootpw --plaintext lab-root-pw (console debug access)
- Network-first boot order (bastion controls boot)
- Vanilla role support, rancher partition support
- Boot screenshots during SSH wait (1/sec rolling buffer)
- Test runner script (run-pxe-test.sh)

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

bastion/tests/integration/pxe-provision.test.ts

@@ -27,13 +27,52 @@ import { ensurePxeNetwork, destroyPxeNetwork, deleteNftablesRejectRules, PXE_NET
 import { createPxeVm, destroyPxeVm, getVmMac, rebootPxeVm, readSerialLog } from "./helpers/pxe-vm.js";
 import { sshExec } from "./helpers/ssh.js";
 
+// --- Boot screenshot capture ---
+const SCREENSHOT_DIR = "/tmp/vm-screenshots";
+
+function startBootScreenshots(vmName: string): { stop: () => void } {
+  try { mkdirSync(SCREENSHOT_DIR, { recursive: true }); } catch {}
+  // Clean old screenshots
+  try {
+    for (const f of require("node:fs").readdirSync(SCREENSHOT_DIR)) {
+      rmSync(join(SCREENSHOT_DIR, f), { force: true });
+    }
+  } catch {}
+
+  let running = true;
+  let seq = 0;
+  const BUFFER_SIZE = 60; // keep last 60 screenshots (1 per second)
+
+  const loop = async () => {
+    while (running) {
+      try {
+        const idx = String(seq % BUFFER_SIZE).padStart(4, "0");
+        const ppm = join(SCREENSHOT_DIR, `tmp-${idx}.ppm`);
+        const png = join(SCREENSHOT_DIR, `boot-${idx}.png`);
+        execSync(`sudo virsh screenshot ${vmName} ${ppm} --screen 0 2>/dev/null`, { timeout: 3000 });
+        execSync(`convert ${ppm} ${png} 2>/dev/null && rm -f ${ppm}`, { timeout: 3000 });
+        seq++;
+      } catch {}
+      await new Promise(r => setTimeout(r, 1000));
+    }
+  };
+  loop();
+
+  return {
+    stop: () => {
+      running = false;
+      log(`Boot screenshots saved to ${SCREENSHOT_DIR}/ (${seq} captured, last ${Math.min(seq, BUFFER_SIZE)} kept)`);
+    },
+  };
+}
+
 // --- Test constants ---
 const VM_NAME = "lab-pxe-test";
 const VM_MEMORY = 4096;     // 4GB (Anaconda needs ~2GB minimum)
 const VM_VCPUS = 12;
 const VM_DISK_GB = 250;      // LVM layout needs ~204GB (swap 27 + root 33 + var 100 + etc). QCOW2 is sparse.
 const HTTP_PORT = 8099;      // Avoid conflicts with real bastion
-const SSH_USER = "root";     // Use root for SSH (admin user key setup has known issue)
+const SSH_USER = "root";     // Use root SSH for baseline testing
 const BASTION_IP = PXE_GATEWAY; // 192.168.251.1
 const DHCP_RANGE_START = `${PXE_SUBNET}.100`;
 const DHCP_RANGE_END = `${PXE_SUBNET}.200`;
@@ -41,7 +80,7 @@ const DHCP_RANGE_END = `${PXE_SUBNET}.200`;
 // Fedora install takes a while
 const DISCOVERY_TIMEOUT_MS = 5 * 60_000;    // 5 min for PXE boot + discovery
 const INSTALL_TIMEOUT_MS = 30 * 60_000;     // 30 min for full Fedora install
-const SSH_TIMEOUT_MS = 15 * 60_000;         // 15 min: PXE (~90s) + first boot + SELinux autorelabel (~3min) + reboot + second PXE (~90s) + boot
+const SSH_TIMEOUT_MS = 10 * 60_000;         // 10 min: OVMF retries PXE/HTTP Boot (~3min) before disk boot + OS startup
 
 function findSshKey(): { pubKey: string; keyPath: string } {
   const homes = [homedir()];
@@ -279,8 +318,9 @@ describe("PXE boot provisioning", () => {
     await sleep(3_000);
     deleteNftablesRejectRules();
 
-    // 10. Wait for SSH
+    // 10. Wait for SSH (with aggressive boot screenshots)
     log("Waiting for SSH access...");
+    const screenshots = startBootScreenshots(VM_NAME);
     try {
       await waitForSsh(vmIp, SSH_USER, SSH_TIMEOUT_MS, sshKeyPath);
     } catch {
@@ -292,7 +332,9 @@ describe("PXE boot provisioning", () => {
       } catch (serialErr) {
         log(`Serial console failed: ${serialErr instanceof Error ? serialErr.message : String(serialErr)}`);
       }
-      throw new Error(`SSH not available on ${vmIp} — check serial console diagnostics above`);
+      throw new Error(`SSH not available on ${vmIp} — check serial console diagnostics above. Screenshots: ${SCREENSHOT_DIR}/`);
+    } finally {
+      screenshots.stop();
     }
 
     log("PXE provision test setup complete.");
@@ -345,10 +387,10 @@ describe("PXE boot provisioning", () => {
     expect(data.progress).toBe("complete");
   });
 
-  it("log lines were captured", async () => {
+  it.skip("log lines were captured", async () => {
+    // Requires log streamer in %post — skipped until re-added
     const res = await fetch(`http://${BASTION_IP}:${HTTP_PORT}/api/logs/${encodeURIComponent(vmMac)}`);
     const data = (await res.json()) as { log_total?: number; log_lines?: Array<{ line: string }> };
-    // Should have at least some log lines from the log streamer
     expect(data.log_total).toBeGreaterThan(0);
   });
 

bastion/tests/integration/run-pxe-test.sh

@@ -0,0 +1,27 @@
+#!/bin/bash
+# One-shot PXE integration test runner.
+# Compiles, runs unit tests, cleans up, and runs the full integration test.
+set -e
+
+cd "$(dirname "$0")/../.."
+
+echo "=== Step 1: Compile ==="
+npx tsc --noEmit
+echo "✓ Compile OK"
+
+echo ""
+echo "=== Step 2: Kickstart unit tests ==="
+npx vitest run src/bastion/tests/kickstart.test.ts 2>&1 | tail -5
+echo "✓ Unit tests OK"
+
+echo ""
+echo "=== Step 3: Clean up ==="
+sudo lsof -ti:8099 2>/dev/null | xargs -r sudo kill -9 || true
+sudo virsh destroy lab-pxe-test 2>/dev/null || true
+sudo virsh undefine lab-pxe-test --nvram 2>/dev/null || true
+sudo rm -f /var/lib/libvirt/images/lab-pxe-test.qcow2
+echo "✓ Cleanup done"
+
+echo ""
+echo "=== Step 4: Integration test ==="
+npx vitest run -c /dev/null tests/integration/pxe-provision.test.ts 2>&1