fix: host_deploy stops/starts and health-checks server-2 too (#56)

Co-authored-by: host <host@test.local>

deploy/tests/test_host_deploy.py

@@ -39,12 +39,17 @@ class HostDeploySim(unittest.TestCase):
             f.write("keep\n")
         with open(os.path.join(self.app, "old.py"), "w") as f:
             f.write("x = 0\n")
-        # 假 systemctl：is-active 一律失败（走 cloud-worker 分支），其余成功。
+        # 假 systemctl：is-active 一律失败（走 cloud-worker 分支），其余成功；
+        # cat codex-cloud-server-2 按 FAKE_SERVER2_PRESENT 决定 unit 是否存在。
         fake_bin = os.path.join(base, "fakebin")
         os.makedirs(fake_bin)
         with open(os.path.join(fake_bin, "systemctl"), "w") as f:
             f.write("#!/bin/sh\n"
                     "echo \"$@\" >> \"$FAKE_SYSTEMCTL_LOG\"\n"
+                    "if [ \"$1\" = \"cat\" ] && [ \"$2\" = \"codex-cloud-server-2\" ]; then\n"
+                    "  if [ \"$FAKE_SERVER2_PRESENT\" = \"0\" ]; then exit 1; fi\n"
+                    "  exit 0\n"
+                    "fi\n"
                     "if [ \"$1\" = \"is-active\" ]; then exit 1; fi\n"
                     "exit 0\n")
         with open(os.path.join(fake_bin, "curl"), "w") as f:
@@ -67,6 +72,7 @@ class HostDeploySim(unittest.TestCase):
         self.env["CODEX_DEPLOY_HEALTH_ATTEMPTS"] = "3"
         self.env["FAKE_HEALTH_JSON"] = self.health_path
         self.env["FAKE_SYSTEMCTL_LOG"] = self.log_path
+        self.env["FAKE_SERVER2_PRESENT"] = "1"
 
     def tearDown(self):
         self.tmp.cleanup()
@@ -86,10 +92,17 @@ class HostDeploySim(unittest.TestCase):
         sha = make_tarball(src, tarball)
         return tarball, sha
 
-    def run_host(self, tarball, sha, commit):
+    def run_host(self, tarball, sha, commit, extra_env=None):
+        env = dict(self.env)
+        if extra_env:
+            env.update(extra_env)
         return subprocess.run(
             ["sh", HOST_SCRIPT, tarball, sha, commit],
-            capture_output=True, text=True, env=self.env)
+            capture_output=True, text=True, env=env)
+
+    def systemctl_log(self):
+        with open(self.log_path) as f:
+            return f.read()
 
     def live_commit(self):
         with open(os.path.join(self.app, "DEPLOYED_COMMIT")) as f:
@@ -142,6 +155,36 @@ class HostDeploySim(unittest.TestCase):
         with open(os.path.join(self.app, "work", "keep.txt")) as f:
             self.assertEqual(f.read().strip(), "keep")
 
+    def test_server2_stopped_started_when_present(self):
+        # 2026-10-27 首推教训：server-2 存在时必须同批停/起，
+        # 否则 LB 混跑新旧内存代码。
+        self.set_health({"ok": True, "backend": "postgres", "db_ok": True,
+                         "commit": "newcommit"})
+        tarball, sha = self.build_new_tarball()
+        r = self.run_host(tarball, sha, "newcommit")
+        self.assertEqual(r.returncode, 0, r.stdout + r.stderr)
+        log = self.systemctl_log()
+        self.assertIn(
+            "stop codex-cloud-worker codex-cloud-server codex-cloud-server-2",
+            log)
+        self.assertIn(
+            "start codex-cloud-server codex-cloud-server-2 codex-cloud-worker",
+            log)
+
+    def test_server2_skipped_when_unit_absent(self):
+        # 老主机没有 server-2 unit 时退回单机行为，不报错、不提它。
+        self.set_health({"ok": True, "backend": "postgres", "db_ok": True,
+                         "commit": "newcommit"})
+        tarball, sha = self.build_new_tarball()
+        r = self.run_host(tarball, sha, "newcommit",
+                          extra_env={"FAKE_SERVER2_PRESENT": "0"})
+        self.assertEqual(r.returncode, 0, r.stdout + r.stderr)
+        # 探测行 cat 本身会提到 server-2（探测失败才判定不存在），
+        # 所以断言停/起行里没有它，而不是全日志没有这个词。
+        log = self.systemctl_log()
+        self.assertIn("stop codex-cloud-worker codex-cloud-server\n", log)
+        self.assertIn("start codex-cloud-server codex-cloud-worker\n", log)
+
 
 if __name__ == "__main__":
     unittest.main()