refactor: update paths for runtime logs

Author: takasehideki

README.md

@@ -251,7 +251,7 @@ Example:
 ```
 
 If SSH startup or readiness check fails on any Host, this command exits with a non-zero status.
-The REST server log is stored on each Host under `<remote-repo-base>/<ws-dir>/<topology>/results/runtime/rest_server.log`.
+The REST server log is stored on each Host under `<remote-repo-base>/<ws-dir>/<topology>/runtime_logs/rest_server.log`.
 For full subcommand and option details (including `wait`, `monitor`, `logs`, and related options), see [manager_scripts/README.md](./manager_scripts/README.md#manage_rest_serverssh).
 
 If the server exits at startup with a chrony sudo permission error, check the chrony sudo setup in [Clock synchronization for REST benchmark (chrony)](#clock-synchronization-for-rest-benchmark-chrony).
@@ -398,7 +398,7 @@ On prepare, the Manager creates `<ws-dir>/<topology>/results/<session_timestamp>
 - Trial logs are collected under `<ws-dir>/<topology>/results/latest-<rmw>/raw_logs/trial<N>/`.
 - Aggregated outputs such as `total_latency.csv`, `throughput.csv`, `host_trials_usage.csv`, and `host_usage_summary.csv` are written under `<ws-dir>/<topology>/results/latest-<rmw>/analysis/`.
 - In `docker`/`native` modes, runtime service logs are collected under `<ws-dir>/<topology>/results/latest-<rmw>/runtime_logs/` (for example, `<host>_rest_server.log`; and `zenohd_router.log` for Zenoh router runs).
-  - Note: `<host>_rest_server.log` is copied from the long-lived REST service log (`results/runtime/rest_server.log`), so it may include entries from earlier benchmark runs unless the REST server was restarted.
+  - Note: `<host>_rest_server.log` is copied from the long-lived REST service log (`<ws-dir>/<topology>/runtime_logs/rest_server.log`), so it may include entries from earlier benchmark runs unless the REST server was restarted.
 
 For details on output directory structure and CSV column definitions, see [performance_test/README.md](./performance_test/README.md).
 

manager_scripts/README.md

@@ -120,7 +120,7 @@ If command execution or readiness check fails on any Host, the script exits with
 
 Runtime log files created by this script are stored on each Host under:
 
-- `<remote-repo-base>/<ws-dir>/<topology>/results/runtime/rest_server.log`
+- `<remote-repo-base>/<ws-dir>/<topology>/runtime_logs/rest_server.log`
 
 The `logs` subcommand reads `rest_server.log`.
 `start` / `stop` / `status` operate on the process listening on fixed REST port `5000` (no PID file management).

manager_scripts/manage_rest_servers.sh

@@ -221,7 +221,7 @@ if [[ -z "${HOSTS_LINE}" ]]; then
     exit 1
 fi
 
-REMOTE_RUNTIME_DIR="${REMOTE_REPO_BASE}/${WS_DIR}/${TOPOLOGY_DIR}/results/runtime"
+REMOTE_RUNTIME_DIR="${REMOTE_REPO_BASE}/${WS_DIR}/${TOPOLOGY_DIR}/runtime_logs"
 
 IFS=',' read -r -a HOSTS_RAW <<< "${HOSTS_LINE}"
 HOSTS=()

performance_test/README.md

@@ -30,15 +30,16 @@ Use this mode for CI or formal evaluations where partially valid totals are not
 
 ## Output Structure
 
-`performance_test.py` creates run-scoped outputs under `<ws-dir>/<topology>/results/<timestamp>-<rmw>/`, and related tools may also manage shared runtime logs under `<ws-dir>/<topology>/results/runtime/`:
+`performance_test.py` creates run-scoped outputs under `<ws-dir>/<topology>/results/<timestamp>-<rmw>/`, and related tools may also manage shared runtime logs under `<ws-dir>/<topology>/runtime_logs/`:
 
 ```
+runtime_logs/
+├── rest_server.log                  # managed by manage_rest_servers.sh
+└── zenohd_router.log                # created when rmw_zenohd is started natively
+
 results/
 ├── latest-fastdds -> 2026-04-26_13-21-45-fastdds/   # symlink per RMW
 ├── latest-zenoh   -> 2026-04-26_14-02-10-zenoh/
-├── runtime/
-│   ├── rest_server.log              # managed by manage_rest_servers.sh
-│   └── zenohd_router.log            # created when rmw_zenohd is started natively
 └── 2026-04-26_13-21-45-fastdds/
     ├── system_perf/
     │   ├── chrony_check/
@@ -81,7 +82,7 @@ results/
 ```
 
 `runtime_logs/<host>_rest_server.log` is a snapshot copy of each Host's long-lived
-`results/runtime/rest_server.log`, so it may include entries from previous runs
+`<ws-dir>/<topology>/runtime_logs/rest_server.log`, so it may include entries from previous runs
 unless the REST server was restarted before benchmarking.
 
 ## CSV Formats

performance_test/runner.py

@@ -292,7 +292,7 @@ def collect_runtime_logs(
 
     remote_repo_root = os.environ.get("ROS2_PERF_REPO_ROOT", remote_repo_base)
     remote_runtime_dir = (
-        f"{remote_repo_root}/{ws_dir}/{topology_name}/results/runtime"
+        f"{remote_repo_root}/{ws_dir}/{topology_name}/runtime_logs"
     )
 
     for host in hosts:
@@ -359,7 +359,7 @@ def collect_runtime_logs(
         dst = os.path.join(runtime_logs_dir, "zenohd_router.log")
         if exec_policy == "native":
             src = os.path.join(
-                local_repo_root, ws_dir, topology_name, "results", "runtime", "zenohd_router.log"
+                local_repo_root, ws_dir, topology_name, "runtime_logs", "zenohd_router.log"
             )
             if os.path.exists(src):
                 shutil.copy2(src, dst)

performance_test/zenoh_runtime.py

@@ -71,7 +71,7 @@ def build_config_override(connect_host):
 
 
 def _zenoh_router_runtime_dir(base_dir, ws_dir, topology_name):
-    return os.path.join(base_dir, ws_dir, topology_name, "results", "runtime")
+    return os.path.join(base_dir, ws_dir, topology_name, "runtime_logs")
 
 
 def _find_local_pid_by_port(port):