feat: activate dormant plugin health/metrics subsystem and surface it in the web UI (#388)

* feat(plugin-system): activate dormant plugin health & metrics subsystem

PluginManager shipped a fully-built health tracker, resource monitor and
circuit breaker that were never instantiated (health_tracker/resource_monitor
were left as None), so the circuit breaker never engaged and the existing
health/metrics API routes always returned "not available".

- DisplayController now wires a PluginHealthTracker and PluginResourceMonitor
  onto the plugin manager, enabling the circuit breaker (a repeatedly-failing
  plugin's update() is skipped after consecutive failures, then retried after
  a cooldown) and per-plugin execution-time metrics. Both persist to the
  shared cache.
- load_plugin() now validates each plugin's config against its JSON schema in
  a strictly warn/degrade-only way: a violation logs a warning and flags the
  plugin degraded in the health tracker, but never changes whether the plugin
  loads or its pass/fail behaviour. Adds PluginHealthTracker.set_degraded(),
  which never touches the circuit breaker.
- ResourceMonitor CPU/memory sampling now reuses a cached psutil.Process and
  reads cpu_percent(interval=None), so monitoring no longer blocks ~100ms per
  call on the display loop's update path.
- Fix DiskCache.get() raising TypeError for max_age=None ("never expires"),
  which silently discarded persisted plugin health/metrics on read and thus
  broke cross-process and post-restart surfacing.
- Fix two dead PluginManager helpers that called non-existent tracker methods.

Tests: new test_resource_monitor, test_plugin_health,
test_plugin_manager_schema_soft; extended test_cache_manager and
test_display_controller.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UvTav268UXv44ub9K11LYq

* feat(web-ui): surface plugin health, metrics and load state

With the health/metrics subsystem now active in the display service, expose it
in the web UI (which runs as a separate process from the display loop):

- Wire a health tracker / resource monitor backed by the shared on-disk cache
  into the web process so /api/v3/plugins/health and /plugins/metrics read the
  data the display service persists.
- Build those route responses per installed plugin id (the tracker's in-memory
  view is empty in a fresh web process) so cross-process data is included.
- Add state + error_info to /plugins/installed entries so the UI can show why a
  plugin isn't running instead of just loaded:false.
- Add a "Plugin Health" panel to the Tools page (circuit status, avg/max update
  time, update count, last error) plus PluginAPI.getPluginMetrics().

Tests: route-level tests for the health/metrics endpoints in test_web_api.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UvTav268UXv44ub9K11LYq

* fix(plugin-metrics): refresh cross-process health/metrics reads; type hints

Addresses CodeRabbit review on #388:

- Major: the web process's health/resource trackers cached the first persisted
  read in an in-memory dict (and the CacheManager memory tier held max_age=None
  entries indefinitely), so a long-lived web process showed the first snapshot
  and never reflected the display service's later updates. Add an opt-in
  force_reload path (get_health_summary/get_health_state/_load_health_state and
  get_metrics_summary/get_metrics) that bypasses the in-memory copy and, via a
  new memory_ttl passthrough on CacheManager.get, the cache manager's memory
  tier — so each /plugins/health and /plugins/metrics poll reads fresh persisted
  state. Default behaviour (force_reload=False) is unchanged for the display
  process and existing callers.
- Minor: DiskCache.get type hint is now Optional[int] with the None ("never
  expires") semantics documented, matching MemoryCache.get.

Tests: new force_reload staleness cases in test_plugin_health and
test_resource_monitor.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UvTav268UXv44ub9K11LYq

---------

Co-authored-by: Claude <noreply@anthropic.com>
This commit is contained in:
Chuck
2026-07-09 07:54:18 -04:00
committed by GitHub
co-authored by Claude Opus 4.8
parent 85d321cf33
commit 3b93024993
16 changed files with 797 additions and 49 deletions
+16
View File
@@ -160,6 +160,22 @@ api_v3.health_monitor = health_monitor
from src.cache_manager import CacheManager
api_v3.cache_manager = CacheManager()
# Wire plugin health/metrics for the web process. The display service records
# health and execution-time metrics to the shared on-disk cache; giving the web
# process its own tracker/monitor backed by that same cache lets the health API
# routes (/api/v3/plugins/health, /plugins/metrics) read that persisted data.
# Guarded so any init failure degrades to "not available" rather than breaking
# the web server.
try:
from src.plugin_system.plugin_health import PluginHealthTracker
from src.plugin_system.resource_monitor import PluginResourceMonitor
plugin_manager.health_tracker = PluginHealthTracker(api_v3.cache_manager)
plugin_manager.resource_monitor = PluginResourceMonitor(api_v3.cache_manager)
except Exception as _hm_err: # pragma: no cover - defensive startup guard
logging.getLogger(__name__).warning(
"Could not enable plugin health/metrics for web UI: %s", _hm_err
)
app.register_blueprint(pages_v3, url_prefix='/v3')
app.register_blueprint(api_v3, url_prefix='/api/v3')
+72 -4
View File
@@ -2073,6 +2073,18 @@ def get_installed_plugins():
return None
def _build_plugin_entry_inner(plugin_info, plugin_id):
# Capture runtime state (state machine + error context) before the
# manifest merge below can shadow the 'state' key. get_all_plugin_info
# attaches this via PluginStateManager.get_state_info(); surfacing it
# lets the UI show *why* a plugin isn't running instead of just
# 'loaded: false'.
state_info = plugin_info.get('state')
plugin_state = None
plugin_error_info = None
if isinstance(state_info, dict):
plugin_state = state_info.get('state')
plugin_error_info = state_info.get('error_info')
# Re-read manifest from disk to ensure we have the latest metadata
manifest_path = Path(api_v3.plugin_manager.plugins_dir) / plugin_id / "manifest.json"
if manifest_path.exists():
@@ -2154,6 +2166,8 @@ def get_installed_plugins():
'enabled': enabled,
'verified': verified,
'loaded': plugin_info.get('loaded', False),
'state': plugin_state,
'error_info': plugin_error_info,
'last_updated': last_updated,
'last_commit': last_commit,
'last_commit_message': last_commit_message,
@@ -2173,6 +2187,31 @@ def get_installed_plugins():
logger.error('Error in get_installed_plugins', exc_info=True)
return jsonify({'status': 'error', 'message': 'An error occurred; see logs for details'}), 500
def _installed_plugin_ids():
"""Best-effort list of installed plugin IDs for the web process.
Health/metrics state is written by the separate display service to the
shared on-disk cache, so the tracker's in-memory set is empty here. We
enumerate the installed plugins and read each one's persisted summary by ID
instead of relying on the tracker's in-memory `get_all_*` view.
"""
pm = api_v3.plugin_manager
manifests = getattr(pm, 'plugin_manifests', None)
if not manifests:
# Only pay for a discovery scan when we haven't discovered anything yet;
# subsequent polls reuse the already-populated manifest map.
try:
pm.discover_plugins()
except Exception:
logger.debug('discover_plugins failed while listing plugin ids', exc_info=True)
manifests = getattr(pm, 'plugin_manifests', None)
try:
return list(manifests.keys()) if manifests else []
except Exception:
logger.debug('listing plugin_manifests failed while building plugin ids', exc_info=True)
return []
@api_v3.route('/plugins/health', methods=['GET'])
def get_plugin_health():
"""Get health metrics for all plugins"""
@@ -2188,8 +2227,23 @@ def get_plugin_health():
'message': 'Health tracking not available'
})
# Get health summaries for all plugins
health_summaries = api_v3.plugin_manager.health_tracker.get_all_health_summaries()
tracker = api_v3.plugin_manager.health_tracker
# Build per-plugin summaries by ID so persisted (cross-process) health
# is included, then fold in any in-memory-only entries.
health_summaries = {}
for pid in _installed_plugin_ids():
try:
# force_reload: this process only reads; bypass the in-memory
# snapshot so each poll reflects the display service's latest
# persisted state.
health_summaries[pid] = tracker.get_health_summary(pid, force_reload=True)
except Exception:
logger.debug('Could not read health summary for %s', pid, exc_info=True)
try:
for pid, summary in tracker.get_all_health_summaries().items():
health_summaries.setdefault(pid, summary)
except Exception:
logger.debug('get_all_health_summaries failed', exc_info=True)
return jsonify({
'status': 'success',
@@ -2264,8 +2318,22 @@ def get_plugin_metrics():
'message': 'Resource monitoring not available'
})
# Get metrics summaries for all plugins
metrics_summaries = api_v3.plugin_manager.resource_monitor.get_all_metrics_summaries()
monitor = api_v3.plugin_manager.resource_monitor
# Build per-plugin summaries by ID so persisted (cross-process) metrics
# are included, then fold in any in-memory-only entries.
metrics_summaries = {}
for pid in _installed_plugin_ids():
try:
# force_reload: read-only path — bypass the in-memory snapshot so
# each poll reflects the display service's latest persisted metrics.
metrics_summaries[pid] = monitor.get_metrics_summary(pid, force_reload=True)
except Exception:
logger.debug('Could not read metrics summary for %s', pid, exc_info=True)
try:
for pid, summary in monitor.get_all_metrics_summaries().items():
metrics_summaries.setdefault(pid, summary)
except Exception:
logger.debug('get_all_metrics_summaries failed', exc_info=True)
return jsonify({
'status': 'success',
@@ -340,11 +340,25 @@ const PluginAPI = {
* @returns {Promise<Object>} Health data
*/
async getPluginHealth(pluginId = null) {
const endpoint = pluginId
const endpoint = pluginId
? `/plugins/health/${pluginId}`
: '/plugins/health';
const response = await this.request(endpoint);
return response.data || {};
},
/**
* Get plugin resource metrics (execution time, memory, cpu).
*
* @param {string} pluginId - Optional plugin identifier (null for all)
* @returns {Promise<Object>} Metrics data keyed by plugin id
*/
async getPluginMetrics(pluginId = null) {
const endpoint = pluginId
? `/plugins/metrics/${pluginId}`
: '/plugins/metrics';
const response = await this.request(endpoint);
return response.data || {};
}
};
@@ -157,6 +157,38 @@
</div>
</div>
<!-- Plugin Health -->
<div class="bg-white rounded-lg shadow p-6">
<div class="border-b border-gray-200 pb-4 mb-6 flex items-start justify-between gap-4">
<div>
<h2 class="text-lg font-semibold text-gray-900">Plugin Health</h2>
<p class="mt-1 text-sm text-gray-600">Circuit-breaker status and per-plugin update timings recorded by the display service. A plugin whose <code class="bg-gray-100 px-1 rounded">update()</code> keeps failing is paused ("Circuit open") and retried automatically after a cooldown.</p>
</div>
<button id="btn-plugin-health-refresh" onclick="refreshPluginHealth(true)"
class="shrink-0 inline-flex items-center px-3 py-2 border border-gray-300 text-sm font-medium rounded-md text-gray-700 bg-white hover:bg-gray-50">
<i class="fas fa-sync-alt mr-2"></i>Refresh
</button>
</div>
<div id="plugin-health-message" class="hidden mb-4 text-sm text-gray-500"></div>
<div class="overflow-x-auto">
<table class="min-w-full divide-y divide-gray-200">
<thead class="bg-gray-50">
<tr>
<th class="px-4 py-3 text-left text-xs font-medium text-gray-500 uppercase tracking-wider">Plugin</th>
<th class="px-4 py-3 text-left text-xs font-medium text-gray-500 uppercase tracking-wider">Status</th>
<th class="px-4 py-3 text-right text-xs font-medium text-gray-500 uppercase tracking-wider">Avg update</th>
<th class="px-4 py-3 text-right text-xs font-medium text-gray-500 uppercase tracking-wider">Max update</th>
<th class="px-4 py-3 text-right text-xs font-medium text-gray-500 uppercase tracking-wider">Updates</th>
<th class="px-4 py-3 text-left text-xs font-medium text-gray-500 uppercase tracking-wider">Last error</th>
</tr>
</thead>
<tbody id="plugin-health-tbody" class="bg-white divide-y divide-gray-200">
<tr><td colspan="6" class="px-4 py-8 text-center text-gray-500">Loading…</td></tr>
</tbody>
</table>
</div>
</div>
</div>
<script>
@@ -397,7 +429,80 @@
});
}
// ── plugin health panel ──────────────────────────────────────────────────
function phEscape(s) {
return String(s).replace(/[&<>"']/g, function (c) {
return { '&': '&amp;', '<': '&lt;', '>': '&gt;', '"': '&quot;', "'": '&#39;' }[c];
});
}
function phFmtSecs(v) {
if (typeof v !== 'number' || !isFinite(v)) return '—';
return v.toFixed(3) + 's';
}
function phStatus(h) {
if (!h) return { label: 'Unknown', cls: 'warning' };
if (h.circuit_state === 'open') return { label: 'Circuit open', cls: 'error' };
if (h.circuit_state === 'half_open') return { label: 'Recovering', cls: 'warning' };
if (h.degraded) return { label: 'Degraded', cls: 'warning' };
if (h.is_healthy) return { label: 'Healthy', cls: 'success' };
return { label: 'Unknown', cls: 'warning' };
}
async function refreshPluginHealth(force) {
const tbody = document.getElementById('plugin-health-tbody');
const msg = document.getElementById('plugin-health-message');
if (!tbody || !window.PluginAPI) return;
try {
if (force && PluginAPI.clearCache) PluginAPI.clearCache();
const results = await Promise.all([
PluginAPI.getPluginHealth(),
PluginAPI.getPluginMetrics()
]);
const health = results[0] || {};
const metrics = results[1] || {};
const ids = Array.from(new Set(Object.keys(health).concat(Object.keys(metrics)))).sort();
if (!ids.length) {
if (msg) {
msg.textContent = 'No plugin health data yet — it appears once the display service has run plugins.';
msg.classList.remove('hidden');
}
tbody.innerHTML = '<tr><td colspan="6" class="px-4 py-8 text-center text-gray-500">No data</td></tr>';
return;
}
if (msg) msg.classList.add('hidden');
let rows = '';
ids.forEach(function (id) {
const h = health[id] || {};
const m = metrics[id] || {};
const st = phStatus(h);
const lastErr = h.degraded_reason || h.last_error || '';
const calls = (typeof m.call_count === 'number') ? m.call_count : '—';
const errCell = lastErr
? '<span title="' + phEscape(lastErr) + '">' + phEscape(lastErr) + '</span>'
: '<span class="text-gray-400">—</span>';
rows += '<tr>' +
'<td class="px-4 py-3 whitespace-nowrap text-sm font-medium text-gray-900">' + phEscape(id) + '</td>' +
'<td class="px-4 py-3 whitespace-nowrap"><span class="status-indicator ' + st.cls + '">' + st.label + '</span></td>' +
'<td class="px-4 py-3 whitespace-nowrap text-right text-sm text-gray-600">' + phFmtSecs(m.avg_execution_time) + '</td>' +
'<td class="px-4 py-3 whitespace-nowrap text-right text-sm text-gray-600">' + phFmtSecs(m.max_execution_time) + '</td>' +
'<td class="px-4 py-3 whitespace-nowrap text-right text-sm text-gray-600">' + calls + '</td>' +
'<td class="px-4 py-3 text-sm text-red-600 max-w-xs truncate">' + errCell + '</td>' +
'</tr>';
});
tbody.innerHTML = rows;
} catch (e) {
const emsg = (e && e.message) ? e.message : String(e);
tbody.innerHTML = '<tr><td colspan="6" class="px-4 py-6 text-center text-red-500">Failed to load plugin health: ' + phEscape(emsg) + '</td></tr>';
}
}
window.refreshPluginHealth = refreshPluginHealth;
// Load on first render; HTMX will have already swapped us in by this point.
loadGitInfo();
refreshPluginHealth(false);
// Refresh periodically. Guard against duplicate timers if this partial is
// re-swapped in by HTMX; the handler re-resolves DOM nodes by id each tick.
if (!window._pluginHealthTimer) {
window._pluginHealthTimer = setInterval(function () { refreshPluginHealth(true); }, 15000);
}
})();
</script>