Přidal jsem Health Monitoring do portálu:

Přehled má nové karty: Zdravé služby, Nezdravé služby, Nedostupné služby.
Přehled zobrazuje sekci Služby vyžadující pozornost pro unhealthy a unreachable.
Snapshot /portal/ws/operations teď obsahuje health data a stránka je aktualizuje přes existující realtime vrstvu.
Seznam služeb zobrazuje aktuální health status a poslední kontrolu, včetně řazení Podle zdraví.
Detail služby má sekce Zdraví služby a Historie kontrol s posledními 50 záznamy.
Přidal jsem audit event service.health.view.
Browser title teď používá formát CSBot Services Portal - ..., takže detail služby odpovídá požadavku.
Nepoužil jsem ORM, vše je přes SQLite dotazy.

app/db/health.py

@@ -0,0 +1,102 @@
+from app.db.database import get_connection
+
+
+HEALTH_STATUSES = ("healthy", "unhealthy", "unreachable")
+
+
+def get_latest_service_health():
+    con = get_connection()
+
+    try:
+        rows = con.execute(
+            """
+            SELECT
+                h.service_id,
+                COALESCE(a.name, h.service_id) AS service_name,
+                h.status,
+                h.http_status,
+                h.response_time_ms,
+                h.error_text,
+                h.checked_at
+            FROM service_health h
+            LEFT JOIN apps a ON a.id = h.service_id
+            INNER JOIN (
+                SELECT service_id, MAX(id) AS max_id
+                FROM service_health
+                GROUP BY service_id
+            ) latest ON latest.max_id = h.id
+            """
+        ).fetchall()
+    except Exception:
+        rows = []
+
+    con.close()
+    return {row["service_id"]: dict(row) for row in rows}
+
+
+def get_service_health(service_id: str):
+    con = get_connection()
+
+    try:
+        row = con.execute(
+            """
+            SELECT
+                h.service_id,
+                COALESCE(a.name, h.service_id) AS service_name,
+                h.status,
+                h.http_status,
+                h.response_time_ms,
+                h.error_text,
+                h.checked_at
+            FROM service_health h
+            LEFT JOIN apps a ON a.id = h.service_id
+            WHERE h.service_id = ?
+            ORDER BY h.id DESC
+            LIMIT 1
+            """,
+            (service_id,),
+        ).fetchone()
+    except Exception:
+        row = None
+
+    con.close()
+    return dict(row) if row else None
+
+
+def get_service_health_history(service_id: str, limit: int = 50):
+    con = get_connection()
+
+    try:
+        rows = con.execute(
+            """
+            SELECT id, service_id, status, http_status, response_time_ms, error_text, checked_at
+            FROM service_health
+            WHERE service_id = ?
+            ORDER BY id DESC
+            LIMIT ?
+            """,
+            (service_id, limit),
+        ).fetchall()
+    except Exception:
+        rows = []
+
+    con.close()
+    return [dict(row) for row in rows]
+
+
+def get_health_summary():
+    latest = get_latest_service_health()
+    summary = {"healthy": 0, "unhealthy": 0, "unreachable": 0}
+    for item in latest.values():
+        status = item.get("status")
+        if status in summary:
+            summary[status] += 1
+    return summary
+
+
+def get_problem_service_health():
+    latest = get_latest_service_health()
+    return [
+        item for item in latest.values()
+        if item.get("status") in {"unhealthy", "unreachable"}
+    ]

app/db/operations.py

@@ -1,4 +1,5 @@
 from app.db.database import get_connection
+from app.db.health import get_health_summary, get_problem_service_health
 from app.db.migrations import run_migrations
 from app.db.workers import is_worker_online
 
@@ -197,6 +198,8 @@ def get_operations_snapshot():
         WHERE LOWER(COALESCE(status, '')) IN ('failed', 'error', 'disabled', 'deleted', 'archived')
         """
     ).fetchone()["count"]
+    health_summary = get_health_summary()
+    health_problems = get_problem_service_health()
 
     jobs_summary = con.execute(
         """
@@ -296,6 +299,12 @@ def get_operations_snapshot():
             "active": active_applications,
             "problematic": problematic_applications,
         },
+        "health": {
+            "healthy": health_summary["healthy"],
+            "unhealthy": health_summary["unhealthy"],
+            "unreachable": health_summary["unreachable"],
+            "problems": health_problems,
+        },
         "workers": {
             "online": workers_online,
             "offline": max(0, total_workers - workers_online),

app/routes/apps.py

@@ -17,6 +17,7 @@ from ..config import (
 )
 from ..db.apps import get_app, get_app_deployments, get_apps, update_app_resources
 from ..db.audit import log_audit_event
+from ..db.health import get_latest_service_health, get_service_health, get_service_health_history
 from ..db.jobs import create_job, get_jobs, has_active_deploy_job
 from ..routes.deployments import render_status_pill
 from ..shell import run_command
@@ -26,6 +27,24 @@ router = APIRouter()
 DEFAULT_PAGE_SIZE = 20
 
 
+def render_health_status(status: str | None) -> str:
+    value = status or ""
+    normalized = value.lower()
+    labels = {
+        "healthy": "zdravá",
+        "unhealthy": "nezdravá",
+        "unreachable": "nedostupná",
+    }
+    class_name = "pill pill-muted"
+    if normalized == "healthy":
+        class_name = "pill pill-success"
+    elif normalized == "unhealthy":
+        class_name = "pill pill-warning"
+    elif normalized == "unreachable":
+        class_name = "pill pill-danger"
+    return f'<span class="{class_name}">{html.escape(labels.get(normalized, value or "neznámá"))}</span>'
+
+
 @router.get("/")
 def portal_home(user=Depends(require_user)):
     return RedirectResponse(url="/portal/operations", status_code=303)
@@ -40,6 +59,7 @@ def apps_page(
     user=Depends(require_user),
 ):
     apps = get_apps()
+    latest_health = get_latest_service_health()
     query = q.strip()
     selected_status = status.strip()
     if query:
@@ -50,6 +70,16 @@ def apps_page(
         ]
     if selected_status:
         apps = [item for item in apps if (item.get("status", "") or "") == selected_status]
+    sort = request.query_params.get("sort", "").strip()
+    if sort == "health":
+        order = {"unreachable": 0, "unhealthy": 1, "healthy": 2}
+        apps = sorted(
+            apps,
+            key=lambda item: (
+                order.get((latest_health.get(item.get("id", "")) or {}).get("status"), 3),
+                item.get("id", ""),
+            ),
+        )
     status_values = sorted({item.get("status", "") for item in get_apps() if item.get("status")})
     total_apps = len(apps)
     total_pages = max(1, math.ceil(total_apps / DEFAULT_PAGE_SIZE))
@@ -68,6 +98,9 @@ def apps_page(
         app_id = html.escape(item.get("id", ""))
         app_url_id = quote(item.get("id", ""), safe="")
         status = html.escape(item.get("status", ""))
+        health = latest_health.get(item.get("id", "")) or {}
+        health_status = render_health_status(health.get("status"))
+        health_checked_at = html.escape(health.get("checked_at", "") or "")
         docs = html.escape(item.get("docs", f"/apps/{app_id}/docs"))
         memory = item.get("memory", "")
         cpus = item.get("cpus", "")
@@ -104,6 +137,7 @@ def apps_page(
                 <span class="muted">/apps/{app_id}</span>
             </td>
             <td><span class="pill">{status}</span></td>
+            <td>{health_status}<br><span class="muted">{health_checked_at}</span></td>
             <td><a href="{docs}">Swagger</a></td>
             <td>
                 <form method="post" action="/portal/update-resources">
@@ -150,7 +184,7 @@ def apps_page(
         """
 
     if not rows:
-        rows = '<tr><td colspan="6">Zatím nejsou nasazené žádné služby.</td></tr>'
+        rows = '<tr><td colspan="7">Zatím nejsou nasazené žádné služby.</td></tr>'
     status_options = ['<option value="">Všechny stavy</option>']
     for value in status_values:
         selected = " selected" if selected_status == value else ""
@@ -165,6 +199,8 @@ def apps_page(
             params["q"] = query
         if selected_status:
             params["status"] = selected_status
+        if sort:
+            params["sort"] = sort
         return f"/portal/apps?{urlencode(params)}"
 
     previous_link = (
@@ -216,6 +252,10 @@ def apps_page(
             <form method="get" action="/portal/apps" class="filter-form">
                 <input name="q" value="{html.escape(query)}" placeholder="Název nebo ID služby">
                 <select name="status">{"".join(status_options)}</select>
+                <select name="sort">
+                    <option value="">Výchozí řazení</option>
+                    <option value="health"{" selected" if sort == "health" else ""}>Podle zdraví</option>
+                </select>
                 <input type="hidden" name="page" value="1">
                 <button type="submit">Filtrovat</button>
                 <a class="btn btn-secondary" href="/portal/apps">Reset</a>
@@ -225,6 +265,7 @@ def apps_page(
                 <tr>
                     <th>Služba</th>
                     <th>Status</th>
+                    <th>Zdraví</th>
                     <th>Dokumentace</th>
                     <th>Prostředky</th>
                     <th>Git</th>
@@ -245,6 +286,13 @@ def app_detail(app_id: str, request: Request, user=Depends(require_user)):
     if not app:
         raise HTTPException(status_code=404, detail="App not found")
 
+    log_audit_event(
+        user,
+        action="service.health.view",
+        target_type="service",
+        target_id=app_id,
+    )
+
     escaped_app_id = html.escape(app.get("id", ""))
     app_url_id = quote(app.get("id", ""), safe="")
     name = html.escape(app.get("name", "") or "")
@@ -254,6 +302,8 @@ def app_detail(app_id: str, request: Request, user=Depends(require_user)):
     memory = html.escape(app.get("memory", "") or "")
     cpus = html.escape(app.get("cpus", "") or "")
     updated_at = html.escape(app.get("updated_at", "") or "")
+    current_health = get_service_health(app.get("id", ""))
+    health_history = get_service_health_history(app.get("id", ""), limit=50)
 
     rows = ""
     for deployment in get_app_deployments(app.get("id", ""), limit=10):
@@ -294,8 +344,29 @@ def app_detail(app_id: str, request: Request, user=Depends(require_user)):
     if not job_rows:
         job_rows = '<tr><td colspan="4">Zatím nejsou evidované žádné úlohy této služby.</td></tr>'
 
+    health_status = render_health_status(current_health.get("status") if current_health else None)
+    health_http_status = html.escape(str(current_health.get("http_status") or "")) if current_health else ""
+    health_response_time = html.escape(str(current_health.get("response_time_ms") or "")) if current_health else ""
+    health_checked_at = html.escape(current_health.get("checked_at", "") or "") if current_health else ""
+    health_rows = ""
+    for item in health_history:
+        error_text = item.get("error_text", "") or ""
+        error_preview = error_text if len(error_text) <= 140 else f"{error_text[:137]}..."
+        health_rows += f"""
+        <tr>
+            <td>{html.escape(item.get("checked_at", "") or "")}</td>
+            <td>{render_health_status(item.get("status"))}</td>
+            <td>{html.escape(str(item.get("http_status") or ""))}</td>
+            <td>{html.escape(str(item.get("response_time_ms") or ""))}</td>
+            <td>{html.escape(error_preview)}</td>
+        </tr>
+        """
+
+    if not health_rows:
+        health_rows = '<tr><td colspan="5">Zatím nejsou evidované žádné kontroly zdraví.</td></tr>'
+
     return page(
-        f"Služba {escaped_app_id}",
+        name or escaped_app_id,
         f"""
         <div class="card">
             <h2>{escaped_app_id}</h2>
@@ -325,6 +396,30 @@ def app_detail(app_id: str, request: Request, user=Depends(require_user)):
             </table>
         </div>
 
+        <div class="card">
+            <h2>Zdraví služby</h2>
+            <table>
+                <tr><th>Aktuální status</th><td>{health_status}</td></tr>
+                <tr><th>HTTP status</th><td>{health_http_status}</td></tr>
+                <tr><th>Odezva</th><td>{health_response_time}</td></tr>
+                <tr><th>Poslední kontrola</th><td>{health_checked_at}</td></tr>
+            </table>
+        </div>
+
+        <div class="card">
+            <h2>Historie kontrol</h2>
+            <table>
+                <tr>
+                    <th>Čas</th>
+                    <th>Status</th>
+                    <th>HTTP status</th>
+                    <th>Odezva</th>
+                    <th>Chyba</th>
+                </tr>
+                {health_rows}
+            </table>
+        </div>
+
         <div class="card">
             <h2>Historie nasazení</h2>
             <table>

app/routes/operations.py

@@ -110,6 +110,46 @@ def render_recent_audit_events(events: list[dict]) -> str:
     return rows
 
 
+def render_health_status(status: str | None) -> str:
+    value = status or ""
+    normalized = value.lower()
+    labels = {
+        "healthy": "zdravá",
+        "unhealthy": "nezdravá",
+        "unreachable": "nedostupná",
+    }
+    class_name = "pill pill-muted"
+    if normalized == "healthy":
+        class_name = "pill pill-success"
+    elif normalized == "unhealthy":
+        class_name = "pill pill-warning"
+    elif normalized == "unreachable":
+        class_name = "pill pill-danger"
+    return f'<span class="{class_name}">{html.escape(labels.get(normalized, value))}</span>'
+
+
+def render_health_problems(items: list[dict]) -> str:
+    rows = ""
+    for item in items:
+        service_id = html.escape(item.get("service_id", "") or "")
+        service_name = html.escape(item.get("service_name") or item.get("service_id", "") or "")
+        error_text = item.get("error_text", "") or ""
+        error_preview = error_text if len(error_text) <= 120 else f"{error_text[:117]}..."
+        rows += f"""
+        <tr>
+            <td><a href="/portal/apps/{quote(service_id, safe='')}">{service_name}</a></td>
+            <td>{render_health_status(item.get("status"))}</td>
+            <td>{html.escape(item.get("checked_at", "") or "")}</td>
+            <td>{html.escape(str(item.get("response_time_ms") or ""))}</td>
+            <td>{html.escape(error_preview)}</td>
+        </tr>
+        """
+
+    if not rows:
+        rows = '<tr><td colspan="5">Žádné služby nevyžadují pozornost.</td></tr>'
+    return rows
+
+
 @router.get("/operations", response_class=HTMLResponse)
 def operations_dashboard(request: Request, user=Depends(require_user)):
     snapshot = get_operations_snapshot()
@@ -145,12 +185,29 @@ def operations_dashboard(request: Request, user=Depends(require_user)):
         <div class="stats-grid">
             <div class="stat-card stat-success"><span>Běžící služby</span><strong data-live-count="apps.active">{snapshot["apps"]["active"]}</strong></div>
             <div class="stat-card stat-danger"><span>Problémové služby</span><strong data-live-count="apps.problematic">{snapshot["apps"]["problematic"]}</strong></div>
+            <div class="stat-card stat-success"><span>Zdravé služby</span><strong data-live-count="health.healthy">{snapshot["health"]["healthy"]}</strong></div>
+            <div class="stat-card stat-warning"><span>Nezdravé služby</span><strong data-live-count="health.unhealthy">{snapshot["health"]["unhealthy"]}</strong></div>
+            <div class="stat-card stat-danger"><span>Nedostupné služby</span><strong data-live-count="health.unreachable">{snapshot["health"]["unreachable"]}</strong></div>
             <div class="stat-card stat-warning"><span>Aktivní úlohy</span><strong data-live-count="jobs.running">{snapshot["jobs"]["running"]}</strong></div>
             <div class="stat-card stat-total"><span>Nasazení dnes</span><strong data-live-count="deployments.today">{snapshot["deployments"]["today"]}</strong></div>
             <div class="stat-card stat-success"><span>Online workery</span><strong data-live-count="workers.online">{snapshot["workers"]["online"]}</strong></div>
             <div class="stat-card stat-danger"><span>Selhané úlohy (24 h)</span><strong data-live-count="jobs.failed_24h">{snapshot["jobs"]["failed_24h"]}</strong></div>
         </div>
 
+        <div class="card">
+            <h2>Služby vyžadující pozornost</h2>
+            <table>
+                <tr>
+                    <th>Služba</th>
+                    <th>Status</th>
+                    <th>Poslední kontrola</th>
+                    <th>Odezva</th>
+                    <th>Chyba</th>
+                </tr>
+                <tbody data-live-table="health.problems">{render_health_problems(snapshot["health"]["problems"])}</tbody>
+            </table>
+        </div>
+
         <div class="card">
             <h2>Poslední incidenty</h2>
             <table>

app/static/operations-live.js

@@ -44,6 +44,21 @@
         ? '<span class="pill pill-success">online</span>'
         : '<span class="pill pill-danger">offline</span>';
 
+    const healthBadge = (status) => {
+        const value = String(status || "");
+        const normalized = value.toLowerCase();
+        const labels = {
+            healthy: "zdravá",
+            unhealthy: "nezdravá",
+            unreachable: "nedostupná",
+        };
+        let className = "pill pill-muted";
+        if (normalized === "healthy") className = "pill pill-success";
+        if (normalized === "unhealthy") className = "pill pill-warning";
+        if (normalized === "unreachable") className = "pill pill-danger";
+        return `<span class="${className}">${escapeHtml(labels[normalized] || value)}</span>`;
+    };
+
     const setText = (selector, value) => {
         document.querySelectorAll(selector).forEach((el) => {
             el.textContent = value ?? 0;
@@ -112,6 +127,24 @@
         return rows || '<tr><td colspan="6">Zatím nejsou evidované žádné auditní události.</td></tr>';
     };
 
+    const renderHealthProblems = (items) => {
+        const rows = (items || []).map((item) => {
+            const errorText = String(item.error_text || "");
+            const errorPreview = errorText.length > 120 ? `${errorText.slice(0, 117)}...` : errorText;
+            const serviceName = item.service_name || item.service_id || "";
+            return `
+                <tr>
+                    <td><a href="/portal/apps/${encodeURIComponent(item.service_id || "")}">${escapeHtml(serviceName)}</a></td>
+                    <td>${healthBadge(item.status)}</td>
+                    <td>${escapeHtml(item.checked_at)}</td>
+                    <td>${escapeHtml(item.response_time_ms ?? "")}</td>
+                    <td>${escapeHtml(errorPreview)}</td>
+                </tr>
+            `;
+        }).join("");
+        return rows || '<tr><td colspan="5">Žádné služby nevyžadují pozornost.</td></tr>';
+    };
+
     const renderSnapshot = (snapshot) => {
         setText("[data-live-count='jobs.queued']", snapshot.jobs?.queued);
         setText("[data-live-count='jobs.running']", snapshot.jobs?.running);
@@ -119,6 +152,9 @@
         setText("[data-live-count='jobs.success_24h']", snapshot.jobs?.success_24h);
         setText("[data-live-count='apps.active']", snapshot.apps?.active);
         setText("[data-live-count='apps.problematic']", snapshot.apps?.problematic);
+        setText("[data-live-count='health.healthy']", snapshot.health?.healthy);
+        setText("[data-live-count='health.unhealthy']", snapshot.health?.unhealthy);
+        setText("[data-live-count='health.unreachable']", snapshot.health?.unreachable);
         setText("[data-live-count='workers.online']", snapshot.workers?.online);
         setText("[data-live-count='workers.offline']", snapshot.workers?.offline);
         setText("[data-live-count='workers.total']", (snapshot.workers?.online || 0) + (snapshot.workers?.offline || 0));
@@ -133,6 +169,7 @@
         document.querySelectorAll("[data-live-table='deployments.recent']").forEach((el) => { el.innerHTML = renderDeployments(snapshot.deployments?.recent); });
         document.querySelectorAll("[data-live-table='deployments.failed']").forEach((el) => { el.innerHTML = renderDeployments(snapshot.deployments?.recent_failed); });
         document.querySelectorAll("[data-live-table='audit.recent']").forEach((el) => { el.innerHTML = renderAudit(snapshot.audit?.recent); });
+        document.querySelectorAll("[data-live-table='health.problems']").forEach((el) => { el.innerHTML = renderHealthProblems(snapshot.health?.problems); });
 
         (snapshot.workers?.recent || []).forEach((worker) => {
             const row = Array.from(document.querySelectorAll("tr[data-worker-id]"))

app/templates/layout.py

@@ -46,7 +46,7 @@ def page(title: str, body: str, user=None) -> str:
     return f"""
     <html>
     <head>
-        <title>{html.escape(title)} | {APP_NAME}</title>
+        <title>{APP_NAME} - {html.escape(title)}</title>
         <link rel="stylesheet" href="{PORTAL_PREFIX}/static/styles.css">
         <script src="{PORTAL_PREFIX}/static/portal.js"></script>
     </head>