Add SNMP monitor collection

worker/app/scheduler.py

@@ -7,11 +7,12 @@ from sqlalchemy.exc import SQLAlchemyError
 from sqlalchemy.orm import Session
 import httpx
 
+from app.collectors.snmp import SnmpCheckConfig, SnmpCheckResult, run_snmp_check
 from app.collectors.website import WebsiteCheckConfig, run_website_check
 from app.collectors.network import PingCheckConfig, TcpCheckConfig, run_ping_check, run_tcp_check
 from app.config import settings
 from app.db import session_scope
-from app.models import AlertRule, Asset, CheckResult, Incident, Monitor, NotificationChannel
+from app.models import AlertRule, Asset, CheckResult, Credential, Incident, Metric, Monitor, NotificationChannel
 from app.secrets import decrypt_secret
 
 logger = logging.getLogger(__name__)
@@ -47,7 +48,7 @@ class Scheduler:
     def _load_due_monitors(self, db: Session) -> list[Monitor]:
         now = datetime.now(UTC)
         monitors = db.scalars(
-            select(Monitor).where(Monitor.monitor_type.in_(["http", "ping", "tcp"])).order_by(Monitor.id).limit(50)
+            select(Monitor).where(Monitor.monitor_type.in_(["http", "ping", "tcp", "snmp"])).order_by(Monitor.id).limit(50)
         ).all()
         due: list[Monitor] = []
         for monitor in monitors:
@@ -60,7 +61,7 @@ class Scheduler:
         return due
 
     async def _run_monitor(self, db: Session, monitor: Monitor) -> None:
-        result = await self._collect_monitor_result(monitor)
+        result = await self._collect_monitor_result(db, monitor)
         now = datetime.now(UTC)
 
         monitor.status = result.status
@@ -76,6 +77,18 @@ class Scheduler:
         )
         db.flush()
 
+        for metric in getattr(result, "metrics", []):
+            db.add(
+                Metric(
+                    monitor_id=monitor.id,
+                    name=metric.name,
+                    value=metric.value,
+                    unit=metric.unit,
+                    observed_at=now,
+                )
+            )
+        db.flush()
+
         if monitor.asset_id is not None:
             asset = db.get(Asset, monitor.asset_id)
             if asset is not None:
@@ -87,7 +100,7 @@ class Scheduler:
 
         logger.info("Checked %s: %s (%s ms)", monitor.name, result.status, result.response_time_ms)
 
-    async def _collect_monitor_result(self, monitor: Monitor):
+    async def _collect_monitor_result(self, db: Session, monitor: Monitor):
         if monitor.monitor_type == "http":
             config = WebsiteCheckConfig(
                 url=monitor.target,
@@ -115,8 +128,36 @@ class Scheduler:
             )
             return await run_tcp_check(config)
 
+        if monitor.monitor_type == "snmp":
+            return await self._collect_snmp_monitor_result(db, monitor)
+
         raise ValueError(f"Unsupported monitor type: {monitor.monitor_type}")
 
+    async def _collect_snmp_monitor_result(self, db: Session, monitor: Monitor) -> SnmpCheckResult:
+        profile_id = monitor.config.get("credential_profile_id")
+        if not isinstance(profile_id, int):
+            return SnmpCheckResult(status="down", response_time_ms=None, message="SNMP credential profile is not configured")
+
+        profile = db.get(Credential, profile_id)
+        if profile is None or profile.credential_type != "snmp":
+            return SnmpCheckResult(status="down", response_time_ms=None, message="SNMP credential profile was not found")
+
+        community = decrypt_secret(profile.encrypted_secret)
+        if not community:
+            return SnmpCheckResult(status="down", response_time_ms=None, message="SNMP credential profile has no usable community string")
+
+        extra = dict(profile.extra or {})
+        config = SnmpCheckConfig(
+            host=monitor.target,
+            community=community,
+            item_id=str(monitor.config.get("item_id") or ""),
+            item_type=str(monitor.config.get("item_type") or ""),
+            port=int(extra.get("port") or 161),
+            timeout_seconds=float(extra.get("timeout_seconds") or 5),
+            retries=int(extra.get("retries") or 1),
+        )
+        return await run_snmp_check(config)
+
     async def _evaluate_rule(self, db: Session, monitor: Monitor, rule: AlertRule, now: datetime, message: str) -> None:
         open_incident = db.scalar(
             select(Incident).where(