feat(monitors): unify ping/dns/http into one Monitor entity + custom targets

Collapse the three former check types into a single core `Monitor` entity
with one management surface (/monitors), one result table (monitor_results),
and a single scheduled task. Every type can now watch a free-standing custom
destination (optional host_id) — not just a registered Host.

- models: Monitor + MonitorResult replace PingResult/DnsResult; Host loses its
  ping/dns facet columns (now Monitor rows linked by host_id).
- checks: monitors/{ping,dns,http}.py pure probes + runner.run_monitor
  dispatcher; one monitor_check scheduler with a per-monitor due-filter.
- status: single monitor_status_source replaces the three sources.
- UI: /monitors blueprint (type-aware add/edit/list/widget); host hub shows a
  host's linked monitors + "add monitor for this host"; nav + widget registry
  + alert metric catalog rewired. http plugin folded into core and removed.
- migration 0022 merges the http branch, data-migrates host facets +
  http_monitors + all three result histories, drops the old tables/columns.

Resolves the per-host ping/dns auto-attach issue (#275): monitors are now
explicit, never auto-added to every host.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016Jg27rgypiW2efULXJDtMC

steward/migrations/versions/0022_unify_monitors.py

@@ -0,0 +1,253 @@
+"""Unify ping/dns/http into one Monitor entity + monitor_results
+
+Merges the standalone "http" branch (http_001_initial) back into the core
+line and collapses the three former check models into `monitors` +
+`monitor_results`:
+
+  * Host ping/dns facets (hosts.ping_enabled/dns_enabled/probe_type/... ) become
+    Monitor rows linked via host_id.
+  * http_monitors rows become hostless Monitor rows (type=http).
+  * ping_results / dns_results / http_results history is copied into
+    monitor_results, then the old tables + host columns are dropped.
+
+The http_monitors/http_results tables have no ORM models anymore (the http
+plugin was folded into core), so they're read/dropped via raw SQL.
+
+Revision ID: 0022_unify_monitors
+Revises: 0021_dashboard_widget_grid, http_001_initial
+Create Date: 2026-06-18
+"""
+import json
+import uuid
+from typing import Sequence, Union
+
+from alembic import op
+import sqlalchemy as sa
+
+revision: str = "0022_unify_monitors"
+down_revision: Union[str, Sequence[str], None] = (
+    "0021_dashboard_widget_grid", "http_001_initial",
+)
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def _new_id() -> str:
+    return str(uuid.uuid4())
+
+
+def upgrade() -> None:
+    conn = op.get_bind()
+
+    # ── 1. New tables ─────────────────────────────────────────────────────────
+    op.create_table(
+        "monitors",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("name", sa.String(128), nullable=False),
+        sa.Column("type", sa.String(16), nullable=False),
+        sa.Column("target", sa.String(2048), nullable=False),
+        sa.Column("host_id", sa.String(36),
+                  sa.ForeignKey("hosts.id", ondelete="CASCADE"), nullable=True),
+        sa.Column("config_json", sa.Text, nullable=False, server_default="{}"),
+        sa.Column("enabled", sa.Boolean, nullable=False, server_default="true"),
+        sa.Column("check_interval_seconds", sa.Integer, nullable=False, server_default="0"),
+        sa.Column("last_checked_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+        # rule 36: new monitor types ship via DROP/ADD of this CHECK constraint.
+        sa.CheckConstraint("type IN ('icmp', 'tcp', 'dns', 'http')", name="ck_monitors_type"),
+    )
+    op.create_index("ix_monitors_host_id", "monitors", ["host_id"])
+
+    op.create_table(
+        "monitor_results",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("monitor_id", sa.String(36),
+                  sa.ForeignKey("monitors.id", ondelete="CASCADE"), nullable=False),
+        sa.Column("checked_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("is_up", sa.Boolean, nullable=False, server_default="false"),
+        sa.Column("response_ms", sa.Float, nullable=True),
+        sa.Column("status_code", sa.Integer, nullable=True),
+        sa.Column("resolved_ip", sa.String(255), nullable=True),
+        sa.Column("content_matched", sa.Boolean, nullable=True),
+        sa.Column("tls_expires_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("error_msg", sa.String(512), nullable=True),
+    )
+    op.create_index("ix_monitor_results_monitor_id", "monitor_results", ["monitor_id"])
+    op.create_index("ix_monitor_results_checked_at", "monitor_results", ["checked_at"])
+    op.create_index("ix_monitor_results_monitor_checked", "monitor_results",
+                    ["monitor_id", "checked_at"])
+
+    ins_monitor = sa.text(
+        "INSERT INTO monitors (id, name, type, target, host_id, config_json, "
+        "enabled, check_interval_seconds, last_checked_at, created_at) VALUES "
+        "(:id, :name, :type, :target, :host_id, :cfg, :enabled, :interval, "
+        ":last, :created)"
+    )
+
+    # ── 2. Host ping/dns facets → Monitor rows ────────────────────────────────
+    ping_map: dict[str, str] = {}   # host_id -> monitor_id
+    dns_map: dict[str, str] = {}
+    hosts = conn.execute(sa.text(
+        "SELECT id, name, address, probe_type, probe_port, ping_enabled, "
+        "dns_enabled, dns_expected_ip, created_at FROM hosts"
+    )).mappings().all()
+    for h in hosts:
+        if h["ping_enabled"]:
+            mid = _new_id()
+            mtype = "icmp" if str(h["probe_type"]) == "icmp" else "tcp"
+            cfg = {} if mtype == "icmp" else {"port": h["probe_port"] or 80}
+            conn.execute(ins_monitor, {
+                "id": mid, "name": h["name"], "type": mtype, "target": h["address"],
+                "host_id": h["id"], "cfg": json.dumps(cfg), "enabled": True,
+                "interval": 0, "last": None, "created": h["created_at"],
+            })
+            ping_map[h["id"]] = mid
+        if h["dns_enabled"]:
+            mid = _new_id()
+            cfg = {"expected_ip": h["dns_expected_ip"]}
+            conn.execute(ins_monitor, {
+                "id": mid, "name": f"{h['name']} (DNS)", "type": "dns",
+                "target": h["address"], "host_id": h["id"], "cfg": json.dumps(cfg),
+                "enabled": True, "interval": 0, "last": None, "created": h["created_at"],
+            })
+            dns_map[h["id"]] = mid
+
+    # ── 3. http_monitors rows → hostless Monitor rows (raw SQL: no ORM) ───────
+    http_map: dict[str, str] = {}   # old http monitor id -> new monitor id
+    http_rows = conn.execute(sa.text(
+        "SELECT id, name, url, method, expected_status, content_match, "
+        "headers_json, timeout_seconds, check_interval_seconds, follow_redirects, "
+        "verify_ssl, enabled, last_checked_at, created_at FROM http_monitors"
+    )).mappings().all()
+    for m in http_rows:
+        mid = _new_id()
+        try:
+            headers = json.loads(m["headers_json"] or "{}")
+        except (ValueError, TypeError):
+            headers = {}
+        cfg = {
+            "method": m["method"], "expected_status": m["expected_status"],
+            "content_match": m["content_match"], "headers": headers,
+            "timeout_seconds": m["timeout_seconds"],
+            "follow_redirects": bool(m["follow_redirects"]),
+            "verify_ssl": bool(m["verify_ssl"]),
+        }
+        conn.execute(ins_monitor, {
+            "id": mid, "name": m["name"], "type": "http", "target": m["url"],
+            "host_id": None, "cfg": json.dumps(cfg), "enabled": bool(m["enabled"]),
+            "interval": m["check_interval_seconds"] or 0,
+            "last": m["last_checked_at"], "created": m["created_at"],
+        })
+        http_map[m["id"]] = mid
+
+    # ── 4. Result history → monitor_results ───────────────────────────────────
+    for host_id, mid in ping_map.items():
+        conn.execute(sa.text(
+            "INSERT INTO monitor_results (id, monitor_id, checked_at, is_up, response_ms) "
+            "SELECT gen_random_uuid()::text, :mid, probed_at, (status = 'up'), response_time_ms "
+            "FROM ping_results WHERE host_id = :hid"
+        ), {"mid": mid, "hid": host_id})
+    for host_id, mid in dns_map.items():
+        conn.execute(sa.text(
+            "INSERT INTO monitor_results (id, monitor_id, checked_at, is_up, resolved_ip) "
+            "SELECT gen_random_uuid()::text, :mid, resolved_at, (status = 'resolved'), resolved_ip "
+            "FROM dns_results WHERE host_id = :hid"
+        ), {"mid": mid, "hid": host_id})
+    for old_id, mid in http_map.items():
+        conn.execute(sa.text(
+            "INSERT INTO monitor_results (id, monitor_id, checked_at, is_up, response_ms, "
+            "status_code, content_matched, tls_expires_at, error_msg) "
+            "SELECT gen_random_uuid()::text, :mid, checked_at, is_up, response_ms, "
+            "status_code, content_matched, tls_expires_at, error_msg "
+            "FROM http_results WHERE monitor_id = :oid"
+        ), {"mid": mid, "oid": old_id})
+
+    # ── 4b. Repoint dashboard widgets (ping/dns/http_monitors → monitors) ─────
+    # The ping/dns/http_monitors widget keys are retired; one `monitors` widget
+    # replaces them. Convert ping in place, drop the now-redundant others.
+    conn.execute(sa.text(
+        "UPDATE dashboard_widgets SET widget_key = 'monitors' WHERE widget_key = 'ping'"
+    ))
+    conn.execute(sa.text(
+        "DELETE FROM dashboard_widgets WHERE widget_key IN ('dns', 'http_monitors')"
+    ))
+
+    # ── 5. Drop the old tables, host columns, and orphaned enum types ─────────
+    op.drop_table("ping_results")
+    op.drop_table("dns_results")
+    op.drop_table("http_results")
+    op.drop_table("http_monitors")
+    op.drop_column("hosts", "ping_enabled")
+    op.drop_column("hosts", "dns_enabled")
+    op.drop_column("hosts", "dns_expected_ip")
+    op.drop_column("hosts", "probe_type")
+    op.drop_column("hosts", "probe_port")
+    op.drop_column("hosts", "poll_interval_seconds")
+    op.execute("DROP TYPE IF EXISTS pingstatus")
+    op.execute("DROP TYPE IF EXISTS dnsstatus")
+    op.execute("DROP TYPE IF EXISTS probetype")
+
+
+def downgrade() -> None:
+    # Dev-only project (no installs to protect): downgrade restores the table
+    # shells so the schema is walkable, but does NOT recover migrated history.
+    op.add_column("hosts", sa.Column("poll_interval_seconds", sa.Integer, nullable=True))
+    op.add_column("hosts", sa.Column("probe_port", sa.Integer, nullable=False, server_default="80"))
+    op.add_column("hosts", sa.Column("probe_type",
+                  sa.Enum("tcp", "icmp", name="probetype"), nullable=False, server_default="tcp"))
+    op.add_column("hosts", sa.Column("dns_expected_ip", sa.String(255), nullable=True))
+    op.add_column("hosts", sa.Column("dns_enabled", sa.Boolean, nullable=False, server_default="false"))
+    op.add_column("hosts", sa.Column("ping_enabled", sa.Boolean, nullable=False, server_default="true"))
+
+    op.create_table(
+        "ping_results",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("host_id", sa.String(36), sa.ForeignKey("hosts.id", ondelete="CASCADE"), nullable=False),
+        sa.Column("probed_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("status", sa.Enum("up", "down", name="pingstatus"), nullable=False),
+        sa.Column("response_time_ms", sa.Float, nullable=True),
+    )
+    op.create_table(
+        "dns_results",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("host_id", sa.String(36), sa.ForeignKey("hosts.id", ondelete="CASCADE"), nullable=False),
+        sa.Column("resolved_at", sa.DateTime(timezone=True), nullable=False),
+        sa.Column("status", sa.Enum("resolved", "failed", name="dnsstatus"), nullable=False),
+        sa.Column("resolved_ip", sa.String(255), nullable=True),
+    )
+    op.create_table(
+        "http_monitors",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("name", sa.String(128), nullable=False),
+        sa.Column("url", sa.String(2048), nullable=False),
+        sa.Column("method", sa.String(8), nullable=False, server_default="GET"),
+        sa.Column("expected_status", sa.Integer, nullable=False, server_default="200"),
+        sa.Column("content_match", sa.String(512), nullable=False, server_default=""),
+        sa.Column("headers_json", sa.Text, nullable=False, server_default="{}"),
+        sa.Column("timeout_seconds", sa.Integer, nullable=False, server_default="10"),
+        sa.Column("check_interval_seconds", sa.Integer, nullable=False, server_default="0"),
+        sa.Column("follow_redirects", sa.Boolean, nullable=False, server_default="1"),
+        sa.Column("verify_ssl", sa.Boolean, nullable=False, server_default="1"),
+        sa.Column("enabled", sa.Boolean, nullable=False, server_default="1"),
+        sa.Column("last_checked_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+    )
+    op.create_table(
+        "http_results",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("monitor_id", sa.String(36), nullable=False, index=True),
+        sa.Column("checked_at", sa.DateTime(timezone=True), nullable=False, index=True),
+        sa.Column("status_code", sa.Integer, nullable=True),
+        sa.Column("response_ms", sa.Float, nullable=True),
+        sa.Column("is_up", sa.Boolean, nullable=False, server_default="0"),
+        sa.Column("content_matched", sa.Boolean, nullable=True),
+        sa.Column("error_msg", sa.String(512), nullable=True),
+        sa.Column("tls_expires_at", sa.DateTime(timezone=True), nullable=True),
+    )
+
+    op.drop_index("ix_monitor_results_monitor_checked", table_name="monitor_results")
+    op.drop_index("ix_monitor_results_checked_at", table_name="monitor_results")
+    op.drop_index("ix_monitor_results_monitor_id", table_name="monitor_results")
+    op.drop_table("monitor_results")
+    op.drop_index("ix_monitors_host_id", table_name="monitors")
+    op.drop_table("monitors")

steward/migrations/versions/http_001_initial.py

@@ -0,0 +1,59 @@
+"""HTTP monitoring tables (relocated from the former http plugin)
+
+This revision originally lived in plugins/http/migrations. The http plugin has
+been folded into core (unified Monitor entity), so its one migration is moved
+here UNCHANGED — same revision id + "http" branch label — purely so the
+revision stays resolvable in existing databases that already stamped it. The
+0022_unify_monitors merge revision then absorbs this branch and drops these
+tables after migrating their data into `monitor_results`.
+
+Revision ID: http_001_initial
+Revises: (none — branch off core via depends_on)
+Create Date: 2026-03-22
+"""
+from typing import Sequence, Union
+from alembic import op
+import sqlalchemy as sa
+
+revision: str = "http_001_initial"
+down_revision: Union[str, None] = None
+branch_labels: Union[str, Sequence[str], None] = "http"
+depends_on: Union[str, Sequence[str], None] = ("0004_app_settings",)
+
+
+def upgrade() -> None:
+    op.create_table(
+        "http_monitors",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("name", sa.String(128), nullable=False),
+        sa.Column("url", sa.String(2048), nullable=False),
+        sa.Column("method", sa.String(8), nullable=False, server_default="GET"),
+        sa.Column("expected_status", sa.Integer, nullable=False, server_default="200"),
+        sa.Column("content_match", sa.String(512), nullable=False, server_default=""),
+        sa.Column("headers_json", sa.Text, nullable=False, server_default="{}"),
+        sa.Column("timeout_seconds", sa.Integer, nullable=False, server_default="10"),
+        sa.Column("check_interval_seconds", sa.Integer, nullable=False, server_default="0"),
+        sa.Column("follow_redirects", sa.Boolean, nullable=False, server_default="1"),
+        sa.Column("verify_ssl", sa.Boolean, nullable=False, server_default="1"),
+        sa.Column("enabled", sa.Boolean, nullable=False, server_default="1"),
+        sa.Column("last_checked_at", sa.DateTime(timezone=True), nullable=True),
+        sa.Column("created_at", sa.DateTime(timezone=True), nullable=False),
+    )
+
+    op.create_table(
+        "http_results",
+        sa.Column("id", sa.String(36), primary_key=True),
+        sa.Column("monitor_id", sa.String(36), nullable=False, index=True),
+        sa.Column("checked_at", sa.DateTime(timezone=True), nullable=False, index=True),
+        sa.Column("status_code", sa.Integer, nullable=True),
+        sa.Column("response_ms", sa.Float, nullable=True),
+        sa.Column("is_up", sa.Boolean, nullable=False, server_default="0"),
+        sa.Column("content_matched", sa.Boolean, nullable=True),
+        sa.Column("error_msg", sa.String(512), nullable=True),
+        sa.Column("tls_expires_at", sa.DateTime(timezone=True), nullable=True),
+    )
+
+
+def downgrade() -> None:
+    op.drop_table("http_results")
+    op.drop_table("http_monitors")