feat: add system health diagnostics

backend/app/system_health/models.py

@@ -0,0 +1,124 @@
+"""Typed contracts for the System Health overview."""
+
+from datetime import date, datetime
+from typing import Literal
+
+from pydantic import BaseModel, Field
+
+
+Availability = Literal["available", "unavailable"]
+MigrationState = Literal["current", "mismatch", "unavailable"]
+
+
+class ApplicationDiagnostics(BaseModel):
+    status: Literal["available"] = "available"
+    environment: str
+    build_commit: str | None = None
+    build_version: str | None = None
+    server_time: datetime
+    expected_migration_head: str | None = None
+
+
+class MigrationDiagnostics(BaseModel):
+    state: MigrationState
+    current_revision: str | None = None
+    expected_revision: str | None = None
+
+
+class TableCount(BaseModel):
+    name: str
+    row_count: int
+
+
+class TableSize(BaseModel):
+    name: str
+    total_bytes: int
+
+
+class TelemetryDiagnostics(BaseModel):
+    snapshot_count: int
+    oldest_observed_at: datetime | None = None
+    latest_observed_at: datetime | None = None
+
+
+class FleetIntegrityDiagnostics(BaseModel):
+    total_miners: int
+    miners_without_asset_id: int
+    unmapped_miners: int
+    miners_in_multiple_positions: int
+    duplicate_current_mac_groups: int
+
+
+class LocationIntegrityDiagnostics(BaseModel):
+    sites: int
+    rooms: int
+    positions: int
+    occupied_positions: int
+    empty_positions: int
+    unscoped_legacy_positions: int
+    expected_mac_mismatches: int
+
+
+class RepairIntegrityDiagnostics(BaseModel):
+    active_modular_repairs: int
+    duplicate_active_repair_groups: int
+    legacy_repair_only_miners: int
+    lifecycle_repair_without_active_repair: int
+    active_repair_lifecycle_mismatches: int
+    waiting_placement_records: int
+    repair_events: int
+    repair_parts: int
+
+
+class RollupDiagnostics(BaseModel):
+    table_present: bool
+    row_count: int | None = None
+    latest_bucket: datetime | date | None = None
+    population: Literal["manual"] = "manual"
+
+
+class RetentionDiagnostics(BaseModel):
+    raw_retention_enabled: Literal[False] = False
+    raw_retention_mode: Literal["not_enabled"] = "not_enabled"
+    hourly: RollupDiagnostics
+    daily: RollupDiagnostics
+    latest_recorded_run_at: datetime | None = None
+    latest_recorded_run_mode: str | None = None
+    completeness: Literal["not_guaranteed"] = "not_guaranteed"
+
+
+class PostgreSQLDiagnostics(BaseModel):
+    status: Availability
+    migration: MigrationDiagnostics
+    table_counts: list[TableCount] = Field(default_factory=list)
+    largest_tables: list[TableSize] = Field(default_factory=list)
+    telemetry: TelemetryDiagnostics | None = None
+    fleet: FleetIntegrityDiagnostics | None = None
+    locations: LocationIntegrityDiagnostics | None = None
+    repairs: RepairIntegrityDiagnostics | None = None
+    retention: RetentionDiagnostics | None = None
+
+
+class RedisDiagnostics(BaseModel):
+    status: Availability
+
+
+class SystemHealthOverviewData(BaseModel):
+    application: ApplicationDiagnostics
+    postgresql: PostgreSQLDiagnostics
+    redis: RedisDiagnostics
+
+
+class SystemHealthOverviewMeta(BaseModel):
+    generated_at: datetime
+    partial: bool
+    unavailable_sections: list[str] = Field(default_factory=list)
+    frontend_request_count: Literal[1] = 1
+    postgresql_statement_count: Literal[3] = 3
+    redis_call_count: Literal[1] = 1
+
+
+class SystemHealthOverviewResponse(BaseModel):
+    data: SystemHealthOverviewData
+    meta: SystemHealthOverviewMeta
+    error: None = None