@oneuptime/common 12.0.14 → 12.0.16

package/Tests/Server/Utils/Monitor/Criteria/ServerMonitorCriteria.test.ts

@@ -0,0 +1,1219 @@
+import ServerMonitorCriteria from "../../../../../Server/Utils/Monitor/Criteria/ServerMonitorCriteria";
+import EvaluateOverTime, {
+  OverTimeCriteriaValue,
+} from "../../../../../Server/Utils/Monitor/Criteria/EvaluateOverTime";
+import BasicInfrastructureMetrics, {
+  BasicDiskMetrics,
+  LoadMetrics,
+} from "../../../../../Types/Infrastructure/BasicMetrics";
+import {
+  CheckOn,
+  CriteriaFilter,
+  EvaluateOverTimeType,
+  FilterType,
+} from "../../../../../Types/Monitor/CriteriaFilter";
+import ServerMonitorResponse, {
+  ServerProcess,
+} from "../../../../../Types/Monitor/ServerMonitor/ServerMonitorResponse";
+import ObjectID from "../../../../../Types/ObjectID";
+import { afterEach, describe, expect, jest, test } from "@jest/globals";
+
+/*
+ * Fixed reference "now" so that the difference-in-minutes computation the
+ * evaluator performs (requestReceivedAt vs timeNow) is fully deterministic
+ * and never touches the wall clock.
+ */
+const BASE_TIME: Date = new Date("2026-08-20T12:00:00.000Z");
+
+/*
+ * Build a BasicInfrastructureMetrics payload. Every sub-metric has a benign
+ * default so an individual test only supplies the field it exercises.
+ */
+function buildMetrics(input: {
+  cpuPercentUsed?: number | undefined;
+  cpuIoWaitPercent?: number | undefined;
+  memoryPercentUsed?: number | undefined;
+  swapPercentUsed?: number | undefined;
+  diskMetrics?: Array<BasicDiskMetrics> | undefined;
+  loadMetrics?: LoadMetrics | undefined;
+}): BasicInfrastructureMetrics {
+  return {
+    cpuMetrics: {
+      percentUsed: input.cpuPercentUsed ?? 0,
+      cores: 4,
+      timeIoWaitPercent: input.cpuIoWaitPercent,
+    },
+    memoryMetrics: {
+      total: 100,
+      free: 50,
+      used: 50,
+      percentUsed: input.memoryPercentUsed ?? 0,
+      percentFree: 100 - (input.memoryPercentUsed ?? 0),
+      swapPercentUsed: input.swapPercentUsed,
+    },
+    diskMetrics: input.diskMetrics ?? [],
+    loadMetrics: input.loadMetrics,
+  };
+}
+
+/*
+ * Build a single disk metric entry. percentUsed / percentFree are optional so
+ * the evaluator's `percentUsed ?? percentFree ?? 0` fallback chain can be
+ * exercised; the cast tolerates the deliberately-absent required field.
+ */
+function buildDiskMetric(input: {
+  diskPath: string;
+  percentUsed?: number | undefined;
+  percentFree?: number | undefined;
+}): BasicDiskMetrics {
+  return {
+    total: 100,
+    free: 40,
+    used: 60,
+    diskPath: input.diskPath,
+    percentUsed: input.percentUsed as number,
+    percentFree: input.percentFree as number,
+  };
+}
+
+function buildProcess(input: {
+  pid?: number | undefined;
+  name?: string | undefined;
+  command?: string | undefined;
+}): ServerProcess {
+  return {
+    pid: input.pid ?? 1,
+    name: input.name ?? "proc",
+    command: input.command ?? "/bin/proc",
+  };
+}
+
+/*
+ * Build a ServerMonitorResponse. `minutesSinceLastCheck` positions
+ * requestReceivedAt that many minutes before the fixed timeNow, which drives
+ * the online/offline decision for IsOnline checks.
+ */
+function buildServerResponse(input: {
+  metrics?: BasicInfrastructureMetrics | undefined;
+  processes?: Array<ServerProcess> | undefined;
+  onlyCheckRequestReceivedAt?: boolean | undefined;
+  minutesSinceLastCheck?: number | undefined;
+}): ServerMonitorResponse {
+  const minutesAgo: number = input.minutesSinceLastCheck ?? 0;
+
+  return {
+    projectId: ObjectID.generate(),
+    monitorId: ObjectID.generate(),
+    hostname: "test-host",
+    requestReceivedAt: new Date(BASE_TIME.getTime() - minutesAgo * 60000),
+    timeNow: BASE_TIME,
+    onlyCheckRequestReceivedAt: input.onlyCheckRequestReceivedAt ?? false,
+    basicInfrastructureMetrics: input.metrics,
+    processes: input.processes,
+  };
+}
+
+async function evaluate(
+  dataToProcess: ServerMonitorResponse,
+  criteriaFilter: CriteriaFilter,
+): Promise<string | null> {
+  return ServerMonitorCriteria.isMonitorInstanceCriteriaFilterMet({
+    dataToProcess,
+    criteriaFilter,
+  });
+}
+
+describe("ServerMonitorCriteria.isMonitorInstanceCriteriaFilterMet", () => {
+  afterEach(() => {
+    jest.restoreAllMocks();
+  });
+
+  test("an unrelated CheckOn is not claimed by this evaluator", async () => {
+    const result: string | null = await evaluate(buildServerResponse({}), {
+      checkOn: CheckOn.ResponseTime,
+      filterType: FilterType.LessThan,
+      value: 1000,
+    });
+
+    expect(result).toBeNull();
+  });
+
+  describe("IsOnline", () => {
+    /*
+     * A recent check (difference < 3 minute default) is treated as online.
+     */
+    test("recent check + True → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 1 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.True,
+          value: undefined,
+        },
+      );
+
+      expect(result).toContain(CheckOn.IsOnline);
+      expect(result).toContain("true");
+    });
+
+    test("recent check + False → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 1 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.False,
+          value: undefined,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * A stale check (difference >= 3 minute default) is treated as offline.
+     */
+    test("stale check + False → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 5 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.False,
+          value: undefined,
+        },
+      );
+
+      expect(result).toContain(CheckOn.IsOnline);
+      expect(result).toContain("false");
+    });
+
+    test("stale check + True → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 5 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.True,
+          value: undefined,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * The offline boundary is inclusive: a difference of exactly the 3 minute
+     * threshold takes the offline branch.
+     */
+    test("difference exactly at the 3 minute threshold → offline branch", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 3 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.False,
+          value: undefined,
+        },
+      );
+
+      expect(result).toContain("false");
+    });
+
+    test("difference just below the 3 minute threshold → online branch", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 2 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.True,
+          value: undefined,
+        },
+      );
+
+      expect(result).toContain("true");
+    });
+
+    /*
+     * The IsOnline branches deliberately ignore onlyCheckRequestReceivedAt —
+     * liveness must still be evaluated on a heartbeat-only payload.
+     */
+    test("evaluates even when onlyCheckRequestReceivedAt is set", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          minutesSinceLastCheck: 1,
+          onlyCheckRequestReceivedAt: true,
+        }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.True,
+          value: undefined,
+        },
+      );
+
+      expect(result).toContain(CheckOn.IsOnline);
+    });
+
+    /*
+     * A filter type the boolean comparator does not handle (a numeric
+     * operator) is undecidable for IsOnline.
+     */
+    test("unhandled filter type → undecidable", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 1 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.GreaterThan,
+          value: 1,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+
+  describe("CPUUsagePercent", () => {
+    test("usage above threshold + GreaterThan → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 80 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+        },
+      );
+
+      expect(result).toContain(CheckOn.CPUUsagePercent);
+      expect(result).toContain("greater than");
+      expect(result).toContain("80");
+    });
+
+    test("usage below threshold + GreaterThan → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 30 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    test("numeric string threshold is accepted", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 80 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: "50",
+        },
+      );
+
+      expect(result).toContain(CheckOn.CPUUsagePercent);
+    });
+
+    test("non-numeric threshold → undecidable", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 80 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: "not-a-number",
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * With no infrastructure metrics present the CPU value falls back to 0,
+     * so a LessThan-1 check is met while a GreaterThan-1 check is not.
+     */
+    test("missing metrics fall back to zero usage", async () => {
+      const result: string | null = await evaluate(buildServerResponse({}), {
+        checkOn: CheckOn.CPUUsagePercent,
+        filterType: FilterType.LessThan,
+        value: 1,
+      });
+
+      expect(result).toContain(CheckOn.CPUUsagePercent);
+    });
+
+    /*
+     * onlyCheckRequestReceivedAt short-circuits every infrastructure metric
+     * branch — a heartbeat payload carries no fresh metric to evaluate.
+     */
+    test("onlyCheckRequestReceivedAt short-circuits the CPU branch", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({ cpuPercentUsed: 90 }),
+          onlyCheckRequestReceivedAt: true,
+        }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+
+  describe("MemoryUsagePercent", () => {
+    test("usage below threshold + LessThan → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({ memoryPercentUsed: 40 }),
+        }),
+        {
+          checkOn: CheckOn.MemoryUsagePercent,
+          filterType: FilterType.LessThan,
+          value: 95,
+        },
+      );
+
+      expect(result).toContain(CheckOn.MemoryUsagePercent);
+    });
+
+    test("usage above threshold + LessThan → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({ memoryPercentUsed: 96 }),
+        }),
+        {
+          checkOn: CheckOn.MemoryUsagePercent,
+          filterType: FilterType.LessThan,
+          value: 95,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    test("usage equal to threshold + GreaterThanOrEqualTo → met (boundary)", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({ memoryPercentUsed: 90 }),
+        }),
+        {
+          checkOn: CheckOn.MemoryUsagePercent,
+          filterType: FilterType.GreaterThanOrEqualTo,
+          value: 90,
+        },
+      );
+
+      expect(result).toContain(CheckOn.MemoryUsagePercent);
+    });
+  });
+
+  describe("SwapUsagePercent", () => {
+    test("swap usage above threshold + GreaterThan → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ swapPercentUsed: 30 }) }),
+        {
+          checkOn: CheckOn.SwapUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 20,
+        },
+      );
+
+      expect(result).toContain(CheckOn.SwapUsagePercent);
+    });
+
+    test("absent swap metric falls back to zero", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({}) }),
+        {
+          checkOn: CheckOn.SwapUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 20,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+
+  describe("CPUIoWaitPercent", () => {
+    test("io wait above threshold + GreaterThan → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({ cpuIoWaitPercent: 15 }),
+        }),
+        {
+          checkOn: CheckOn.CPUIoWaitPercent,
+          filterType: FilterType.GreaterThan,
+          value: 10,
+        },
+      );
+
+      expect(result).toContain(CheckOn.CPUIoWaitPercent);
+    });
+
+    test("absent io wait metric falls back to zero", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({}) }),
+        {
+          checkOn: CheckOn.CPUIoWaitPercent,
+          filterType: FilterType.GreaterThan,
+          value: 10,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+
+  describe("LoadAverage", () => {
+    const load: LoadMetrics = { load1: 2, load5: 1.5, load15: 0.5 };
+
+    test("LoadAverage1Min reads the 1 minute figure", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ loadMetrics: load }) }),
+        {
+          checkOn: CheckOn.LoadAverage1Min,
+          filterType: FilterType.GreaterThan,
+          value: 1,
+        },
+      );
+
+      expect(result).toContain(CheckOn.LoadAverage1Min);
+    });
+
+    test("LoadAverage5Min reads the 5 minute figure", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ loadMetrics: load }) }),
+        {
+          checkOn: CheckOn.LoadAverage5Min,
+          filterType: FilterType.GreaterThan,
+          value: 1,
+        },
+      );
+
+      expect(result).toContain(CheckOn.LoadAverage5Min);
+    });
+
+    /*
+     * The 15 minute figure (0.5) is below the threshold, so the same filter
+     * that matched the 1 and 5 minute figures does not fire here — proving
+     * the correct load slot is selected per CheckOn.
+     */
+    test("LoadAverage15Min reads the 15 minute figure → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ loadMetrics: load }) }),
+        {
+          checkOn: CheckOn.LoadAverage15Min,
+          filterType: FilterType.GreaterThan,
+          value: 1,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    test("absent load metrics fall back to zero", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({}) }),
+        {
+          checkOn: CheckOn.LoadAverage1Min,
+          filterType: FilterType.LessThan,
+          value: 1,
+        },
+      );
+
+      expect(result).toContain(CheckOn.LoadAverage1Min);
+    });
+  });
+
+  describe("DiskUsagePercent", () => {
+    /*
+     * A trailing slash on the requested path and none on the reported path
+     * still resolve to the same disk after normalization.
+     */
+    test("matches disk after path normalization and uses percentUsed", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [
+              buildDiskMetric({ diskPath: "/var/log", percentUsed: 75 }),
+            ],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          serverMonitorOptions: { diskPath: "/var/log/" },
+        },
+      );
+
+      expect(result).toContain(CheckOn.DiskUsagePercent);
+      expect(result).toContain("/var/log/");
+    });
+
+    /*
+     * Backslash separators and mixed case both normalize away before the
+     * lookup, so a Windows-style reported path matches a forward-slash
+     * lower-case request.
+     */
+    test("normalizes backslashes and case when matching the disk", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [
+              buildDiskMetric({ diskPath: "C:\\Data", percentUsed: 88 }),
+            ],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          serverMonitorOptions: { diskPath: "c:/data" },
+        },
+      );
+
+      expect(result).toContain(CheckOn.DiskUsagePercent);
+    });
+
+    /*
+     * With no serverMonitorOptions the requested path defaults to "/".
+     */
+    test("defaults to the root disk when no path option is given", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [buildDiskMetric({ diskPath: "/", percentUsed: 40 })],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThanOrEqualTo,
+          value: 40,
+        },
+      );
+
+      expect(result).toContain(CheckOn.DiskUsagePercent);
+    });
+
+    /*
+     * When percentUsed is absent the evaluator falls back to percentFree.
+     */
+    test("falls back to percentFree when percentUsed is absent", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [
+              buildDiskMetric({
+                diskPath: "/data",
+                percentUsed: undefined,
+                percentFree: 65,
+              }),
+            ],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          serverMonitorOptions: { diskPath: "/data" },
+        },
+      );
+
+      expect(result).toContain(CheckOn.DiskUsagePercent);
+    });
+
+    /*
+     * A requested path that matches no reported disk yields a usage of 0.
+     */
+    test("unmatched disk path yields zero usage", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [
+              buildDiskMetric({ diskPath: "/other", percentUsed: 99 }),
+            ],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.LessThan,
+          value: 10,
+          serverMonitorOptions: { diskPath: "/data" },
+        },
+      );
+
+      expect(result).toContain(CheckOn.DiskUsagePercent);
+    });
+  });
+
+  describe("ServerProcessName", () => {
+    test("IsExecuting + process present → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsExecuting,
+          value: "nginx",
+        },
+      );
+
+      expect(result).toBe("Process nginx is executing.");
+    });
+
+    /*
+     * Matching is case-insensitive and trimmed, but the original threshold
+     * text is echoed back in the message.
+     */
+    test("IsExecuting matches case-insensitively", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsExecuting,
+          value: "NGINX",
+        },
+      );
+
+      expect(result).toBe("Process NGINX is executing.");
+    });
+
+    test("IsExecuting + process absent → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsExecuting,
+          value: "redis",
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    test("IsNotExecuting + process absent → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsNotExecuting,
+          value: "redis",
+        },
+      );
+
+      expect(result).toBe("Process redis is not executing.");
+    });
+
+    test("IsNotExecuting + process present → not met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsNotExecuting,
+          value: "nginx",
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * An absent processes array is treated as an empty list, so nothing is
+     * executing and IsNotExecuting is satisfied.
+     */
+    test("absent processes list satisfies IsNotExecuting", async () => {
+      const result: string | null = await evaluate(buildServerResponse({}), {
+        checkOn: CheckOn.ServerProcessName,
+        filterType: FilterType.IsNotExecuting,
+        value: "nginx",
+      });
+
+      expect(result).toBe("Process nginx is not executing.");
+    });
+
+    /*
+     * A falsy threshold (undefined value) skips the whole process block via
+     * the `threshold &&` guard, leaving the check undecidable.
+     */
+    test("undefined threshold skips the process block", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsExecuting,
+          value: undefined,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * A filter type that is neither executing nor not-executing falls through
+     * the process block without a decision.
+     */
+    test("unhandled filter type → undecidable", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.EqualTo,
+          value: "nginx",
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    test("onlyCheckRequestReceivedAt short-circuits the process block", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ name: "nginx" })],
+          onlyCheckRequestReceivedAt: true,
+        }),
+        {
+          checkOn: CheckOn.ServerProcessName,
+          filterType: FilterType.IsExecuting,
+          value: "nginx",
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+
+  describe("ServerProcessPID", () => {
+    test("IsExecuting matches a numeric PID threshold", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ pid: 4321 })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessPID,
+          filterType: FilterType.IsExecuting,
+          value: 4321,
+        },
+      );
+
+      expect(result).toBe("Process with PID 4321 is executing.");
+    });
+
+    test("IsNotExecuting + PID absent → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ pid: 4321 })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessPID,
+          filterType: FilterType.IsNotExecuting,
+          value: "9999",
+        },
+      );
+
+      expect(result).toBe("Process with PID 9999 is not executing.");
+    });
+
+    /*
+     * The PID block has an explicit trailing null return, so an unhandled
+     * filter type resolves to null without falling through to later blocks.
+     */
+    test("unhandled filter type → undecidable", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ pid: 4321 })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessPID,
+          filterType: FilterType.EqualTo,
+          value: 4321,
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+
+  describe("ServerProcessCommand", () => {
+    test("IsExecuting matches an exact command", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ command: "/usr/bin/node server.js" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessCommand,
+          filterType: FilterType.IsExecuting,
+          value: "/usr/bin/node server.js",
+        },
+      );
+
+      expect(result).toBe(
+        "Process with command /usr/bin/node server.js is executing.",
+      );
+    });
+
+    test("IsExecuting matches after trimming and lower-casing", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ command: "/usr/bin/node server.js" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessCommand,
+          filterType: FilterType.IsExecuting,
+          value: "  /USR/bin/Node server.js  ",
+        },
+      );
+
+      expect(result).toContain("is executing.");
+    });
+
+    test("IsNotExecuting + command absent → met", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          processes: [buildProcess({ command: "/usr/bin/node server.js" })],
+        }),
+        {
+          checkOn: CheckOn.ServerProcessCommand,
+          filterType: FilterType.IsNotExecuting,
+          value: "/bin/bash",
+        },
+      );
+
+      expect(result).toBe("Process with command /bin/bash is not executing.");
+    });
+  });
+
+  /*
+   * The evaluate-over-time path replaces the current metric value with the
+   * window recorded for the monitor. The DB-backed lookup is spied so these
+   * tests stay deterministic.
+   */
+  describe("evaluate over time", () => {
+    /** The window produced a usable value. */
+    function usable(
+      value: Array<number | boolean> | number | boolean,
+    ): OverTimeCriteriaValue {
+      return { earlyReturn: null, value: value };
+    }
+
+    /** The window could not back the filter, so nothing is compared. */
+    function noUsableWindow(): OverTimeCriteriaValue {
+      return { earlyReturn: { result: null }, value: undefined };
+    }
+
+    function mockOverTime(
+      result: OverTimeCriteriaValue,
+    ): ReturnType<typeof jest.spyOn> {
+      return jest
+        .spyOn(EvaluateOverTime, "getOverTimeValueForCriteriaFilter")
+        .mockResolvedValue(result) as ReturnType<typeof jest.spyOn>;
+    }
+
+    test("uses the over-time boolean series for IsOnline while online", async () => {
+      const spy: ReturnType<typeof jest.spyOn> = mockOverTime(
+        usable([true, false]),
+      );
+
+      const response: ServerMonitorResponse = buildServerResponse({
+        minutesSinceLastCheck: 5,
+      });
+
+      const result: string | null = await evaluate(response, {
+        checkOn: CheckOn.IsOnline,
+        filterType: FilterType.True,
+        value: undefined,
+        evaluateOverTime: true,
+        evaluateOverTimeOptions: {
+          timeValueInMinutes: 10,
+          evaluateOverTimeType: EvaluateOverTimeType.AnyValue,
+        },
+      });
+
+      expect(result).toContain(CheckOn.IsOnline);
+      expect(spy).toHaveBeenCalledWith(
+        expect.objectContaining({
+          projectId: response.projectId,
+          monitorId: response.monitorId,
+          criteriaFilter: expect.objectContaining({
+            checkOn: CheckOn.IsOnline,
+          }),
+        }),
+      );
+    });
+
+    /*
+     * timeValueInMinutes overrides the default offline threshold: a 5 minute
+     * gap is still "online" when the window is 10 minutes, which would have
+     * been offline under the default of 3.
+     */
+    test("timeValueInMinutes overrides the offline threshold", async () => {
+      mockOverTime(usable([false, false]));
+
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 5 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.False,
+          value: undefined,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 3,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+        },
+      );
+
+      expect(result).toContain("false");
+    });
+
+    /*
+     * "Is Online" is the one filter that keeps its fallback: for a server
+     * monitor the absence of data IS the signal, and the last-check-in
+     * arithmetic above already implements the "wait this long" behaviour. So
+     * an unusable window still falls back to the "recent check -> online"
+     * default rather than declining to evaluate.
+     */
+    test("empty over-time series falls back to the liveness default", async () => {
+      mockOverTime(noUsableWindow());
+
+      const result: string | null = await evaluate(
+        buildServerResponse({ minutesSinceLastCheck: 1 }),
+        {
+          checkOn: CheckOn.IsOnline,
+          filterType: FilterType.True,
+          value: undefined,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 3,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+        },
+      );
+
+      expect(result).toContain("true");
+    });
+
+    /*
+     * A numeric over-time series takes precedence over the point-in-time CPU
+     * metric. The array (70, 80) breaches while the live value (10) would not.
+     */
+    test("over-time numeric series takes precedence over the live CPU metric", async () => {
+      mockOverTime(usable([70, 80]));
+
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 10 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+        },
+      );
+
+      expect(result).toContain(CheckOn.CPUUsagePercent);
+      expect(result).toContain("70");
+    });
+
+    /*
+     * Only the disk-usage series carries a diskPath attribute. Scoping any
+     * other series by it would filter against an attribute no row has, and
+     * an unusable window now means the filter waits rather than comparing
+     * the live value - so the stray option would silence it outright.
+     */
+    test("scopes the window by disk path only for disk usage", async () => {
+      const spy: ReturnType<typeof jest.spyOn> = mockOverTime(usable([95]));
+
+      await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [
+              buildDiskMetric({ diskPath: "/data", percentUsed: 95 }),
+            ],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 90,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+          serverMonitorOptions: { diskPath: "/data" },
+        },
+      );
+
+      expect(spy).toHaveBeenCalledWith(
+        expect.objectContaining({ miscData: { diskPath: "/data" } }),
+      );
+    });
+
+    test("does not scope a CPU window by a stray disk path", async () => {
+      const spy: ReturnType<typeof jest.spyOn> = mockOverTime(usable([70, 80]));
+
+      await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 75 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+          serverMonitorOptions: { diskPath: "/" },
+        },
+      );
+
+      expect(spy).toHaveBeenCalledWith(
+        expect.objectContaining({ miscData: undefined }),
+      );
+    });
+
+    /*
+     * Disk usage read its over-time window and then threw the result away,
+     * comparing the reading from this check instead - so "evaluate over
+     * time" on a disk filter did nothing at all.
+     */
+    test("disk usage compares the over-time window, not just this check", async () => {
+      mockOverTime(usable([95, 96, 97]));
+
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [buildDiskMetric({ diskPath: "/", percentUsed: 10 })],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 90,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+          serverMonitorOptions: { diskPath: "/" },
+        },
+      );
+
+      expect(result).toContain("All values of");
+      expect(result).toContain("95");
+    });
+
+    test("disk usage does not fire while its window is unusable", async () => {
+      mockOverTime(noUsableWindow());
+
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [buildDiskMetric({ diskPath: "/", percentUsed: 99 })],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 90,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.AllValues,
+          },
+          serverMonitorOptions: { diskPath: "/" },
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * A disk filter that does NOT evaluate over time keeps comparing the
+     * reading from this check.
+     */
+    test("disk usage without over-time still compares this check", async () => {
+      const result: string | null = await evaluate(
+        buildServerResponse({
+          metrics: buildMetrics({
+            diskMetrics: [buildDiskMetric({ diskPath: "/", percentUsed: 99 })],
+          }),
+        }),
+        {
+          checkOn: CheckOn.DiskUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 90,
+          serverMonitorOptions: { diskPath: "/" },
+        },
+      );
+
+      expect(result).toContain("99");
+    });
+
+    /*
+     * An average of exactly zero is a real measurement, not a missing one.
+     * It used to be swallowed by a truthiness check and replaced with the
+     * live CPU reading, so "average CPU over 5 minutes above 50%" could fire
+     * on a single busy sample while the average itself was 0.
+     */
+    test("a zero aggregate is compared as zero, not replaced by the live metric", async () => {
+      mockOverTime(usable(0));
+
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 90 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.Average,
+          },
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+
+    /*
+     * A window that could not be read says nothing about the last five
+     * minutes, so the filter waits rather than firing off the single live
+     * reading that happens to have arrived with this check.
+     */
+    test("an unreadable window does not fire off the live CPU metric", async () => {
+      mockOverTime(noUsableWindow());
+
+      const result: string | null = await evaluate(
+        buildServerResponse({ metrics: buildMetrics({ cpuPercentUsed: 90 }) }),
+        {
+          checkOn: CheckOn.CPUUsagePercent,
+          filterType: FilterType.GreaterThan,
+          value: 50,
+          evaluateOverTime: true,
+          evaluateOverTimeOptions: {
+            timeValueInMinutes: 5,
+            evaluateOverTimeType: EvaluateOverTimeType.Average,
+          },
+        },
+      );
+
+      expect(result).toBeNull();
+    });
+  });
+});