@oneuptime/common 12.0.33 → 12.0.34

package/Tests/Server/Utils/LogRedaction.test.ts

@@ -58,6 +58,9 @@ describe("isSensitiveLogKey", () => {
       "userHandle",
       "rawId",
       "signature",
+      "parameters",
+      "parameterValues",
+      "bindParameters",
     ]) {
       expect(isSensitiveLogKey(key)).toBe(true);
     }
@@ -329,6 +332,28 @@ describe("redactLogString", () => {
     ).not.toContain("hunter2");
   });
 
+  it("redacts Telegram bot credentials embedded in request paths", () => {
+    const botToken: string = "1234567890:ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi";
+    const url: string = `https://api.telegram.org/bot${botToken}/sendMessage`;
+
+    const output: string = redactLogString(
+      `Request to ${url} failed with ECONNRESET`,
+    );
+
+    expect(output).not.toContain(botToken);
+    expect(output).toContain(`/bot${REDACTED}/sendMessage`);
+    expect(output).toContain("ECONNRESET");
+  });
+
+  it("redacts a Telegram bot URL nested inside an Error", () => {
+    const botToken: string = "9876543210:ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghi";
+    const error: Error = new Error(
+      `POST https://api.telegram.org/bot${botToken}/sendMessage failed`,
+    );
+
+    expect(serialize(error)).not.toContain(botToken);
+  });
+
   it("leaves ordinary log lines untouched", () => {
     for (const message of [
       "Slack token exchange completed. ok: true",
@@ -339,6 +364,7 @@ describe("redactLogString", () => {
       "User logged in: user@example.com",
       "Monitor probe finished in 42ms",
       "GET https://oneuptime.com/api/status-page/1234 -> 200",
+      "GET https://api.telegram.org/bot/status -> 200",
     ]) {
       expect(redactLogString(message)).toBe(message);
     }

package/Tests/Server/Utils/LoggerCredentialLeak.test.ts

@@ -1,4 +1,6 @@
 import ConfigLogLevel from "../../../Server/Types/ConfigLogLevel";
+import { inspect } from "util";
+import { QueryFailedError } from "typeorm";
 import {
   afterEach,
   beforeEach,
@@ -93,15 +95,7 @@ const collectSinkText: CollectSinkTextFunction = (): string => {
               return argument;
             }
 
-            if (argument instanceof Error) {
-              return `${argument.message} ${argument.stack || ""}`;
-            }
-
-            try {
-              return JSON.stringify(argument);
-            } catch {
-              return String(argument);
-            }
+            return inspect(argument, { depth: null });
           })
           .join(" "),
       );
@@ -235,6 +229,26 @@ describe("logger never writes credentials to any sink", () => {
     expect(collectSinkText()).not.toContain(SENTINEL);
   });
 
+  it("drops database bind parameters carried on QueryFailedError", () => {
+    const verificationToken: string = `0abc1234_${"A".repeat(43)}`;
+    const driverError: Error & { code?: string } = Object.assign(
+      new Error("database connection lost"),
+      { code: "ECONNRESET" },
+    );
+    const error: QueryFailedError = new QueryFailedError(
+      'UPDATE "UserTelegram" SET "isVerified" = $1 WHERE "verificationCode" = $2',
+      [true, verificationToken],
+      driverError,
+    );
+
+    logAtEveryLevel(error);
+
+    const output: string = collectSinkText();
+    expect(output).not.toContain(verificationToken);
+    expect(output).toContain("[REDACTED]");
+    expect(output).toContain("UserTelegram");
+  });
+
   it("still logs an ordinary message unchanged", () => {
     logger.debug("Exchanging Slack authorization code for an access token.");
 

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

@@ -0,0 +1,584 @@
+import CompareCriteria from "../../../../../Server/Utils/Monitor/Criteria/CompareCriteria";
+import {
+  CheckOn,
+  CriteriaFilter,
+  EvaluateOverTimeType,
+  FilterType,
+} from "../../../../../Types/Monitor/CriteriaFilter";
+
+/*
+ * Two defects, both visible in one real alert email.
+ *
+ * The customer received:
+ *   "Any value of (used_cpu / limit_cpu) * 100 is 72.35, 81.54, 79.95,
+ *    91.53, 87.73 % which is greater than 90 %."
+ * Four of those five numbers are below the threshold the sentence claims
+ * they all exceed — `getCompareMessage` printed the whole evaluation window
+ * and then asserted the filter's verdict over it.
+ *
+ * Separately, every comparator branched on `AnyValue` and let EVERY other
+ * evaluation type fall through to `.every()`. Average, Sum, Maximum Value
+ * and Minimum Value — all four selectable in the criteria UI — were
+ * silently evaluated as "All Values".
+ */
+
+// The exact window from the customer's ALT-113.
+const WINDOW: Array<number> = [72.35, 81.54, 79.95, 91.53, 87.73];
+
+function metricFilter(
+  filterType: FilterType,
+  evaluationType: EvaluateOverTimeType,
+): CriteriaFilter {
+  return {
+    checkOn: CheckOn.MetricValue,
+    filterType: filterType,
+    value: 90,
+    metricMonitorOptions: {
+      metricAggregationType: evaluationType,
+      metricAlias: "pod_cpu_limit_saturation",
+    },
+  } as CriteriaFilter;
+}
+
+describe("CompareCriteria aggregation semantics", () => {
+  describe("reduceWindow", () => {
+    test("Average collapses the window to its mean", () => {
+      expect(
+        CompareCriteria.reduceWindow({
+          values: [10, 20, 30],
+          evaluationType: EvaluateOverTimeType.Average,
+        }),
+      ).toEqual([20]);
+    });
+
+    test("Sum collapses the window to its total", () => {
+      expect(
+        CompareCriteria.reduceWindow({
+          values: [10, 20, 30],
+          evaluationType: EvaluateOverTimeType.Sum,
+        }),
+      ).toEqual([60]);
+    });
+
+    test("Maximum Value collapses to the largest sample", () => {
+      expect(
+        CompareCriteria.reduceWindow({
+          values: [10, 30, 20],
+          evaluationType: EvaluateOverTimeType.MaximumValue,
+        }),
+      ).toEqual([30]);
+    });
+
+    test("Minimum Value collapses to the smallest sample", () => {
+      expect(
+        CompareCriteria.reduceWindow({
+          values: [30, 10, 20],
+          evaluationType: EvaluateOverTimeType.MunimumValue,
+        }),
+      ).toEqual([10]);
+    });
+
+    test.each([
+      EvaluateOverTimeType.AnyValue,
+      EvaluateOverTimeType.AllValues,
+      undefined,
+    ])(
+      "%s quantifies over the untouched window",
+      (evaluationType: EvaluateOverTimeType | undefined) => {
+        expect(
+          CompareCriteria.reduceWindow({
+            values: [10, 20, 30],
+            evaluationType: evaluationType,
+          }),
+        ).toEqual([10, 20, 30]);
+      },
+    );
+
+    test("an empty window is returned unchanged rather than reduced to NaN", () => {
+      expect(
+        CompareCriteria.reduceWindow({
+          values: [],
+          evaluationType: EvaluateOverTimeType.Average,
+        }),
+      ).toEqual([]);
+    });
+  });
+
+  describe("the four reducing aggregations are actually applied", () => {
+    /*
+     * The mean of the production window is 82.62. Under the old code
+     * "Average > 90" was evaluated as "every sample > 90" — false, but for
+     * the wrong reason. These assertions pin the reason.
+     */
+    test("Average > 90 is false because the mean is 82.62, not because a sample was low", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: 90,
+        }),
+      ).toBe(false);
+
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: 80,
+        }),
+      ).toBe(true);
+    });
+
+    test("Average > 80 was WRONG before this fix: AllValues semantics said false", () => {
+      /*
+       * If Average were still falling through to `.every()`, 79.95 would
+       * sink it. It does not, because the mean is what is compared.
+       */
+      expect(
+        WINDOW.every((v: number) => {
+          return v > 80;
+        }),
+      ).toBe(false);
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: 80,
+        }),
+      ).toBe(true);
+    });
+
+    test("Maximum Value fires on the single high sample", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.MaximumValue,
+          threshold: 90,
+        }),
+      ).toBe(true);
+    });
+
+    test("Minimum Value does not", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.MunimumValue,
+          threshold: 90,
+        }),
+      ).toBe(false);
+    });
+
+    test("Sum aggregates across the window", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: [1, 2, 3],
+          evaluationType: EvaluateOverTimeType.Sum,
+          threshold: 5,
+        }),
+      ).toBe(true);
+      expect(
+        CompareCriteria.greaterThan({
+          value: [1, 2, 3],
+          evaluationType: EvaluateOverTimeType.Sum,
+          threshold: 6,
+        }),
+      ).toBe(false);
+    });
+
+    test("every comparator honours the aggregation, not just greaterThan", () => {
+      const mean: number = 20; // mean of [10, 20, 30]
+      const values: Array<number> = [10, 20, 30];
+
+      expect(
+        CompareCriteria.lessThan({
+          value: values,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: mean + 1,
+        }),
+      ).toBe(true);
+      expect(
+        CompareCriteria.greaterThanOrEqual({
+          value: values,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: mean,
+        }),
+      ).toBe(true);
+      expect(
+        CompareCriteria.lessThanOrEqual({
+          value: values,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: mean,
+        }),
+      ).toBe(true);
+      expect(
+        CompareCriteria.equalTo({
+          value: values,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: mean,
+        }),
+      ).toBe(true);
+      expect(
+        CompareCriteria.notEqualTo({
+          value: values,
+          evaluationType: EvaluateOverTimeType.Average,
+          threshold: mean,
+        }),
+      ).toBe(false);
+    });
+  });
+
+  describe("AnyValue and AllValues keep their existing semantics", () => {
+    test("AnyValue fires when one sample breaches", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.AnyValue,
+          threshold: 90,
+        }),
+      ).toBe(true);
+    });
+
+    test("AllValues does not", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: WINDOW,
+          evaluationType: EvaluateOverTimeType.AllValues,
+          threshold: 90,
+        }),
+      ).toBe(false);
+    });
+
+    test("an unspecified evaluation type still defaults to AllValues", () => {
+      expect(
+        CompareCriteria.greaterThan({ value: WINDOW, threshold: 90 }),
+      ).toBe(false);
+    });
+
+    test("a scalar value is compared directly", () => {
+      expect(CompareCriteria.greaterThan({ value: 91, threshold: 90 })).toBe(
+        true,
+      );
+    });
+
+    test("an empty window never matches", () => {
+      expect(
+        CompareCriteria.greaterThan({
+          value: [],
+          evaluationType: EvaluateOverTimeType.AnyValue,
+          threshold: 90,
+        }),
+      ).toBe(false);
+    });
+  });
+
+  describe("the root cause quotes only the samples that breached", () => {
+    test("the exact sentence a customer received is no longer produced", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: WINDOW,
+        threshold: 90,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.AnyValue,
+        ),
+        metricDisplayName: "(used_cpu / limit_cpu) * 100",
+        unit: "%",
+      });
+
+      // The four non-breaching samples are gone.
+      expect(message).not.toContain("72.35");
+      expect(message).not.toContain("81.54");
+      expect(message).not.toContain("79.95");
+      expect(message).not.toContain("87.73");
+
+      /*
+       * The one that did breach is named, and the reader is told how much
+       * of the window it represents.
+       */
+      expect(message).toContain("91.53");
+      expect(message).toContain(
+        "1 of 5 samples in the evaluation window breached this threshold.",
+      );
+    });
+
+    test("a fully breaching window keeps the shorter sentence", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: [95, 96, 97],
+        threshold: 90,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.AnyValue,
+        ),
+        metricDisplayName: "CPU",
+        unit: "%",
+      });
+
+      expect(message).toBe(
+        "Any value of CPU is 95, 96, 97 % which is greater than 90 %.",
+      );
+    });
+
+    test("a below-threshold recovery message quotes only the recovering samples", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: [95, 80, 70],
+        threshold: 90,
+        criteriaFilter: metricFilter(
+          FilterType.LessThanOrEqualTo,
+          EvaluateOverTimeType.AnyValue,
+        ),
+        metricDisplayName: "CPU",
+        unit: "%",
+      });
+
+      expect(message).toContain("80, 70");
+      expect(message).not.toContain("95");
+    });
+
+    test("AllValues is unaffected — every sample breached by definition", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: [95, 96],
+        threshold: 90,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.AllValues,
+        ),
+        metricDisplayName: "CPU",
+        unit: "%",
+      });
+
+      expect(message).toBe(
+        "All values of CPU is 95, 96 % which is greater than 90 %.",
+      );
+    });
+
+    test("a reducing aggregation quotes the aggregate, not the raw window", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: [10, 20, 30],
+        threshold: 15,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.Average,
+        ),
+        metricDisplayName: "CPU",
+        unit: "%",
+      });
+
+      expect(message).toContain("The average of CPU is 20");
+      expect(message).not.toContain("10, 20, 30");
+    });
+
+    test.each([
+      [EvaluateOverTimeType.Average, "The average of"],
+      [EvaluateOverTimeType.Sum, "The sum of"],
+      [EvaluateOverTimeType.MaximumValue, "The maximum of"],
+      [EvaluateOverTimeType.MunimumValue, "The minimum of"],
+      [EvaluateOverTimeType.AnyValue, "Any value of"],
+      [EvaluateOverTimeType.AllValues, "All values of"],
+    ])(
+      "%s is named in the sentence as %s",
+      (evaluationType: EvaluateOverTimeType, expectedPrefix: string) => {
+        const message: string = CompareCriteria.getCompareMessage({
+          values: [100],
+          threshold: 90,
+          criteriaFilter: metricFilter(FilterType.GreaterThan, evaluationType),
+          metricDisplayName: "CPU",
+          unit: "%",
+        });
+
+        expect(message.startsWith(expectedPrefix)).toBe(true);
+      },
+    );
+
+    test("a message rendered for a filter that did not match degrades gracefully", () => {
+      /*
+       * Nothing in the window is above 200; the whole window is quoted
+       * rather than an empty list.
+       */
+      const message: string = CompareCriteria.getCompareMessage({
+        values: WINDOW,
+        threshold: 200,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.AnyValue,
+        ),
+        metricDisplayName: "CPU",
+        unit: "%",
+      });
+
+      expect(message).toContain("72.35");
+      expect(message).not.toContain("samples in the evaluation window");
+    });
+
+    test("a non-numeric window is left alone", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: [true, false],
+        threshold: 1,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.AnyValue,
+        ),
+        metricDisplayName: "Up",
+      });
+
+      expect(message).toContain("true, false");
+    });
+
+    test("getBreachingValues returns only the matching samples", () => {
+      expect(
+        CompareCriteria.getBreachingValues({
+          values: WINDOW,
+          evaluationType: EvaluateOverTimeType.AnyValue,
+          predicate: (value: number) => {
+            return value > 90;
+          },
+        }),
+      ).toEqual([91.53]);
+    });
+  });
+
+  /*
+   * The unit reaching getCompareMessage is the raw OTel/UCUM code the
+   * exporter declared, not a display label. Two of those codes are not
+   * units a reader should ever see:
+   *
+   *   "1"      UCUM's dimensionless marker for a ratio metric. Appended
+   *            verbatim it produced "is 0.85 1 which is greater than
+   *            0.8 1" — a sentence readers parse as a truncated number or
+   *            a typo, and which claims 1% for a value that is 85%.
+   *            Shipped today by web_vital.cls (thresholdUnit "1"),
+   *            oneuptime.host.heartbeat, and every kubeletstats ratio
+   *            gauge once the native-unit map reaches the infra workers.
+   *
+   *   "{cpu}"  UCUM annotation-only units. The braces are descriptive;
+   *            the value itself is the quantity.
+   *
+   * Everything else must survive untouched — this suppression is exactly
+   * two cases wide, not a general unit filter.
+   */
+  describe("the root cause never prints a unit that is not a unit", () => {
+    function unitMessage(unit: string | undefined): string {
+      return CompareCriteria.getCompareMessage({
+        values: 0.85,
+        threshold: 0.8,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThan,
+          EvaluateOverTimeType.AnyValue,
+        ),
+        metricDisplayName: "Metric Value",
+        unit: unit,
+      });
+    }
+
+    test("the dimensionless '1' is dropped from the value AND the threshold", () => {
+      const message: string = unitMessage("1");
+
+      expect(message).toBe(
+        "Any value of Metric Value is 0.85 which is greater than 0.8.",
+      );
+      // The two halves of the defect, pinned separately.
+      expect(message).not.toContain("0.85 1");
+      expect(message).not.toContain("0.8 1");
+      // No stray dimensionless token anywhere in the sentence.
+      expect(message).not.toMatch(/\d\s1\b/);
+    });
+
+    test("the CLS sentence a customer received is no longer produced", () => {
+      const message: string = CompareCriteria.getCompareMessage({
+        values: [0.31],
+        threshold: 0.25,
+        criteriaFilter: metricFilter(
+          FilterType.GreaterThanOrEqualTo,
+          EvaluateOverTimeType.Average,
+        ),
+        metricDisplayName: "Cumulative Layout Shift",
+        unit: "1",
+      });
+
+      expect(message).toContain("0.31 which is");
+      expect(message).not.toMatch(/\d\s1\b/);
+    });
+
+    test("annotation-only units are dropped, braces and all", () => {
+      const message: string = unitMessage("{cpu}");
+
+      expect(message).not.toContain("{cpu}");
+      expect(message).not.toContain("{");
+      expect(message).toBe(
+        "Any value of Metric Value is 0.85 which is greater than 0.8.",
+      );
+
+      expect(unitMessage("{packets}")).not.toContain("{packets}");
+      expect(unitMessage("{errors}")).not.toContain("{errors}");
+    });
+
+    test("real units still survive on BOTH the value and the threshold", () => {
+      /*
+       * The guard against over-suppression. If this ever goes quiet the
+       * fix has widened into a general unit filter.
+       */
+      expect(unitMessage("By")).toBe(
+        "Any value of Metric Value is 0.85 By which is greater than 0.8 By.",
+      );
+      expect(unitMessage("%")).toBe(
+        "Any value of Metric Value is 0.85 % which is greater than 0.8 %.",
+      );
+      expect(unitMessage("ms")).toBe(
+        "Any value of Metric Value is 0.85 ms which is greater than 0.8 ms.",
+      );
+      // "1" is suppressed; a unit that merely CONTAINS a 1 is not.
+      expect(unitMessage("m/s2")).toContain("0.85 m/s2");
+      expect(unitMessage("10*3/uL")).toContain("0.85 10*3/uL");
+    });
+
+    test("a stored unit with stray whitespace renders once, trimmed", () => {
+      const message: string = unitMessage(" ms ");
+
+      expect(message).toBe(
+        "Any value of Metric Value is 0.85 ms which is greater than 0.8 ms.",
+      );
+      expect(message).not.toContain("  ");
+    });
+
+    test("a whitespace-only unit is treated as no unit at all", () => {
+      expect(unitMessage("   ")).toBe(
+        "Any value of Metric Value is 0.85 which is greater than 0.8.",
+      );
+      expect(unitMessage(undefined)).toBe(
+        "Any value of Metric Value is 0.85 which is greater than 0.8.",
+      );
+      // A padded "1" is still the dimensionless marker.
+      expect(unitMessage(" 1 ")).not.toMatch(/\d\s1\b/);
+    });
+
+    test("every comparator branch takes the same suffix", () => {
+      /*
+       * The suffix is appended in seven places — once on the observed
+       * values and once per comparator. A partial fix would leave the
+       * threshold half carrying "1" while the value half dropped it.
+       */
+      const comparators: Array<FilterType> = [
+        FilterType.GreaterThan,
+        FilterType.GreaterThanOrEqualTo,
+        FilterType.LessThan,
+        FilterType.LessThanOrEqualTo,
+        FilterType.EqualTo,
+        FilterType.NotEqualTo,
+      ];
+
+      for (const filterType of comparators) {
+        const message: string = CompareCriteria.getCompareMessage({
+          values: 0.85,
+          threshold: 0.8,
+          criteriaFilter: metricFilter(
+            filterType,
+            EvaluateOverTimeType.AnyValue,
+          ),
+          metricDisplayName: "Metric Value",
+          unit: "1",
+        });
+
+        expect(message).not.toMatch(/\d\s1\b/);
+        expect(message).toContain("0.85");
+        expect(message).toContain("0.8.");
+      }
+    });
+  });
+});