Error alerting

Author: matt-aitken

apps/webapp/app/models/projectAlert.server.ts

@@ -32,3 +32,9 @@ export const ProjectAlertSlackStorage = z.object({
 });
 
 export type ProjectAlertSlackStorage = z.infer<typeof ProjectAlertSlackStorage>;
+
+export const ErrorAlertConfig = z.object({
+  evaluationIntervalMs: z.number().min(60_000).default(300_000),
+});
+
+export type ErrorAlertConfig = z.infer<typeof ErrorAlertConfig>;

apps/webapp/app/presenters/v3/ErrorGroupPresenter.server.ts

@@ -2,7 +2,7 @@ import { z } from "zod";
 import { type ClickHouse, msToClickHouseInterval } from "@internal/clickhouse";
 import { TimeGranularity } from "~/utils/timeGranularity";
 import { ErrorId } from "@trigger.dev/core/v3/isomorphic";
-import { type PrismaClientOrTransaction } from "@trigger.dev/database";
+import { type ErrorGroupStatus, type PrismaClientOrTransaction } from "@trigger.dev/database";
 import { timeFilterFromTo } from "~/components/runs/v3/SharedFilters";
 import { type Direction, DirectionSchema } from "~/components/ListPagination";
 import { findDisplayableEnvironment } from "~/models/runtimeEnvironment.server";
@@ -61,6 +61,18 @@ function parseClickHouseDateTime(value: string): Date {
   return new Date(value.replace(" ", "T") + "Z");
 }
 
+export type ErrorGroupState = {
+  status: ErrorGroupStatus;
+  resolvedAt: Date | null;
+  resolvedInVersion: string | null;
+  resolvedBy: string | null;
+  ignoredUntil: Date | null;
+  ignoredReason: string | null;
+  ignoredByUserId: string | null;
+  ignoredUntilOccurrenceRate: number | null;
+  ignoredUntilTotalOccurrences: number | null;
+};
+
 export type ErrorGroupSummary = {
   fingerprint: string;
   errorType: string;
@@ -70,6 +82,7 @@ export type ErrorGroupSummary = {
   firstSeen: Date;
   lastSeen: Date;
   affectedVersions: string[];
+  state: ErrorGroupState;
 };
 
 export type ErrorGroupOccurrences = Awaited<ReturnType<ErrorGroupPresenter["getOccurrences"]>>;
@@ -114,7 +127,7 @@ export class ErrorGroupPresenter extends BasePresenter {
       defaultPeriod: "7d",
     });
 
-    const [summary, affectedVersions, runList] = await Promise.all([
+    const [summary, affectedVersions, runList, stateRow] = await Promise.all([
       this.getSummary(organizationId, projectId, environmentId, fingerprint),
       this.getAffectedVersions(organizationId, projectId, environmentId, fingerprint),
       this.getRunList(organizationId, environmentId, {
@@ -128,10 +141,22 @@ export class ErrorGroupPresenter extends BasePresenter {
         cursor,
         direction,
       }),
+      this.getState(environmentId, fingerprint),
     ]);
 
     if (summary) {
       summary.affectedVersions = affectedVersions;
+      summary.state = stateRow ?? {
+        status: "UNRESOLVED",
+        resolvedAt: null,
+        resolvedInVersion: null,
+        resolvedBy: null,
+        ignoredUntil: null,
+        ignoredReason: null,
+        ignoredByUserId: null,
+        ignoredUntilOccurrenceRate: null,
+        ignoredUntilTotalOccurrences: null,
+      };
     }
 
     return {
@@ -290,6 +315,31 @@ export class ErrorGroupPresenter extends BasePresenter {
     return sortVersionsDescending(versions).slice(0, 5);
   }
 
+  private async getState(
+    environmentId: string,
+    fingerprint: string
+  ): Promise<ErrorGroupState | null> {
+    const row = await this.replica.errorGroupState.findFirst({
+      where: {
+        environmentId,
+        errorFingerprint: fingerprint,
+      },
+      select: {
+        status: true,
+        resolvedAt: true,
+        resolvedInVersion: true,
+        resolvedBy: true,
+        ignoredUntil: true,
+        ignoredReason: true,
+        ignoredByUserId: true,
+        ignoredUntilOccurrenceRate: true,
+        ignoredUntilTotalOccurrences: true,
+      },
+    });
+
+    return row;
+  }
+
   private async getRunList(
     organizationId: string,
     environmentId: string,

apps/webapp/app/presenters/v3/ErrorsListPresenter.server.ts

@@ -9,7 +9,7 @@ const errorsListGranularity = new TimeGranularity([
   { max: "3 months", granularity: "1w" },
   { max: "Infinity", granularity: "30d" },
 ]);
-import { type PrismaClientOrTransaction } from "@trigger.dev/database";
+import { type ErrorGroupStatus, type PrismaClientOrTransaction } from "@trigger.dev/database";
 import { type Direction } from "~/components/ListPagination";
 import { timeFilterFromTo } from "~/components/runs/v3/SharedFilters";
 import { findDisplayableEnvironment } from "~/models/runtimeEnvironment.server";
@@ -23,6 +23,7 @@ export type ErrorsListOptions = {
   // filters
   tasks?: string[];
   versions?: string[];
+  status?: ErrorGroupStatus;
   period?: string;
   from?: number;
   to?: number;
@@ -41,6 +42,7 @@ export const ErrorsListOptionsSchema = z.object({
   projectId: z.string(),
   tasks: z.array(z.string()).optional(),
   versions: z.array(z.string()).optional(),
+  status: z.enum(["UNRESOLVED", "RESOLVED", "IGNORED"]).optional(),
   period: z.string().optional(),
   from: z.number().int().nonnegative().optional(),
   to: z.number().int().nonnegative().optional(),
@@ -90,7 +92,11 @@ function decodeCursor(cursor: string): ErrorGroupCursor | null {
   }
 }
 
-function cursorFromRow(row: { occurrence_count: number; error_fingerprint: string; task_identifier: string }): string {
+function cursorFromRow(row: {
+  occurrence_count: number;
+  error_fingerprint: string;
+  task_identifier: string;
+}): string {
   return encodeCursor({
     occurrenceCount: row.occurrence_count,
     fingerprint: row.error_fingerprint,
@@ -126,6 +132,7 @@ export class ErrorsListPresenter extends BasePresenter {
       projectId,
       tasks,
       versions,
+      status,
       period,
       search,
       from,
@@ -161,6 +168,7 @@ export class ErrorsListPresenter extends BasePresenter {
       (tasks !== undefined && tasks.length > 0) ||
       (versions !== undefined && versions.length > 0) ||
       (search !== undefined && search !== "") ||
+      status !== undefined ||
       !time.isDefault;
 
     const possibleTasksAsync = getAllTaskIdentifiers(this.replica, environmentId);
@@ -262,15 +270,14 @@ export class ErrorsListPresenter extends BasePresenter {
 
     // Fetch global first_seen / last_seen from the errors_v1 summary table
     const fingerprints = errorGroups.map((e) => e.error_fingerprint);
-    const globalSummaryMap = await this.getGlobalSummary(
-      organizationId,
-      projectId,
-      environmentId,
-      fingerprints
-    );
+    const [globalSummaryMap, stateMap] = await Promise.all([
+      this.getGlobalSummary(organizationId, projectId, environmentId, fingerprints),
+      this.getErrorGroupStates(environmentId, errorGroups),
+    ]);
 
-    const transformedErrorGroups = errorGroups.map((error) => {
+    let transformedErrorGroups = errorGroups.map((error) => {
       const global = globalSummaryMap.get(error.error_fingerprint);
+      const state = stateMap.get(`${error.task_identifier}:${error.error_fingerprint}`);
       return {
         errorType: error.error_type,
         errorMessage: error.error_message,
@@ -279,9 +286,16 @@ export class ErrorsListPresenter extends BasePresenter {
         firstSeen: global?.firstSeen ?? new Date(),
         lastSeen: global?.lastSeen ?? new Date(),
         count: error.occurrence_count,
+        status: state?.status ?? "UNRESOLVED",
+        resolvedAt: state?.resolvedAt ?? null,
+        ignoredUntil: state?.ignoredUntil ?? null,
       };
     });
 
+    if (status) {
+      transformedErrorGroups = transformedErrorGroups.filter((g) => g.status === status);
+    }
+
     return {
       errorGroups: transformedErrorGroups,
       pagination: {
@@ -291,6 +305,7 @@ export class ErrorsListPresenter extends BasePresenter {
       filters: {
         tasks,
         versions,
+        status,
         search,
         period: time,
         from: effectiveFrom,
@@ -376,6 +391,51 @@ export class ErrorsListPresenter extends BasePresenter {
     return { data };
   }
 
+  /**
+   * Batch-fetch ErrorGroupState rows from Postgres for the given ClickHouse error groups.
+   * Returns a map keyed by `${taskIdentifier}:${errorFingerprint}`.
+   */
+  private async getErrorGroupStates(
+    environmentId: string,
+    errorGroups: Array<{ task_identifier: string; error_fingerprint: string }>
+  ) {
+    type StateValue = {
+      status: ErrorGroupStatus;
+      resolvedAt: Date | null;
+      ignoredUntil: Date | null;
+    };
+
+    const result = new Map<string, StateValue>();
+    if (errorGroups.length === 0) return result;
+
+    const states = await this.replica.errorGroupState.findMany({
+      where: {
+        environmentId,
+        OR: errorGroups.map((e) => ({
+          taskIdentifier: e.task_identifier,
+          errorFingerprint: e.error_fingerprint,
+        })),
+      },
+      select: {
+        taskIdentifier: true,
+        errorFingerprint: true,
+        status: true,
+        resolvedAt: true,
+        ignoredUntil: true,
+      },
+    });
+
+    for (const state of states) {
+      result.set(`${state.taskIdentifier}:${state.errorFingerprint}`, {
+        status: state.status,
+        resolvedAt: state.resolvedAt,
+        ignoredUntil: state.ignoredUntil,
+      });
+    }
+
+    return result;
+  }
+
   /**
    * Fetches global first_seen / last_seen for a set of fingerprints from errors_v1.
    */

apps/webapp/app/routes/_app.orgs.$organizationSlug.projects.$projectParam.env.$envParam.alerts/route.tsx

@@ -63,6 +63,7 @@ import {
   v3NewProjectAlertPath,
   v3ProjectAlertsPath,
 } from "~/utils/pathBuilder";
+import { alertsWorker } from "~/v3/alertsWorker.server";
 
 export const meta: MetaFunction = () => {
   return [
@@ -156,6 +157,17 @@ export const action = async ({ request, params }: ActionFunctionArgs) => {
         data: { enabled: true },
       });
 
+      if (alertChannel.alertTypes.includes("ERROR_GROUP")) {
+        await alertsWorker.enqueue({
+          id: `evaluateErrorAlerts:${project.id}`,
+          job: "v3.evaluateErrorAlerts",
+          payload: {
+            projectId: project.id,
+            scheduledAt: Date.now(),
+          },
+        });
+      }
+
       return redirectWithSuccessMessage(
         v3ProjectAlertsPath({ slug: organizationSlug }, { slug: projectParam }, { slug: envParam }),
         request,
@@ -556,7 +568,7 @@ export function alertTypeTitle(alertType: ProjectAlertType): string {
     case "DEPLOYMENT_SUCCESS":
       return "Deployment success";
     default: {
-      assertNever(alertType);
+      throw new Error(`Unknown alertType: ${alertType}`);
     }
   }
 }