Skip to content

armgit / dub

Public Imported from GitHub
0

The modern link attribution platform. Loved by world-class engineering teams at Twilio, Zillow, Jobber, Vercel, Raycast, and more.

Add skipDuplicates and docs to partner seeding script

Rafi9151149f1bf8
diff --git a/apps/web/scripts/dev/seed-100k-partners.ts b/apps/web/scripts/dev/seed-100k-partners.ts
index 83e1264310..5bce5bddcd 100644
--- a/apps/web/scripts/dev/seed-100k-partners.ts
+++ b/apps/web/scripts/dev/seed-100k-partners.ts
@@ -1,3 +1,34 @@
+/**
+ * ====================================================================================
+ * 🚀 High-Scale Partner Data Seeding Script
+ * ====================================================================================
+ *
+ * PURPOSE:
+ * Bulk-generates and inserts realistic partner records into the local database
+ * for local development, performance benchmarking, and testing partner search at scale.
+ *
+ * GENERATED DATA MODEL:
+ * 1. Partner Name (`name`): Realistic first & last name combinations.
+ * 2. Partner Email (`email`): Includes prefix & substring test patterns for search evaluation.
+ * 3. Partner Company Name (`companyName`): Business & agency names.
+ * 4. Partner Description (`description`): Marketing & creator profile text.
+ * 5. Partner Platforms (`PartnerPlatform`): Assigns 1-2 web/social platforms per partner
+ * (website, youtube, twitter, linkedin, instagram, tiktok), generating ~150,000 total platform rows.
+ * 6. Partner Short Links (`Link`): Generates valid `https://` short referral links per partner.
+ *
+ * PERFORMANCE & ARCHITECTURE DECISIONS:
+ * - Chunked Bulk Insertions: Processes generation in memory and bulk-inserts using
+ * `prisma.createMany` in chunks of 2,500 records to insert ~400,000+ total rows in seconds.
+ * - Pre-Computed Password Hash: Pre-computes `"password"` hash once to optimize generation time.
+ * - Seed Fingerprinting: Generates deterministic namespace hashes (`seedFingerprint`) to ensure
+ * unique, non-colliding records across runs.
+ *
+ * CLI USAGE:
+ * cd apps/web
+ * pnpm run script dev/seed-100k-partners [--count=100000] [--programId=prog_123] [--seed=custom-seed]
+ * ====================================================================================
+ */
+
import { createId } from "@/lib/api/create-id";
import { hashPassword } from "@/lib/auth/password";
import { prisma } from "@/lib/prisma";
@@ -291,12 +322,18 @@ const insertPartnerChunk = async ({
}: PartnerChunk) => {
// Keep every chunk atomic so a failed write cannot leave partial relations.
const [, partnerResult] = await prisma.$transaction([
- prisma.user.createMany({ data: users }),
- prisma.partner.createMany({ data: partners }),
- prisma.partnerUser.createMany({ data: partnerUsers }),
- prisma.programEnrollment.createMany({ data: enrollments }),
- prisma.partnerPlatform.createMany({ data: platforms }),
- prisma.link.createMany({ data: links }),
+ prisma.user.createMany({ data: users, skipDuplicates: true }),
+ prisma.partner.createMany({ data: partners, skipDuplicates: true }),
+ prisma.partnerUser.createMany({ data: partnerUsers, skipDuplicates: true }),
+ prisma.programEnrollment.createMany({
+ data: enrollments,
+ skipDuplicates: true,
+ }),
+ prisma.partnerPlatform.createMany({
+ data: platforms,
+ skipDuplicates: true,
+ }),
+ prisma.link.createMany({ data: links, skipDuplicates: true }),
]);
return partnerResult.count;
@@ -339,6 +376,8 @@ async function main() {
console.log(` Workspace: "${workspace.name}" (${workspace.id})\n`);
// Step 3: Build the stable namespace shared by every generated chunk.
+ // Pre-compute the password hash for 'password' once to avoid computing
+ // 100,000 separate bcrypt hashes during seeding.
const passwordHash = await hashPassword("password");
const seedNamespace = `${program.id}:${seed}`;
const seedFingerprint = createHash("sha256")