🎯 Sistem Terbaru: Staggered Start + Parallel Execution untuk performa maksimal!
- Apa itu Rate Limit?
- Sistem Baru v3 Beta: Staggered Start
- 2 Mode Rate Limiting
- Rekomendasi Formula by Sistem
- Cara Kerja Worker Rotation
- Simulasi Lengkap 5 Skenario
- Error dan Solusinya
- Perkiraan Waktu
- Panduan Penggunaan Lengkap
- Checklist Sebelum Generate
- Tips Pro
- Troubleshooting Cepat
Rate Limit adalah batasan jumlah request yang bisa dilakukan ke server dalam waktu tertentu.
Bayangkan restoran McDonald's:
- Kasir = API Server (Google/GenAI)
- Pelanggan = Worker (program generate video)
- Rate Limit = "Maksimal 10 pelanggan per menit per kasir"
❌ 20 pelanggan datang BERSAMAAN (burst)
→ Kasir kewalahan
→ Server tolak request
→ Error 429 "Too Many Requests"
→ Generate gagal!
✅ Pelanggan 1 → Masuk antrian
Tunggu 15 detik...
✅ Pelanggan 2 → Masuk antrian
Tunggu 15 detik...
✅ Pelanggan 3 → Masuk antrian
Tunggu 15 detik...
✅ Pelanggan 4 → Masuk antrian
→ Kasir melayani dengan tenang
→ Tidak ada yang ditolak
→ Semua generate sukses!
Staggered Start = Workers dimulai secara bertahap dengan delay antar start, tapi setelah start, semua jalan PARALLEL (bersamaan).
┌─────────────────────────────────────────────────────┐
│ STAGGERED START (Mencegah Burst Request) │
├─────────────────────────────────────────────────────┤
│ │
│ t=0s: Worker 1 → START ✅ │
│ Main: Tunggu 15 detik... 😴 │
│ │
│ t=15s: Worker 2 → START ✅ │
│ Main: Tunggu 15 detik... 😴 │
│ │
│ t=30s: Worker 3 → START ✅ │
│ Main: Tunggu 15 detik... 😴 │
│ │
│ t=45s: Worker 4 → START ✅ │
│ Main: Selesai! 🎉 │
│ │
├─────────────────────────────────────────────────────┤
│ PARALLEL EXECUTION (Setelah Semua Start) │
├─────────────────────────────────────────────────────┤
│ │
│ Worker 1: Generating... (parallel) 🟢 │
│ Worker 2: Generating... (parallel) 🟢 │
│ Worker 3: Generating... (parallel) 🟢 │
│ Worker 4: Generating... (parallel) 🟢 │
│ │
│ ✅ Semua workers jalan BERSAMAAN! │
│ ✅ TIDAK ada yang tunggu worker lain selesai! │
│ │
└─────────────────────────────────────────────────────┘
| Aspek | Sistem Lama (v3.1) | Sistem Baru (v3.06 Beta) |
|---|---|---|
| Start Workers | Bersamaan (burst) ❌ | Bertahap (staggered) ✅ |
| Blocking | Ada queue blocking ❌ | Tidak ada blocking ✅ |
| Execution | Serial (satu-satu) ❌ | Parallel (bersamaan) ✅ |
| Performa | Lambat 🐢 | Cepat 🚀 |
| Rate Limit | Sering error 429 ❌ | Jarang error ✅ |
Aplikasi menyediakan 2 mode sederhana untuk mengatur rate limiting:
Mode Otomatis - Sistem menghitung optimal settings berdasarkan jumlah akun yang terdeteksi.
✅ Keuntungan:
- Tidak perlu mikir pengaturan
- Sudah ditest optimal
- Auto-adjust berdasarkan akun
- Recommended untuk pemula
❌ Kekurangan:
- Tidak bisa custom
- Fixed formula
Cara Pakai:
- Centang radio button "Formula by Sistem"
- Klik Start
- Sistem auto-detect jumlah akun
- Sistem auto-set optimal workers & delay
- Done! ✅
Mode Manual - Kamu atur sendiri max worker per akun dan delay.
✅ Keuntungan:
- Full control
- Bisa eksperimen
- Cocok untuk advanced user
❌ Kekurangan:
- Perlu paham konsep
- Salah setting bisa rate limit
- Butuh trial & error
Cara Pakai:
- Centang radio button "Setting Your Rate Limiting"
- Atur Max Worker Per Akun (default: 4, recommended: 4)
- Atur Delay Antar Worker (default: 15s, recommended: 10-20s)
- Atur delay lainnya (Generate/Status/Upload/Upscale)
- Klik Start
Settings Manual:
┌─────────────────────────────────────────────────┐
│ Max Worker Per Akun: [4] (rekomendasi: 4) │
│ Delay Antar Worker: [15] detik (10-20s) │
│ Delay After Generate: [3] detik │
│ Delay After Status: [3] detik │
│ Delay After Upload: [3] detik (I2V) │
│ Delay After Upscale: [3] detik (4K) │
└─────────────────────────────────────────────────┘
Info:
ℹ️ Terdeteksi: 3 akun
Total Workers: 12 (4 per akun × 3)
Berikut formula optimal yang digunakan sistem otomatis:
| Jumlah Akun | Max Worker/Akun | Total Workers | Delay Antar Worker | Delay Generate/Status |
|---|---|---|---|---|
| 1 akun | 4 | 4 | 15 detik | 3 detik |
| 2 akun | 4 | 8 | 12 detik | 3 detik |
| 3 akun | 4 | 12 | 7 detik | 3 detik |
| 4-5 akun | 4 | 16-20 | 5 detik | 3 detik |
| 6+ akun | 4 | 24+ | 3 detik | 3 detik |
1️⃣ Single Account (1 Akun):
Max Worker: 4
Delay: 15 detik antar start
Total: 4 workers
Timeline:
t=0s: Worker 1 START → Generate prompt #1
t=15s: Worker 2 START → Generate prompt #2
t=30s: Worker 3 START → Generate prompt #3
t=45s: Worker 4 START → Generate prompt #4
Kenapa 4 workers?
✅ Aman untuk 1 akun (tidak overload)
✅ Delay 15s mencegah burst request
✅ Performa cukup baik untuk 1 akun
2️⃣ Multiple Accounts (2-3 Akun):
2 Akun:
- Max Worker/Akun: 4
- Total: 8 workers (4×2)
- Delay: 12 detik
3 Akun:
- Max Worker/Akun: 4
- Total: 12 workers (4×3)
- Delay: 7 detik
Worker Distribution:
W1 → Akun A
W2 → Akun B
W3 → Akun C (jika ada)
W4 → Akun A (rotate)
W5 → Akun B (rotate)
...
Kenapa delay lebih kecil?
✅ Load terbagi ke multiple akun
✅ Rate limit per akun lebih aman
✅ Total throughput lebih tinggi
3️⃣ Enterprise Setup (6+ Akun):
6+ Akun:
- Max Worker/Akun: 4
- Total: 24+ workers
- Delay: 3 detik (fastest!)
Performa:
✅ Maximum throughput
✅ Load distribution optimal
✅ Minimal rate limit risk
✅ Production-ready setup
Worker Rotation adalah sistem dimana:
- Worker selesai generate → tunggu cooldown → generate lagi
- Masing-masing worker punya cooldown sendiri (TIDAK tunggu worker lain)
- Workers jalan PARALLEL (bersamaan)
═══════════════════════════════════════════════════════════════════
STAGGERED START PHASE
═══════════════════════════════════════════════════════════════════
t=0s: [Worker-1] START ✅
Status: Generating prompt #1...
Main Thread: Sleep 15 detik...
t=15s: [Worker-2] START ✅
Status: Generating prompt #2...
[Worker-1] masih generating... (parallel)
Main Thread: Sleep 15 detik...
t=30s: [Worker-3] START ✅
Status: Generating prompt #3...
[Worker-1] masih generating... (parallel)
[Worker-2] masih generating... (parallel)
Main Thread: Sleep 15 detik...
t=45s: [Worker-4] START ✅
Status: Generating prompt #4...
[Worker-1] masih generating... (parallel)
[Worker-2] masih generating... (parallel)
[Worker-3] masih generating... (parallel)
═══════════════════════════════════════════════════════════════════
PARALLEL EXECUTION PHASE
═══════════════════════════════════════════════════════════════════
t=60s: [Worker-1] SELESAI ✅ (video #1 done!)
[Worker-1] Action: Tunggu cooldown 15s...
[Worker-2] masih generating...
[Worker-3] masih generating...
[Worker-4] masih generating...
t=70s: [Worker-2] SELESAI ✅ (video #2 done!)
[Worker-2] Action: Tunggu cooldown 15s...
[Worker-1] cooldown... (5s remaining)
[Worker-3] masih generating...
[Worker-4] masih generating...
t=75s: [Worker-1] RESTART! ✅
[Worker-1] Status: Generating prompt #5...
[Worker-2] cooldown... (10s remaining)
[Worker-3] masih generating...
[Worker-4] masih generating...
t=80s: [Worker-3] SELESAI ✅ (video #3 done!)
[Worker-3] Action: Tunggu cooldown 15s...
[Worker-1] generating prompt #5...
[Worker-2] cooldown... (5s remaining)
[Worker-4] masih generating...
t=85s: [Worker-2] RESTART! ✅
[Worker-2] Status: Generating prompt #6...
[Worker-1] masih generating #5...
[Worker-3] cooldown... (10s remaining)
[Worker-4] masih generating...
t=90s: [Worker-4] SELESAI ✅ (video #4 done!)
[Worker-4] Action: Tunggu cooldown 15s...
[Worker-1] masih generating #5...
[Worker-2] masih generating #6...
[Worker-3] cooldown... (5s remaining)
t=95s: [Worker-3] RESTART! ✅
[Worker-3] Status: Generating prompt #7...
[Worker-1] masih generating #5...
[Worker-2] masih generating #6...
[Worker-4] cooldown... (10s remaining)
═══════════════════════════════════════════════════════════════════
CONTINUOUS ROTATION
═══════════════════════════════════════════════════════════════════
Dan seterusnya...
Semua workers continue rotate dengan cooldown masing-masing!
KEY POINTS:
✅ Workers TIDAK tunggu satu sama lain
✅ Masing-masing punya cooldown sendiri
✅ Execution 100% PARALLEL
✅ Maximum throughput!
Time →
0s 15s 30s 45s 60s 75s 90s 105s 120s
W1: |START|=====GENERATE=====|COOL|START|====GEN====|
W2: |START|=====GENERATE=====|COOL|START|====
W3: |START|=====GENERATE=====|COOL|STA
W4: |START|=====GENERATE=====|CO
Legend:
START = Worker dimulai (staggered)
GEN = Generating (parallel)
COOL = Cooldown delay
Setup:
Akun: 1
Workers: 12
Delay: 3 detik
Mode: Manual (salah setting!)
Yang Terjadi:
t=0s: 12 workers START bersamaan (BURST!)
t=0s: W1-W12 request API bersamaan
t=1s: Server Google: "Too many requests!"
Error 429 Rate Limit Exceeded
Generate GAGAL SEMUA! ❌
Kenapa Gagal?
- ❌ Terlalu banyak workers untuk 1 akun
- ❌ Delay terlalu kecil (burst request)
- ❌ Server langsung reject
Solusi:
✅ Gunakan "Formula by Sistem"
✅ Atau manual: 4 workers, delay 15s
Setup:
Akun: 1
Workers: 4 (auto dari formula)
Delay: 15 detik
Mode: Formula by Sistem ✅
Timeline:
[12:00:00] ℹ️ Starting generation (ultra - t2v)
[12:00:00] ℹ️ Prompts: 100, Workers: 4
[12:00:00] ℹ️ Formula Sistem: 1 Akun → 4 Workers, Delay 15s
[12:00:00] ✅ Generation started successfully!
[12:00:00] ℹ️ Worker 1 started, waiting 15s...
[12:00:00] ℹ️ [Worker-1] Starting generation (seed: 7600)
[12:00:07] ℹ️ [Worker-1] Waiting 3s after generate...
[12:00:10] ℹ️ [Worker-1] Checking status... (attempt 1/30)
[12:00:15] ℹ️ Worker 2 started, waiting 15s...
[12:00:15] ℹ️ [Worker-2] Starting generation (seed: 5432)
[12:00:30] ℹ️ Worker 3 started, waiting 15s...
[12:00:30] ℹ️ [Worker-3] Starting generation (seed: 8901)
[12:00:45] ℹ️ Worker 4 started
[12:00:45] ℹ️ [Worker-4] Starting generation (seed: 2345)
[12:02:30] ✅ [Worker-1] Video saved: prompt_001.mp4
[12:02:30] ℹ️ [Worker-1] Cooldown 15s before restart...
[12:02:45] ℹ️ [Worker-1] Starting generation (seed: 6789)
... (continue rotation)
[12:45:30] ✅ Generation completed!
Total: 100, Success: 100, Failed: 0
Duration: 45 menit 30 detik
Kenapa Sukses?
- ✅ 4 workers optimal untuk 1 akun
- ✅ Delay 15s mencegah burst
- ✅ Parallel execution after staggered start
- ✅ Tidak ada rate limit error!
Performa:
100 prompts ÷ 4 workers = 25 prompts per worker
Avg time per video: ~6-8 menit
Total time: ~45 menit
Throughput: ~2.2 video per menit
Setup:
Akun: 3
Workers: 12 (auto dari formula)
Delay: 7 detik
Mode: Formula by Sistem ✅
Account Distribution:
Worker 1 → Akun A (akun1@example.com)
Worker 2 → Akun B (akun2@example.com)
Worker 3 → Akun C (akun3@example.com)
Worker 4 → Akun A (rotate)
Worker 5 → Akun B (rotate)
Worker 6 → Akun C (rotate)
Worker 7 → Akun A (rotate)
Worker 8 → Akun B (rotate)
Worker 9 → Akun C (rotate)
Worker 10 → Akun A (rotate)
Worker 11 → Akun B (rotate)
Worker 12 → Akun C (rotate)
Result: 4 workers per akun ✅
Timeline:
[12:00:00] ℹ️ Starting generation (ultra - t2v)
[12:00:00] ℹ️ Prompts: 100, Workers: 12
[12:00:00] ℹ️ Formula Sistem: 3 Akun → 4 Workers per akun, Delay 7s
[12:00:00] ℹ️ Loaded 3 Google Ultra account(s)
[12:00:00] ✅ Generation started successfully!
[12:00:00] ℹ️ Worker 1 started (Akun A), waiting 7s...
[12:00:07] ℹ️ Worker 2 started (Akun B), waiting 7s...
[12:00:14] ℹ️ Worker 3 started (Akun C), waiting 7s...
[12:00:21] ℹ️ Worker 4 started (Akun A), waiting 7s...
[12:00:28] ℹ️ Worker 5 started (Akun B), waiting 7s...
[12:00:35] ℹ️ Worker 6 started (Akun C), waiting 7s...
[12:00:42] ℹ️ Worker 7 started (Akun A), waiting 7s...
[12:00:49] ℹ️ Worker 8 started (Akun B), waiting 7s...
[12:00:56] ℹ️ Worker 9 started (Akun C), waiting 7s...
[12:01:03] ℹ️ Worker 10 started (Akun A), waiting 7s...
[12:01:10] ℹ️ Worker 11 started (Akun B), waiting 7s...
[12:01:17] ℹ️ Worker 12 started (Akun C)
[12:01:17] ✅ All 12 workers running in PARALLEL!
... (workers generating in parallel)
[12:16:30] ✅ Generation completed!
Total: 100, Success: 100, Failed: 0
Duration: 16 menit 30 detik
Kenapa Super Cepat?
- ✅ 12 workers PARALLEL (3× faster than 1 akun)
- ✅ Load terbagi ke 3 akun (4 workers per akun aman)
- ✅ Delay lebih kecil (7s) karena multiple akun
- ✅ Maximum throughput!
Performa:
100 prompts ÷ 12 workers = 8-9 prompts per worker
Avg time per video: ~6-8 menit
Total time: ~16 menit (3× faster!)
Throughput: ~6 video per menit 🚀
Setup:
Akun: 10
Workers: 40 (manual: 4 per akun)
Delay: 3 detik (fastest!)
Mode: Manual (advanced user)
Account Distribution:
40 workers ÷ 10 akun = 4 workers per akun ✅
Load per akun: SANGAT RINGAN
Rate limit risk: MINIMAL
Timeline:
[12:00:00] ℹ️ Starting generation (ultra - t2v)
[12:00:00] ℹ️ Prompts: 1000, Workers: 40
[12:00:00] ℹ️ Manual mode: 10 Akun → 4 Workers per akun, Delay 3s
[12:00:00] ✅ Generation started successfully!
[12:00:00-12:02:00] Starting 40 workers dengan delay 3s antar start
[12:02:00] ✅ All 40 workers running in PARALLEL!
... (massive parallel processing)
[12:14:30] ✅ Generation completed!
Total: 1000, Success: 1000, Failed: 0
Duration: 14 menit 30 detik
Performa INSANE:
1000 prompts ÷ 40 workers = 25 prompts per worker
Avg time per video: ~6-8 menit
Total time: ~14 menit
Throughput: ~69 video per menit! 💎
Use Case:
- ✅ Production content agency
- ✅ Mass video generation
- ✅ Enterprise deployment
- ✅ Time-critical projects
Setup (SALAH TOTAL):
Akun: 1
Workers: 35 (max system)
Delay: 0 detik (no delay!)
Mode: Manual (salah parah!)
Yang Terjadi:
[12:00:00] ℹ️ Starting generation (ultra - t2v)
[12:00:00] ℹ️ Prompts: 100, Workers: 35
[12:00:00] ❌ BURST REQUEST 35 WORKERS BERSAMAAN!
[12:00:01] ❌ [Worker-1] Error 429 Rate Limit Exceeded
[12:00:01] ❌ [Worker-2] Error 429 Rate Limit Exceeded
[12:00:01] ❌ [Worker-3] Error 429 Rate Limit Exceeded
... (30 more errors)
[12:00:05] ❌ [Worker-4] Error 429 Rate Limit Exceeded
[12:00:05] ❌ [Worker-5] Error 429 Rate Limit Exceeded
[12:00:10] ⚠️ Semua workers kena rate limit!
[12:00:10] ⚠️ Retry 3×... GAGAL SEMUA!
[12:05:00] ❌ Generation completed!
Total: 100, Success: 0, Failed: 100
Duration: 5 menit
TOTAL DISASTER! ☠️
Kenapa Gagal Total?
- ❌ 35 workers untuk 1 akun = OVERLOAD PARAH
- ❌ Delay 0 detik = BURST REQUEST (semua bersamaan)
- ❌ Server langsung block semua request
- ❌ Tidak ada yang berhasil!
Lesson Learned:
🚫 JANGAN PERNAH:
- Set workers > 6 untuk 1 akun
- Set delay < 5 detik untuk single akun
- Disable rate limiting
- Ignore warning messages
Tampilan Error:
❌ [Worker-3] Error 429 Rate Limit Exceeded
❌ Too many requests from this account
❌ Please wait and try again
Arti: Anda mengirim terlalu banyak request dalam waktu singkat. Server menolak request untuk melindungi sistemnya.
Penyebab:
- Workers terlalu banyak untuk jumlah akun
- Delay antar worker terlalu kecil
- Burst request (semua worker start bersamaan)
Solusi:
A. Segera (Emergency):
1. Klik tombol STOP ⏹️
2. Tunggu 2-3 menit (cooldown server)
3. Aktifkan "Formula by Sistem"
4. Klik START lagi
B. Setting Manual:
Jika punya:
- 1 akun → Max 4 workers, delay 15s
- 2 akun → Max 8 workers, delay 12s
- 3 akun → Max 12 workers, delay 7s
- 4+ akun → Max 20 workers, delay 5s
C. Tambah Akun (Best Solution):
Rate limit = per akun
1 akun:
- Max safe: 4-6 workers
- Throughput: ~2-3 video/menit
3 akun:
- Max safe: 12-18 workers
- Throughput: ~6-9 video/menit (3× faster!)
💡 Tambah akun = Throughput naik linear!
Tampilan Error:
❌ [Worker-1] Error 401 Unauthorized
❌ Cookie expired or invalid
❌ Please refresh account
Arti: Cookie akun Anda sudah expired (kadaluarsa). Perlu login ulang.
Penyebab:
- Cookie expired (biasanya 7-30 hari)
- Session timeout
- Google/GenAI logout otomatis
Solusi:
1. Pergi ke tab "Accounts"
2. Cari akun yang error
3. Klik tombol 🔄 Refresh Cookie
4. Chrome akan terbuka otomatis
5. Login ulang (jika perlu)
6. Cookie akan update otomatis
7. Kembali ke tab generation
8. Klik START lagi
Pencegahan:
✅ Refresh cookie setiap 2-3 minggu
✅ Jangan logout dari browser
✅ Gunakan "Remember Me" saat login
Tampilan Error:
⏱️ [Worker-2] Generation timeout (exceeded 15 minutes)
❌ Request took too long
❌ Moving to next prompt
Arti: Generate video memakan waktu terlalu lama (> 15 menit). Kemungkinan server overload atau prompt terlalu complex.
Penyebab:
- Server Google/GenAI sedang lambat
- Prompt terlalu kompleks (heavy rendering)
- Peak hours (jam sibuk)
Solusi:
✅ Wait & Retry: Biasanya prompt selanjutnya berhasil
✅ Simplify Prompt: Kurangi detail yang terlalu complex
✅ Off-Peak Hours: Generate di malam hari (sepi server)
✅ Check Failed Prompts: Review prompts_failed.txt
Tidak Perlu Panik:
- ✅ Worker otomatis lanjut ke prompt berikutnya
- ✅ Timeout prompt disave di "failed" folder
- ✅ Bisa retry manual nanti
Tampilan Error:
❌ [Worker-4] Connection error
❌ Failed to connect to server
❌ Check your internet connection
Arti: Tidak bisa connect ke server Google/GenAI. Problem network/internet.
Penyebab:
- Internet mati/putus
- DNS problem
- Firewall blocking
- VPN issue
Solusi:
1. Check Internet:
- Buka browser → Google.com
- Pastikan internet normal
2. Restart Network:
- Matikan WiFi → Nyalakan lagi
- Atau restart modem
3. Check Firewall:
- Windows Defender → Allow app
- Antivirus → Add exception
4. Retry:
- Klik STOP
- Tunggu 10 detik
- Klik START lagi
Update:
✅ Error ini TIDAK MUNGKIN terjadi di v3.06 Beta!
✅ Anti-Collision Queue sudah DISABLED
✅ No more queue blocking!
Jika Anda masih lihat error ini:
→ Update aplikasi ke v3.06 Beta latest!
Rata-rata waktu per video: 6-8 menit
- Generate: 6-7 menit (Google processing)
- Download: 30-60 detik
- Upscale (optional): +2-3 menit
- Mute (optional): +5-10 detik
| Setup | Workers | Time/100 Prompts | Throughput |
|---|---|---|---|
| 1 akun | 4 | ~45 menit | ~2.2 video/min |
| 2 akun | 8 | ~25 menit | ~4 video/min |
| 3 akun | 12 | ~16 menit | ~6 video/min |
| 5 akun | 20 | ~12 menit | ~8.3 video/min |
| 10 akun | 40 | ~8 menit | ~12.5 video/min |
Contoh Real:
1 Akun + 4 Workers + 1000 Prompts:
= 1000 ÷ 4 = 250 prompts per worker
= 250 × 7 menit = 1750 menit
= ~29 jam (non-stop)
= ~1.2 hari
3 Akun + 12 Workers + 1000 Prompts:
= 1000 ÷ 12 = 84 prompts per worker
= 84 × 7 menit = 588 menit
= ~9.8 jam (non-stop)
= ~10 jam total (3× faster!)
Rata-rata waktu per image: 30-60 detik
- Generate: 20-40 detik
- Download: 5-10 detik
- Upscale (optional): +15-20 detik
| Setup | Workers | Time/100 Prompts | Throughput |
|---|---|---|---|
| 1 akun | 4 | ~8 menit | ~12.5 img/min |
| 2 akun | 8 | ~5 menit | ~20 img/min |
| 3 akun | 12 | ~3.5 menit | ~28 img/min |
| 5 akun | 20 | ~2.5 menit | ~40 img/min |
| 10 akun | 40 | ~1.5 menit | ~66 img/min |
Contoh Real:
1 Akun + 4 Workers + 1000 Images:
= 1000 ÷ 4 = 250 images per worker
= 250 × 45 detik = 11250 detik
= ~187 menit = ~3.1 jam
3 Akun + 12 Workers + 1000 Images:
= 1000 ÷ 12 = 84 images per worker
= 84 × 45 detik = 3780 detik
= ~63 menit = ~1 jam (3× faster!)
A. Tambah Akun Google Ultra atau GenAI:
1. Buka aplikasi Veogen
2. Klik tab "Accounts"
3. Klik "Add Data" untuk provider yang diinginkan
4. Chrome akan terbuka otomatis
5. Login dengan akun Google/GenAI Anda
6. Cookie akan diambil otomatis
7. Akun akan muncul di daftar
Tips:
✅ Gunakan akun berbeda (jangan sama)
✅ Recommended: 3-5 akun untuk performa optimal
✅ Bisa mix: 2 GenAI + 3 Ultra (total 5)
B. Verifikasi Akun:
Di tab "Accounts", pastikan:
✅ Status: Active (hijau)
✅ Cookie: Valid (tidak expired)
✅ Last Used: Recent
Jika ada yang merah:
❌ Klik 🔄 Refresh Cookie
A. Provider:
GenAI Pro:
- Model: Veo 3.1 (Google AI Studio)
- Speed: Fast
- Quality: Excellent
- Best for: Professional video
Google Ultra:
- Model: Veo 3.1 (Google AI Console)
- Speed: Fast
- Quality: Excellent
- Best for: Alternative account source
B. Mode:
T2V (Text-to-Video):
- Input: Text prompts
- Output: Video dari teks
- Best for: Storyboarding, concepts
I2V (Image-to-Video):
- Input: Images + prompts (optional)
- Output: Video dari gambar
- Best for: Animation, product demo
A. Untuk Pemula (Recommended):
1. Scroll ke section "🤖 Pengaturan Rate Limiting"
2. Pilih radio button: "Formula by Sistem"
3. Done! Sistem akan auto-detect & set optimal
B. Untuk Advanced User:
1. Pilih radio button: "Setting Your Rate Limiting"
2. Settings panel akan muncul
3. Atur:
- Max Worker Per Akun: 4 (recommended)
- Delay Antar Worker: 15 detik
- Delay After Generate: 3 detik
- Delay After Status: 3 detik
- Delay After Upload: 3 detik
- Delay After Upscale: 3 detik
4. Info box akan show:
"Terdeteksi: X akun | Total Workers: Y"
Settings Explanation:
| Setting | Fungsi | Recommended |
|---|---|---|
| Max Worker Per Akun | Jumlah workers per akun | 4 |
| Delay Antar Worker | Jeda start antar worker (staggered) | 10-20 detik |
| Delay After Generate | Jeda setelah request generate | 3 detik |
| Delay After Status | Jeda setelah cek status | 3 detik |
| Delay After Upload | Jeda setelah upload image (I2V) | 3 detik |
| Delay After Upscale | Jeda setelah upscale 4K | 3 detik |
A. Untuk Batch Video Generation:
1. Buat file TXT dengan prompts (satu prompt per baris)
2. Contoh prompts.txt:
A cinematic shot of a cat walking in rain
Epic drone view of mountain sunset
Close-up of coffee being poured
3. Di aplikasi, klik "Load Prompts"
4. Pilih file prompts.txt
5. Toast muncul: "Loaded 3 prompts from prompts.txt"
B. Untuk Image-to-Video (I2V):
1. Siapkan folder images:
my_images/
├── image_001.jpg
├── image_002.jpg
├── image_003.png
└── image_004.jpg
2. Di aplikasi, pilih mode "I2V"
3. Klik "Browse" folder images
4. Pilih folder my_images
5. Images akan load otomatis
6. Info: "Loaded 4 images"
7. Pilih prompt mode:
- Global: Satu prompt untuk semua images
- Multi: Beda prompt per image (load TXT)
A. Aspect Ratio:
- Landscape (16:9): Default, widescreen
- Portrait (9:16): TikTok, Instagram Reels
- Square (1:1): Instagram Post
B. Video Model:
- Fast: Prioritas normal (recommended)
- Lower Priority: Lebih lambat, hemat quota
C. Auto Features:
✅ Auto-Mute:
- Hilangkan audio otomatis
- Requires FFmpeg installed
✅ Auto-Upscale 1080p:
- Upscale ke 1080p (free)
- Topaz AI (if enabled)
✅ Auto-Upscale 4K:
- Upscale ke 4K (premium)
- Requires Topaz API key
- Disable 1080p otomatis
D. Output Directory:
Default: video_download/
Struktur folder otomatis:
video_download/
├── sukses/
│ ├── prompt_001.mp4
│ ├── prompt_002.mp4
│ └── ...
├── failed/
│ └── (failed videos)
└── TrackingPrompt/
└── batch_20251126_120000/
├── sukses/
│ └── prompts_success.txt
├── gagal/
│ └── prompts_failed.txt
└── sisa/
└── prompts_remaining.txt
A. Final Check:
✅ Akun: Terdeteksi & valid
✅ Prompts: Loaded (atau images loaded)
✅ Rate Limiting: Configured
✅ Settings: Reviewed
✅ Output Dir: Set
B. Klik START Button:
[12:00:00] ✅ Application started successfully
[12:00:00] ℹ️ Ready to generate videos
[12:00:00] ✅ Rate limiting: Staggered start enabled (v3.06 Beta)
[12:00:00] ℹ️ Starting generation (ultra - t2v)
[12:00:00] ℹ️ Prompts: 100, Workers: 12
[12:00:00] ℹ️ Formula Sistem: 3 Akun → 4 Workers per akun
[12:00:00] ✅ Generation started successfully!
[12:00:00] ℹ️ Worker 1 started, waiting 7s...
[12:00:07] ℹ️ Worker 2 started, waiting 7s...
...
C. Monitor Progress:
Tab kanan akan show:
┌─────────────────────────────┐
│ 📊 Statistics │
├─────────────────────────────┤
│ Active Workers: 12 │
│ Completed: 45/100 (45%) │
│ Failed: 2 │
│ Remaining: 55 │
│ │
│ Progress: [████████░░] 45% │
└─────────────────────────────┘
Log Activity:
✅ [Worker-1] Video saved: prompt_001.mp4
✅ [Worker-2] Video saved: prompt_002.mp4
⏳ [Worker-3] Checking status... (attempt 3/30)
...
D. Controls:
⏸️ PAUSE: Pause semua workers (temporary)
⏹️ STOP: Stop generation (permanent)
🔄 RESUME: Continue dari pause
A. Check Output Folder:
video_download/sukses/
✅ prompt_001.mp4 (sukses)
✅ prompt_002.mp4 (sukses)
✅ prompt_003.mp4 (sukses)
...
video_download/failed/
❌ prompt_023.mp4 (jika ada yang gagal)
B. Check Tracking:
TrackingPrompt/batch_20251126_120000/sukses/prompts_success.txt:
A cinematic shot of a cat walking in rain
Epic drone view of mountain sunset
Close-up of coffee being poured
...
TrackingPrompt/batch_20251126_120000/gagal/prompts_failed.txt:
(prompts yang gagal - bisa retry manual)
C. Statistics:
Final Report:
✅ Total: 100
✅ Success: 98
❌ Failed: 2
⏱️ Duration: 16 menit 30 detik
🚀 Throughput: 6 video/menit
Success Rate: 98% ✅
A. Manual Retry:
1. Buka prompts_failed.txt
2. Copy prompts yang gagal
3. Paste ke prompt box di aplikasi
4. Klik START (retry with same settings)
B. Adjust & Retry:
Jika banyak yang failed:
1. Kurangi workers (e.g., 12 → 8)
2. Tambah delay (e.g., 7s → 10s)
3. Retry failed prompts
Print checklist ini sebelum batch generation besar:
□ AKUN
□ Minimal 1 akun sudah ditambahkan
□ Status akun: Active (hijau)
□ Cookie valid (tidak expired)
□ Jika multiple akun, pastikan BEDA akun (bukan duplikat)
□ PROMPTS/IMAGES
□ File prompts.txt sudah disiapkan (T2V)
□ Atau folder images sudah ready (I2V)
□ Prompts sudah di-load ke aplikasi
□ Jumlah prompts sesuai (check info box)
□ RATE LIMITING
□ Mode dipilih: Formula by Sistem atau Manual
□ Jika Manual: Settings sudah diatur optimal
□ Info box show: "Terdeteksi: X akun | Total Workers: Y"
□ Total workers masuk akal (tidak terlalu besar)
□ SETTINGS
□ Provider dipilih (GenAI/Ultra)
□ Mode dipilih (T2V/I2V)
□ Aspect ratio sesuai kebutuhan
□ Auto-mute/upscale sesuai kebutuhan
□ Output directory set
□ SYSTEM
□ Internet connection stabil
□ Disk space cukup (min 10GB per 100 video)
□ Aplikasi running smooth (no lag)
□ FFmpeg installed (if auto-mute enabled)
□ Topaz API key (if 4K upscale enabled)
□ FINAL CHECK
□ Semua checkbox di atas ✅
□ Siap monitor progress
□ Tidak akan close aplikasi saat generate
□ PC tidak akan sleep/hibernate
✅ ALL CLEAR! Klik START! 🚀
Recommended Account Strategy:
Pemula (Budget: $0):
- 1-2 akun gratis
- Workers: 4-8
- Throughput: 2-4 video/min
- Good for: Testing, small batches
Intermediate (Budget: Moderate):
- 3-5 akun
- Workers: 12-20
- Throughput: 6-8 video/min
- Good for: Regular production
Professional (Budget: High):
- 6-10 akun
- Workers: 24-40
- Throughput: 12-16 video/min
- Good for: Agency, mass production
Enterprise (Budget: Corporate):
- 10+ akun
- Workers: 40+
- Throughput: 16+ video/min
- Good for: Enterprise deployment
Server load berpengaruh pada speed:
Peak Hours (Lambat):
🔴 08:00-12:00 WIB (Pagi)
🔴 13:00-17:00 WIB (Siang)
🔴 19:00-22:00 WIB (Malam)
Avg time: 8-10 menit per video
Rate limit risk: HIGHER
Off-Peak Hours (Cepat):
✅ 00:00-06:00 WIB (Dini hari)
✅ 12:00-13:00 WIB (Lunch break)
✅ 23:00-00:00 WIB (Late night)
Avg time: 5-7 menit per video
Rate limit risk: LOWER
Pro Tip:
Generate di malam hari (00:00-06:00):
- 30-40% lebih cepat
- Rate limit lebih longgar
- Bisa set workers lebih tinggi
Jangan generate semuanya sekaligus!
Small Batch First (Testing):
1. Test batch: 10-20 prompts
2. Check results quality
3. Check error rate
4. Adjust settings if needed
5. Then scale up!
Production Batch:
Pemula: 50-100 prompts per batch
Intermediate: 100-300 prompts per batch
Professional: 300-1000 prompts per batch
Split besar batch:
1000 prompts → 10× batch @100
- Safer (jika error, tidak semua gagal)
- Easier to monitor
- Can adjust per batch
Good Prompts = Higher Success Rate
Good Prompt Example:
✅ "A cinematic shot of a red sports car driving through a winding mountain road at sunset, camera following from behind"
Why good:
- Clear subject (red sports car)
- Clear action (driving)
- Clear location (mountain road)
- Clear time (sunset)
- Clear camera (following from behind)
Bad Prompt Example:
❌ "car"
Why bad:
- Too vague
- No details
- AI confused
- Higher fail rate
Pro Tips:
✅ Be specific (colors, actions, camera angles)
✅ Use cinematic terms (drone shot, close-up, etc.)
✅ 10-30 words optimal length
✅ Avoid contradictions ("day and night")
✅ Test prompts individually first
Watch These Metrics:
✅ Active Workers: Should match settings
✅ Success Rate: Should be > 90%
✅ Failed Count: Should be minimal
✅ Progress %: Should increase steady
Red Flags:
❌ Workers stuck (tidak ada progress)
❌ Success rate < 70%
❌ Banyak error 429
❌ Connection errors bertubi-tubi
What To Do:
If Success Rate < 90%:
1. Click STOP
2. Review failed prompts log
3. Check error patterns
4. Adjust settings
5. Restart with smaller batch
If Error 429 Multiple Times:
1. Click STOP immediately
2. Wait 5 minutes (cooldown)
3. Reduce workers by 30-50%
4. Increase delays
5. Restart
Gejala:
[12:00:00] ✅ Generation started successfully!
(tidak ada log worker...)
Cek:
□ Prompts sudah diload? (check prompt box)
□ Images sudah diload? (I2V mode)
□ Akun ada & valid?
□ Internet connection OK?
Solusi:
1. Reload prompts/images
2. Check accounts tab
3. Restart aplikasi
4. Try again
Gejala:
❌ [Worker-1] Error 429
❌ [Worker-2] Error 429
❌ [Worker-3] Error 429
(semua workers kena)
Solusi Immediate:
1. STOP generation
2. Wait 5 minutes
3. Pilih "Formula by Sistem"
4. START lagi
Jika masih error:
1. Kurangi Max Worker Per Akun: 4 → 3
2. Naikkan Delay: 15s → 20s
3. Try again
Gejala:
1 video = 15-20 menit (harusnya 6-8 menit)
Kemungkinan:
□ Peak hours (server overload)
□ Internet lambat
□ Workers terlalu sedikit
□ Prompt terlalu complex
Solusi:
1. Check internet speed (speedtest.net)
2. Generate di off-peak hours (malam)
3. Tambah akun → tambah workers
4. Simplify prompts (kurangi detail)
Gejala:
✅ Success: 60
❌ Failed: 40
(40% fail rate!)
Analisa:
1. Buka prompts_failed.txt
2. Cek pattern error:
- Semua timeout? → Server lambat
- Semua error 429? → Rate limit
- Semua error 401? → Cookie expired
- Random? → Prompts issue
Solusi:
Pattern Timeout:
→ Generate di off-peak hours
→ Increase timeout setting
Pattern 429:
→ Reduce workers
→ Increase delays
Pattern 401:
→ Refresh cookie semua akun
→ Retry failed
Random:
→ Check prompt quality
→ Remove problematic prompts
→ Retry batch
Gejala:
Aplikasi not responding
Workers stuck
No log updates
Emergency Steps:
1. Task Manager → End Process
2. Restart aplikasi
3. Check logs:
- TrackingPrompt/batch_xxx/sisa/
- Ada remaining prompts?
4. Continue dari remaining prompts
Pencegahan:
✅ Jangan set workers > 35
✅ Jangan minimize saat generate
✅ Disk space cukup (min 20GB free)
✅ Close aplikasi lain (hemat RAM)
Gejala:
❌ All workers: Error 401 Unauthorized
Solusi Mass Refresh:
1. Tab "Accounts"
2. Klik 🔄 untuk akun pertama
3. Login → Cookie updated
4. Ulangi untuk semua akun
5. Verifikasi semua status: Active
6. Back to generation tab
7. START lagi
Scheduling:
Set reminder:
□ Refresh cookie setiap 2 minggu
□ Check status sebelum batch besar
□ Keep browser logged in
-
Sistem v3.06 Beta = Staggered Start + Parallel Execution
- Workers start bertahap (delay 10-20s)
- Setelah start, semua jalan parallel
- Tidak ada queue blocking
- Maximum throughput!
-
2 Mode Sederhana:
- Formula by Sistem (auto, recommended)
- Setting Your Rate Limiting (manual, advanced)
-
Golden Rule:
- 1 akun = max 4-6 workers
- Multiple akun = linear scalability
- Delay 10-20 detik antar start
- Tambah akun = fastest way to scale
-
Error Prevention:
- Use "Formula by Sistem"
- Don't overload (respect rate limit)
- Generate di off-peak hours
- Test small batch first
-
Monitoring:
- Watch success rate (should be > 90%)
- Stop immediately if error 429 multiple
- Review logs regularly
- Track remaining prompts
Documentation:
- README.md (main documentation)
- CHANGELOG.md (version history)
- This file (rate limiting guide)
Troubleshooting:
- Check logs: TrackingPrompt/
- Review error messages
- Use troubleshooting section above
Best Practices:
- Start small, scale gradually
- Use Formula by Sistem first
- Monitor first batch carefully
- Keep cookies fresh
- Generate at off-peak hours
Developed with ❤️
🎉 Selamat Generate! Semoga Sukses! 🚀
Last Updated: 2025-11-26 Version: 3.06 Beta Document: RUMUS_WORKER.md