Compute Tiers - Replov Docs
Skip to main content

Compute Tiers

Choose the right compute resources for your workload

Every service runs on one of five compute tiers, billed per service from your wallet (the first two Nano Static Sites are free). Micro and above are billed from your wallet by the hour or the month. Change tiers at any time.

Available Tiers

Nano

Perfect for testing and small projects

₹199
Nano / 30 days
Memory
512 MB
CPU
0.5 vCPU
Storage
3 GB
Bandwidth
10 GB

Best For:

  • • Development and testing
  • • Small static sites
  • • Lightweight APIs
  • • Learning and experimentation
Nano: ₹199 per service for 30 days; the first two Nano Static Sites per account are free

Micro

Popular
₹0.83
/hour
Memory
1 GB
CPU
1 vCPU
Storage
10 GB
Bandwidth
100 GB

Best For:

  • • Small production apps
  • • REST APIs
  • • Background workers
  • • Simple AI agents
Monthly Cost (24/7): ₹599/month on monthly billing

Small

Standard production workloads

₹1.67
/hour
Memory
2 GB
CPU
2 vCPU
Storage
25 GB
Bandwidth
500 GB

Best For:

  • • Production web apps
  • • Database-backed APIs
  • • Complex AI agents
  • • Medium traffic sites
Monthly Cost (24/7): ₹1,199/month on monthly billing

Medium

High-traffic applications

₹3.33
/hour
Memory
4 GB
CPU
4 vCPU
Storage
50 GB
Bandwidth
1000 GB

Best For:

  • • High-traffic web apps
  • • Data processing
  • • Multiple services
  • • Resource-intensive agents
Monthly Cost (24/7): ₹2,399/month on monthly billing

Large

Enterprise-grade performance

₹6.94
/hour
Memory
8 GB
CPU
8 vCPU
Storage
100 GB
Bandwidth
5000 GB

Best For:

  • • Enterprise applications
  • • Heavy data processing
  • • High-concurrency APIs
  • • Mission-critical services
Monthly Cost (24/7): ₹4,999/month on monthly billing

Which Tier Fits

TierMemoryCPUStorageBandwidthPrice/Hour
Nano512 MB0.5 vCPU3 GB10 GB₹199 / 30 days
Micro1 GB1 vCPU10 GB100 GB₹0.83
Small2 GB2 vCPU25 GB500 GB₹1.67
Medium4 GB4 vCPU50 GB1000 GB₹3.33
Large8 GB8 vCPU100 GB5000 GB₹6.94

Choosing the Right Tier

Consider Your Workload

  • • Memory-intensive: Choose based on RAM needs (databases, caching)
  • • CPU-intensive: Choose based on vCPU count (data processing, AI)
  • • I/O-intensive: Consider storage and bandwidth limits

Start Small, Scale Up

Begin with a smaller tier and monitor resource usage. Upgrade when you consistently hit 70-80% utilization. You can change tiers at any time; the service restarts with the new limits, so expect a brief interruption.

Monitor Resource Usage

Check your service dashboard for real-time metrics:

  • • CPU usage percentage
  • • Memory consumption
  • • Storage used
  • • Bandwidth consumed

Upgrading or Downgrading

1

Go to Service Settings

Navigate to your service → Settings → Compute Tier

2

Select New Tier

Choose your desired tier from the dropdown

3

Apply Changes

Click "Update Tier", then monitor the service while Docker applies the new limits

Plan changes may restart the service
Replov updates Docker resource constraints when the tier changes. A brief interruption is possible, so schedule production changes and verify the service afterward.

Cost Optimization Tips

1. Use Hourly Billing Wisely

Stop services when not needed (development/staging). You're only charged for running hours.

text
Example: a development service on hourly billing
- Runs 8 hours a day, 5 days a week: about 160 hours a month
- A full month is 730 hours
- So it costs about 22% of running the same tier 24/7

2. Right-Size Your Services

Don't over-provision. Monitor usage and downgrade if consistently under 50% utilization.

3. Optimize Your Code

Efficient code uses fewer resources. Profile and optimize hot paths to reduce tier requirements.

4. Use Caching

Implement caching (Redis, in-memory) to reduce CPU and database load, potentially allowing a smaller tier.

Resource Limits

  • Memory: Hard limit. Service is killed if exceeded. Choose appropriate tier.
  • CPU: Throttled if exceeded. Performance degrades but service continues.
  • Storage: Soft limit. Warning at 80%, service suspended at 100%.
  • Bandwidth: Soft limit. Overage charges apply beyond tier limit.
Memory Limits
Memory limits are strictly enforced. If your service exceeds its memory limit, it will be killed and restarted. Monitor memory usage closely.

Next Steps