TH10 ‐ Difference between hot‐hot, hot‐warm, hot‐cold - DevWithNehaCloud/Oracle-Cloud-Infrastructure-wiki GitHub Wiki

Difference Between Hot-Hot, Hot-Warm, and Hot-Cold Architectures

Overview

Hot-Hot, Hot-Warm, and Hot-Cold are Disaster Recovery (DR) and High Availability (HA) architectures used to ensure application availability during failures.

The main difference lies in how active the secondary environment is and how quickly services can be restored after a disaster.

1. Hot-Hot Architecture (Active-Active)

Definition

In a Hot-Hot architecture, both primary and secondary environments are fully active and serving user traffic simultaneously.

If one environment fails, the other continues serving requests with minimal or near-zero downtime.

Characteristics

  • Both sites are active.
  • Traffic is distributed across both sites.
  • Data is continuously synchronized.
  • Automatic failover is generally used.
  • Minimal or near-zero downtime.

Example

Suppose an e-commerce application is deployed in:

  • OCI Mumbai Region
  • OCI Hyderabad Region

A Global Load Balancer distributes traffic between both regions.

Users
   |
Global Load Balancer
   |
-------------------------
|                       |
Mumbai Region      Hyderabad Region
(Active)            (Active)

If the Mumbai region becomes unavailable, users can be redirected to Hyderabad automatically.

Advantages

  • Very high availability.
  • Near-zero downtime.
  • Excellent user experience.
  • Fast and automated failover.

Disadvantages

  • Most expensive architecture.
  • Complex implementation.
  • Continuous data synchronization is required.
  • Both environments need to be maintained and monitored.

Typical RTO and RPO

Metric Value
RTO Near Zero
RPO Near Zero

2. Hot-Warm Architecture (Active-Passive Ready)

Definition

In a Hot-Warm architecture, the primary site is fully active, while the secondary site is partially active and ready to take over when needed.

The standby environment is already provisioned but may run with reduced capacity.

Characteristics

  • Primary site handles production traffic.
  • Secondary site remains ready for disaster recovery.
  • Data replication is ongoing.
  • Failover can be manual or automated.
  • Lower cost than Hot-Hot.

Example

An application runs in:

  • Mumbai Region — Production
  • Hyderabad Region — Standby

The Hyderabad environment may contain:

  • VCN
  • Subnets
  • Compute instances
  • Database replicas

However, users are not actively connected to Hyderabad during normal operations.

Users
   |
Mumbai Region
(Active)
   |
Replication
   |
Hyderabad Region
(Warm Standby)

If Mumbai fails, traffic is switched to Hyderabad.

Advantages

  • Lower cost than Hot-Hot.
  • Faster recovery than Hot-Cold.
  • Good balance between cost and availability.
  • Secondary infrastructure is already provisioned.

Disadvantages

  • Some failover time exists.
  • Secondary resources may be underutilized.
  • May require additional steps during failover.
  • Standby capacity may be lower than the primary environment.

Typical RTO and RPO

Metric Value
RTO Minutes
RPO Minutes

3. Hot-Cold Architecture (Backup Site)

Definition

In a Hot-Cold architecture, the primary site is active, while the secondary site mainly contains backups and/or basic infrastructure.

Production services are not normally running in the DR site.

If a disaster occurs, resources must be restored or provisioned before services can become available.

Characteristics

  • Only the primary site is active.
  • Backup data is stored in the DR location.
  • Compute resources may not exist in the DR environment.
  • Recovery is usually manual or requires significant recovery steps.
  • Lowest-cost option among these three architectures.

Example

Primary environment:

  • Mumbai Region

DR environment:

  • Hyderabad Region
  • Object Storage backups
  • Database backups

No application servers are running in Hyderabad during normal operations.

Users
   |
Mumbai Region
(Active)
   |
Backups
   |
Hyderabad Region
(Cold Site)

If Mumbai fails:

  1. Restore backups.
  2. Provision or launch compute instances.
  3. Recover databases.
  4. Configure required networking and security.
  5. Start application services.
  6. Redirect users to the DR environment.

Only after these recovery steps can users access the application.

Advantages

  • Lowest cost.
  • Simple to maintain during normal operations.
  • Suitable for workloads with less stringent recovery requirements.
  • Minimal infrastructure cost for the DR environment.

Disadvantages

  • Long recovery time.
  • Significant downtime is possible.
  • More manual recovery steps.
  • Infrastructure and services may need to be provisioned or restored during a disaster.

Typical RTO and RPO

Metric Value
RTO Hours to Days
RPO Hours

Comparison Table

Feature Hot-Hot Hot-Warm Hot-Cold
Secondary Site Status Active Standby Ready Backup / Recovery Site
User Traffic Both Sites Primary Only Primary Only
Failover Time Near Zero Minutes Hours/Days
Cost Highest Medium Lowest
Complexity High Medium Low
Data Replication Continuous Continuous / Frequent Scheduled / Backup Based
Availability Very High High Lower
RTO Near Zero Minutes Hours/Days
RPO Near Zero Minutes Hours

Simple Way to Remember

Hot-Hot  → Both are Active
Hot-Warm → One is Active, One is Ready
Hot-Cold → One is Active, One is Backup/Recovery

Easy Example

Hot-Hot

Mumbai      → Active
Hyderabad   → Active
                ↓
            Both serve traffic


Hot-Warm

Mumbai      → Active
Hyderabad   → Standby
                ↓
          Takes over after failure


Hot-Cold

Mumbai      → Active
Hyderabad   → Backups
                ↓
          Restore/Provision first

OCI Interview Answer (1–2 Minutes)

In a Hot-Hot architecture, both primary and secondary environments are active and serving traffic simultaneously. This provides very low recovery time and can provide near-zero downtime when properly designed.

In a Hot-Warm architecture, the primary environment is active while the secondary environment is already provisioned and ready to take over during a failure. The secondary environment may have reduced capacity and does not normally serve production traffic. Recovery is generally in the minutes range, depending on the design.

In a Hot-Cold architecture, only the primary environment is active, while the DR environment mainly contains backups or basic infrastructure. During a disaster, infrastructure and data must be restored or provisioned before the application becomes available, resulting in a longer recovery time.

So, in simple terms: Hot-Hot means both are active, Hot-Warm means one is active and one is ready, and Hot-Cold means one is active while the other is mainly used for backup and recovery.