Cold standby

As explained in Think Like a CTO

Cold standby is a recovery configuration in which standby services are not actively running and must be started before they can serve workloads.

Cold standby is a disaster-recovery configuration in which only minimal infrastructure remains running. During failover, standby machines must be powered on, started, synchronized with current data, and connected to workloads. section 8.4.1

It costs less than hot standby because it avoids continuously running and synchronizing replacement servers, but recovery takes longer because machines must boot and data must catch up. Its suitability depends on whether the business can tolerate minutes, hours, or days of downtime. section 8.4.1

A workable cold-standby plan should:

• Define the outage conditions and service-impact thresholds that trigger failover. section 8.4.4
• Securely record contacts, network details, passwords, service startup order, and data-population procedures. section 8.4.4
• Update public and private DNS records, allowing time for propagation, then validate roles, configurations, and services in the recovery zone. section 8.4.4
• Test the procedure regularly during off-hours and maintain a reverse plan for returning to the primary zone. section 8.4.4

Cloud as a cold standby

An enterprise can stream data to cloud storage while keeping replacement infrastructure inactive. When needed, it creates the required environment, starts the services, restores current data, and pays for additional capacity only during use. section 8.4.5

Figure 8.3 Downtime economics

Compare downtime tolerance with the cost and operational burden of standby configurations. section 8.4.1

How Think Like a CTO covers it

How deep it goes

Explains the configuration, cost-speed trade-off, failover logistics, DNS implications, testing, and return to the primary zone.

Examples

Uses booting standby machines, synchronizing data, changing DNS, and creating temporary cloud infrastructure as practical examples.

What it assumes

Assumes familiarity with enterprise services and networks while making downtime tolerance and recovery economics explicit.

What you can do after

Operate/configure a cold-standby recovery plan with explicit triggers, startup steps, network changes, validation, testing, and rollback.

How it connects

Connects cold standby with disaster recovery, hot standby, failover, data synchronization, DNS, and cloud infrastructure.

liveBookmanning.com

Read Think Like a CTO on liveBook Get the book at manning.com

Related concepts

• failoveruses · Failover is the transfer of service or workload from a failed or unavailable component to a functioning alternate or standby component.

This explainer was generated by Manning's knowledge graph from the text of Think Like a CTO. The section links point to the parts of the book it draws on.