IS Operations & Resilience: Common Mistakes
Confusing a business continuity plan (BCP) with a disaster recovery plan (DRP).
A BCP addresses the continuation of critical business functions during and after a disruption. A DRP specifically focuses on restoring IT infrastructure and systems. Treating them as interchangeable results in gaps; business processes may lack continuity plans, or IT recovery may lack business context.
Correct approach: Develop both a BCP and a DRP. The BCP covers business processes, personnel, and facilities. The DRP covers IT systems, data, and infrastructure recovery. Both should align with the business impact analysis.
Setting recovery objectives (RTO and RPO) without conducting a business impact analysis (BIA).
The BIA identifies critical business processes and the impact of their disruption over time. Without it, recovery time objectives (RTO) and recovery point objectives (RPO) are arbitrary and may not reflect actual business needs. This can lead to over-investment in non-critical systems or under-protection of essential ones.
Correct approach: Perform a BIA first to identify critical processes and their dependencies. Use the BIA results to set RTO and RPO values that align with business requirements and risk tolerance.
Believing that daily full backups are always the best strategy.
Full backups consume significant storage and time. A combination of full, incremental, and differential backups is usually more efficient and can meet RPO requirements with lower resource consumption. The optimal strategy depends on data volume, change rate, and recovery needs.
Correct approach: Design the backup strategy based on the RPO, data volume, and available resources. Combine full backups with incremental or differential backups to balance recovery speed with storage efficiency.
Failing to test backup restoration regularly.
Backups are worthless if they cannot be restored. Media degradation, configuration errors, and software incompatibilities can all prevent successful restoration. Without regular testing, the organization discovers these problems only during an actual disaster.
Correct approach: Test backup restorations on a regular schedule. Verify that data can be recovered completely and within the required timeframe. Document test results and address any failures.
Treating incident management and problem management as the same process.
Incident management focuses on restoring normal service as quickly as possible. Problem management focuses on identifying and eliminating the root cause to prevent recurrence. Merging these processes leads to either slow incident response (due to root cause analysis during outages) or persistent recurring incidents (due to lack of root cause investigation).
Correct approach: Maintain separate incident and problem management processes. Resolve incidents quickly to restore service, then initiate problem management to investigate root causes and implement permanent fixes.
Assuming a hot site is always the best disaster recovery option.
A hot site provides the fastest recovery but is also the most expensive option. A warm site or cold site may be sufficient depending on the organization's RTO and budget. Choosing a hot site without considering actual recovery requirements wastes resources.
Correct approach: Select the recovery site type based on the RTO established in the BIA. Hot sites suit near-zero RTO needs; warm and cold sites are appropriate for longer acceptable recovery times.
Neglecting IT asset management and configuration management.
Without accurate records of IT assets and their configurations, the organization cannot effectively manage changes, respond to incidents, or plan capacity. Shadow IT, unauthorized software, and unpatched systems thrive in environments with poor asset management.
Correct approach: Maintain a comprehensive configuration management database (CMDB) or asset inventory. Regularly reconcile the inventory against actual deployed assets and update it after every change.
Overlooking the importance of operations documentation and run books.
Operations staff rely on documented procedures for routine tasks, troubleshooting, and escalation. Without current documentation, the organization becomes dependent on individual knowledge, creating a single point of failure. Staff turnover or absence can then cause service disruptions.
Correct approach: Maintain up-to-date operations manuals, run books, and standard operating procedures. Review and update documentation after every significant change to systems or processes.
Ignoring capacity planning and performance monitoring.
Without proactive capacity planning, systems may become overloaded, leading to performance degradation or outages. Reactive approaches (waiting for problems to occur) result in emergency spending, poor user experience, and potential SLA breaches.
Correct approach: Implement continuous performance monitoring and use trend analysis to forecast capacity needs. Plan and budget for capacity upgrades before thresholds are breached.
Storing backup media at the same location as the primary data center.
If a disaster (fire, flood, earthquake) destroys the primary site, co-located backups will be destroyed as well. This defeats the purpose of having backups for disaster recovery. Geographic separation is essential for resilience.
Correct approach: Store backup media at an offsite location that is geographically separated from the primary data center. The offsite location should have appropriate environmental and security controls.