weekly updates

posts/career/ha-dr.org

@@ -0,0 +1,87 @@
+#+TITLE: High Availability, Disaster Recovery and Business Continuity
+#+OPTIONS: num:nil
+#+DATE: <2026-03-11 Wed 17:18>
+#+filetags: :learning:notes:
+#+WIP: 
+#+COMMENTS: t
+#+SLUG: high-availability-disaster-recovery
+
+* HA, DR and BC Competency Log
+Understanding High Availability (HA), Disaster Recovery (DR), and
+Business Continuity Planning (BCP) is important when designing reliable
+systems and analysing service incidents. These concepts help Microlise
+(and organisations in general) to minimise downtime, recover from
+failures, and maintain service availability.
+
+*** High Availability (HA)
+High Availability focuses on preventing service disruption by using
+redundancy. Systems are designed with backup components so that if one
+fails, another can take over automatically.
+
+Examples include:
+
+- Redundant servers or network paths
+- Load balancing across multiple systems
+- Automatic failover modes
+
+This ensures services continue operating even when individual components
+fail. We have this at Microlise where if there is a failure on one of
+the data centres, we can failover to the other one.
+
+An analogy from the article compares this to a bicycle with two brakes.
+The bike can operate with only one brake, but having two provides
+redundancy in case one fails.
+
+*** Disaster Recovery (DR)
+Disaster Recovery focuses on restoring systems after a major failure.
+This includes recovering infrastructure, applications, and data so
+services can resume operation.
+
+Some of the common DR tools and methods include:
+
+- Offsite backups
+- Replicated environments (e.g., multiple data centres with sync)
+- Recovery procedures and restoration tools
+
+Unlike High Availability, which tries to prevent downtime, DR assumes a
+failure has already occurred. In the bicycle analogy, after the crash
+the rider takes a bus home and then drives to work to complete their
+important task. This represents a recovery process after a major
+disruption.
+
+*** Business Continuity Planning (BCP)
+BCP ensures the business can continue operating during or after a
+disruption. This may involve alternative systems, temporary processes,
+or backup locations to keep services running.
+
+In the analogy, the car at home represents BCP because it allows the
+rider to continue their journey despite the broken bicycle. A Business
+Impact Assessment (BIA) helps determine the priority of services and the
+acceptable level of downtime.
+
+*** RTO and RPO
+Two key recovery metrics are:
+
+- Recovery Time Objective (RTO): The maximum time allowed to restore a
+  service after failure. For example, if a critical service must be
+  restored within 4 hours, the disaster recovery process must ensure
+  systems are operational within that timeframe.
+- Recovery Point Objective (RPO): The maximum acceptable amount of data
+  loss.
+
+RPO example:
+
+- An RPO of 0 minutes means no data loss is acceptable.
+- An RPO of 1 hour means up to one hour of data could be lost.
+
+These values help define the required level of HA and DR design.
+
+*** Application in Real Scenarios
+These concepts are useful when performing incident debugging, root cause
+analysis, or explaining service issues to customer support teams. For
+example, engineers may check whether failover worked correctly, whether
+recovery met the RTO, or whether backups allowed data to be restored
+within the RPO.
+
+In summary, the reason why we need to understand HA, DR, and BCP is to
+maintain service availability and ensure business continuity.