> ## Documentation Index > Fetch the complete documentation index at: https://grandcentral.backbase.io/llms.txt > Use this file to discover all available pages before exploring further. # High availability and disaster recovery > Enterprise-grade resilience, observability, and business continuity for banking operations Grand Central iPaaS provides enterprise-grade high availability and disaster recovery capabilities for critical banking operations. The platform minimizes downtime and data loss through automated recovery and self-healing infrastructure. ## Service level agreements Grand Central offers tiered availability commitments: | Tier | Uptime SLA | Use case | | --------- | ---------- | ----------------------------------- | | Signature | 99.9% | Mission-critical banking operations | | Premium | 99.5% | Production banking services | | Essential | 99.5% | Standard banking workloads | **Recovery commitments:** * **Recovery Time Objective (RTO):** 6 hours for nominal situations * **Recovery Point Objective (RPO):** Varies by data criticality; critical financial data uses synchronous replication for minimal data loss ## High availability architecture The platform uses redundant infrastructure and automated recovery to maintain continuous operation during failures. ### Multi-zone deployment The platform deploys services across multiple Azure availability zones within each region: * **Availability zones:** Each zone operates independently with isolated power, cooling, and networking. * **Container distribution:** Kubernetes pods spread across zones using anti-affinity rules and zone-aware scheduling. * **Data redundancy:** Databases and storage replicate synchronously across zones for automatic failover without data loss. * **Network resilience:** Load balancers distribute traffic across zones and reroute automatically when zones become unhealthy. * **Active-active-active deployment:** Services run across three availability zones. The platform remains operational if two zones fail, provided the surviving zone has sufficient capacity to handle the workload. ### Auto-scaling and load balancing The platform automatically adjusts capacity based on demand: | Component | Layer | Function | | ------------------------- | ------- | ----------------------------------------------------------------------------- | | Application Gateway | Layer 7 | External ingress with HTTP/HTTPS routing, SSL termination, and WAF protection | | Azure Load Balancer | Layer 4 | Internal cluster TCP/UDP distribution for high-throughput scenarios | | Horizontal Pod Autoscaler | - | Scales pod count based on CPU, memory, or custom metrics | | Vertical Pod Autoscaler | - | Adjusts resource limits for right-sizing | ### Health monitoring Kubernetes probes ensure application health: * **Readiness probes:** Check if pods are ready to receive traffic. * **Liveness probes:** Restart unhealthy containers automatically. * **Startup probes:** Allow extended initialization time for slow-starting containers. ### Fault tolerance patterns The platform implements patterns to prevent cascade failures: * **Circuit breakers:** Detect failing services and fail fast to protect the system. * **Exponential backoff:** Retry with increasing delays to prevent retry storms. * **Fallback services:** Serve cached responses when primary services are unavailable. * **Dead letter queues:** Route permanently failed requests for later analysis. ## GitOps and self-healing The `applications-live` repository acts as the single source of truth for your runtime environment. ArgoCD continuously monitors cluster state: * **Drift detection:** ArgoCD detects when cluster resources deviate from Git configuration, such as manual changes. * **Automated sync:** The SelfHeal feature automatically reverts the cluster to the desired state defined in Git. * **Configuration reload:** Reloader watches for ConfigMap and Secret changes and performs rolling restarts to apply new configurations. ## Observability stack The platform includes a pre-configured monitoring stack. Access is controlled through roles assigned in the `self-service.tfvars` file. ### Grafana Visualization dashboards for cluster health, performance metrics, and business logic. | Attribute | Value | | ---------- | ----------------------------------------------------------------- | | Access URL | See the platform landing page for the latest Grafana link | | Access | Requires platform credentials with appropriate self-service roles | ### Additional observability tools | Tool | Purpose | | ---------- | ----------------------------------------------- | | Prometheus | Metrics collection from all platform components | | Jaeger | Distributed tracing across services | | ELK Stack | Log aggregation with full-text search | ## Disaster recovery strategy The platform implements multi-region failover to ensure business continuity during regional outages. ### Cross-region replication Data and services replicate across multiple Azure regions for geographic redundancy: | Region type | Function | | ----------- | ---------------------------------------------------------- | | Primary | Active production workloads with real-time data processing | | Secondary | Standby environments with continuous data replication | ### Data replication patterns | Pattern | RPO | Use case | | ------------ | ------------------ | ------------------------------------------------------------- | | Synchronous | 0 (zero data loss) | Financial transactions, account balances, authentication data | | Asynchronous | \< 15 minutes | Analytics, audit logs, configuration | ### Backup strategy | Backup type | Frequency | Description | | ------------- | ---------- | ---------------------------------------------- | | Full | Weekly | Complete backup of all data and configurations | | Incremental | Daily | Changed data since last backup | | Point-in-time | Continuous | Change data capture for precise recovery | Backup storage uses geo-redundant storage with encryption at rest, immutable storage, and long-term retention. ### Failover procedures Failover triggers include: * **Critical failures:** Region outages, database corruption, security breaches * **Performance degradation:** Response time thresholds, error rate spikes, SLA breaches * **Planned maintenance:** Scheduled updates, infrastructure upgrades The failover process executes through coordinated phases: 1. **Detection:** Automated monitoring identifies issues 2. **Assessment:** Evaluate scope and determine recovery strategy 3. **Redirection:** Update DNS and load balancer configurations 4. **Validation:** Health checks verify recovered services ### Recovery testing | Test type | Frequency | Scope | | ------------------ | ----------- | ------------------------------------------------------------- | | Component testing | Ongoing | Database failover, load balancer switching, backup validation | | Partial DR testing | Quarterly | Service layer failover, data replication | | Full DR testing | Bi-annually | Complete region failover with end-to-end validation | The platform also uses chaos engineering to inject controlled failures and validate resilience. ## Incident management ### Incident classification | Priority | Description | Response time | | ----------- | ----------------- | ------------- | | P1 Critical | Service down | 15 minutes | | P2 High | Major degradation | 1 hour | | P3 Medium | Minor impact | 4 hours | | P4 Low | No service impact | 24 hours | ### Escalation procedures * **Automated escalation:** Time-based rules, severity-based routing, on-call integration * **Communication protocols:** Stakeholder notifications, status page updates, customer communication ## Regulatory compliance The platform supports banking regulatory requirements for business continuity: | Standard | Description | | --------- | --------------------------------------------- | | Basel III | Operational resilience requirements | | PCI DSS | Availability standards for payment processing | | ISO 22301 | Business continuity management | | ISO 27031 | ICT readiness for business continuity | ## Next steps Explore security controls and compliance frameworks Learn about network architecture and connectivity patterns Review platform technologies and versions