---
title: "Cloud architecture and backups"
slug: "cloud-backups"
updated: 2025-11-26T15:03:09Z
published: 2025-12-10T14:08:19Z
canonical: "docs.grassfish.com/cloud-backups"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://docs.grassfish.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Cloud architecture and backups

Cloud architecture and backups are based on the following models:

****Azure topology****

The system is hosted in Microsoft Azure, with Europe West (Amsterdam) as the current default region.

**AKS resource group**

Includes the core runtime components:

- Azure Kubernetes Service (AKS) cluster
- SQL Managed Instance
- Storage Accounts
- GitOps and autoscaler extensions

**Network resource group**

Provides the networking foundation:

- Virtual Network and subnets
- Application Gateway with Web Application Firewall (WAF)
- Network Security Groups (NSGs)

**AKS-managed resource group**

Automatically created by Azure to host:

- Node pools
- Load balancer–related resources
- Other AKS-managed infrastructure components

****Redundancy****

Redundancy covers the following aspects:

- **Intra-region redundancy:** Achieved through multiple AKS nodes and pod replicas.
- **Multi-Availability Zone (AZ) and secondary regions:** Supported by the platform but implemented only when required, based on project-specific cost–benefit analysis.
- **Database redundancy:** Generally implemented using an active–passive replication approach.
- **Application redundancy:** AKS workloads can run in an active–active configuration for improved resilience.

****Backups and DR****

Backups and disaster recovery (DR) cover the following strategy and objectives:

**Backup strategy**

A typical backup cycle includes:

- **7 daily** backups
- **5 weekly** backups
- **3 monthly** backups

In container setups, backups cover databases, content, and other essential data. In legacy VM setups, full virtual machine backups are performed.

**DR objectives**

Disaster recovery is based on:

- **RTO/RPO:** Defined per project, usually within a minutes-to-hours range.
- **DR testing:** Supported and can be incorporated into operational agreements.