Azure 99.999 Availability Zones for High Reliability

Azure's Availability Zones are designed to provide high reliability with a 99.999 uptime guarantee. This means that Azure's infrastructure is built to minimize downtime and ensure that your applications and services remain accessible at all times.

Each Availability Zone is a separate physical location within a region, with its own power, cooling, and networking infrastructure. This ensures that if one zone goes down, the others can continue to operate and keep your applications running.

With multiple Availability Zones, you can deploy your applications across multiple zones for even higher reliability. This is especially important for mission-critical workloads that require 99.999 uptime.

Azure's 99.999 availability is a game-changer for businesses that can't afford downtime. Azure Cosmos DB is a multitenant service that manages all details of individual compute nodes transparently, guaranteeing SLAs for availability and P99 latency through all automatic maintenance operations.

This means you don't have to worry about patching or planned maintenance, and your data is always available. Individual node outage resiliency is also a key feature, with Azure Cosmos DB automatically mitigating replica outages by guaranteeing at least three replicas of your data in each Azure region for your account within a four-replica quorum.

You can enable availability zone support for your Azure Cosmos DB account using the Azure portal, Azure CLI, or Azure Resource Manager templates. This provides an RTO of 0 and an RPO of 0, even in a zone outage.

To migrate your Cosmos DB account to availability zone support, see the instructions in the Azure documentation. Availability zones are physically separate groups of data centers within a region, each with independent power, cooling, and networking infrastructure.

Deploying VMs and disks across multiple availability zones provides the highest SLA, with a guaranteed connectivity of 99.99 percent. This reduces expected downtime by a factor of ten, making it ideal for applications that can't afford even an hour of downtime per month.

Here are the methods to distribute VMs and disks across availability zones:

Each of these methods provides redundancy in VMs and disks across multiple data centers in a region, allowing you to fail over to another zone if there's a data center or zonal outage. By using availability zones, you can minimize downtime and service disruption, making it a key component of Azure's 99.999 availability.

A Service Level Agreement (SLA) is a promise by Azure to ensure a certain level of availability for your applications. Azure's SLAs are higher for accounts with availability zones, reaching 99.995% compared to 99.99% for single-region accounts without availability zones.

Availability zones provide distinct power sources, networks, and cooling, making them more resilient to outages. Enabling availability zones increases the cost by 25% for accounts without multi-region writes and autoscale mode.

The SLA for Azure services varies depending on the configuration. For example, a single-region account without availability zones has a 99.99% write availability SLA, while a single-region account with availability zones has a 99.995% write availability SLA.

Here's a summary of the SLAs for different Azure configurations:

It's essential to note that SLAs are not cast-iron guarantees of availability and may not account for unforeseen events that cause outages.

Azure Cosmos DB provides two backup modes to protect against complete data loss: continuous backups and periodic backups. Continuous backups back up each region every 100 seconds, enabling you to restore your data to any point in time with 1-second granularity.

Azure Cosmos DB accounts deployed in multiple regions have data durability that depends on the consistency level configured on the account. Here's a table detailing the RPO (Recovery Point Objective) for each consistency level:

Managed disks in Azure provide at least 99.999999999% (11 9's) of durability and are designed for 99.999% availability.

Azure Cosmos DB provides robust durability features to ensure your data remains accessible and intact. Data loss is generally not a concern in a single region deployment.

Data access is restored after Azure Cosmos DB services recover in the affected region. However, data loss might occur with an unrecoverable disaster in the Azure Cosmos DB region.

To mitigate this risk, Azure Cosmos DB offers two backup modes: continuous backups and periodic backups. Continuous backups back up each region every 100 seconds, allowing you to restore your data to any point in time with 1-second granularity.

Periodic backups fully back up all partitions from all containers under your account, with no synchronization across partitions. The minimum backup interval is 1 hour.

Data durability in multiple regions depends on the consistency level you configure on the account. Here's a breakdown of the RPO for an Azure Cosmos DB account deployed in at least two regions:

For bounded staleness, the minimum value of K and T is 100,000 write operations or 300 seconds. This value defines the minimum RPO for data when using bounded staleness.

Managed disks, on the other hand, offer 99.999% availability and at least 99.999999999% (11 9's) of durability. Your data is replicated three times, ensuring high fault tolerance and persistence.

Testing for high availability is crucial to ensure your application remains accessible even in the event of an outage. You can temporarily disable service-managed failover for your Azure Cosmos DB account to test end-to-end high availability.

To do this, invoke manual failover using PowerShell, the Azure CLI, or the Azure portal, and then monitor your application. This will simulate a failover scenario without actually disrupting your data.

It's essential to note that manual failover requires region connectivity to maintain data consistency, so it won't succeed during an Azure Cosmos DB outage on either the source or destination region.

Deploying VMs and disks across multiple fault domains is another way to ensure high availability. By doing so, you can prevent multiple VMs from going down in case of a storage fault domain outage.

Fault domains define groups of VMs that share a common power source and a network switch. To deploy resources across multiple fault domains, you can use regional Virtual Machine Scale Sets or availability sets.

Deploying your Virtual Machines (VMs) and disks across multiple fault domains is a great way to ensure high uptime and durability, especially if you can't deploy them across availability zones or have ultra-low latency requirements.

This method provides the second highest uptime SLA, after distributing VMs across three availability zones.

Fault domains define groups of VMs that share a common power source and a network switch, making them a reliable option for redundancy.

To deploy resources across multiple fault domains, you can use regional Virtual Machine Scale Sets or availability sets.

The storage fault domains of the disks are aligned with the compute fault domains of their respective parent VMs, preventing multiple VMs from going down if a single storage fault domain experiences an outage.

Multiple VMs can be deployed across fault domains using either regional Virtual Machine Scale Sets or availability sets, providing a reliable and durable solution.

The following diagram depicts the alignment of compute and storage fault domains when using either regional Virtual Machine Scale Sets or availability sets.

Disaster recovery is about recovering from high-impact events, such as natural disasters or failed deployments that result in downtime and data loss.

Microsoft uses the shared responsibility model for disaster recovery, where they ensure the baseline infrastructure and platform services are available, but you're responsible for setting up a disaster recovery plan for your workload.

To ensure business continuity, we recommend setting up your Azure Cosmos DB account with a single write region and at least a second (read) region and enabling service-managed failover.

Service-managed failover allows Azure Cosmos DB to fail over the write region of a multiple-region account to preserve business continuity at the cost of data loss.

Even with service-managed failover enabled, partial outage may require manual intervention for the Azure Cosmos DB service team, and it may take up to 1 hour (or more) for failover to take effect.

To achieve better write availability during partial outages, we recommend enabling availability zones in addition to service-managed failover.

Region outages are outages that affect all Azure Cosmos DB nodes in an Azure region, across all availability zones, and can be configured to support various outcomes of durability and availability.