Event Grid frequently asked questions

Availability and disaster recovery

Does Azure Event Grid support resiliency?

Yes, Azure Event Grid it's a high available service by zone-redundant support replicated in three availability zones when available to provide automatic in-region recovery of your data if a failure occurs in the region. See availability zones in Azure Event Grid and Highly available services for more details.

How does resiliency works in Azure Event Grid?

During a zone-wide outage, no action is required during zone recovery, Azure Event Grid will self-heal and rebalance itself to take advantage of the healthy zone automatically.

Does Azure Event Grid support availability zone resiliency across all the SKUs?

Yes, Azure Event Grid has only one SKU (Basic) and supports availability zones resiliency.

Do I need to configure Azure Event Grid to enable resiliency?

No, you don't need to configure anything. Azure Event Grid automatically replicates data across data centers in the selected region. See availability zones in Azure Event Grid for more details.

How do I recover from a failover?

If you opt in "cross-geo" in data residency configuration of the Azure Event Grid topics and domains, event subscription configuration data, also known as "metadata" will be replicated to a predefined secondary region. This allows Microsoft-initiated failover to the paired region if a failure occurs in a region. See Azure cross-region replication pairings for all geographies to learn more about the supported paired regions. Otherwise, if you opt in "regional", no data of any kind is replicated to another region by Microsoft. You may define your own disaster recovery plan. For more information, see Build your own disaster recovery for custom topics in Event Grid.

Microsoft-initiated failover is exercised by Microsoft in rare situations to fail over Event Grid resources from an affected region to the corresponding geo-paired region. This process doesn't require an intervention from the user. Microsoft reserves the right to make a determination of when this option will be exercised. This mechanism doesn't involve a user consent before the user's traffic is failed over.

Data residency

Where does Azure Event Grid store data?

Azure Event Grid provides you the flexibility to choose where data is stored and processed. By default, your data is replicated to a paired region. However, you can choose to keep it within the same region you deployed the service instance in.

The selection of the kind of data residency currently supported in topics and domains has implications in the disaster recovery configuration. When you create topics and domains, you can opt to replicate data to predefined secondary regions in case Microsoft decides to fail over the traffic by choosing Cross-Geo option, or can opt to keep your data in the region selected without any kind of data replication by choosing Regional option.

See Create a custom topic or a domain in Azure Event Grid to learn more about how to configure data residency when creating a topic or domain or see Configure a custom topic or a domain in Azure Event Grid to learn how to configure data residency in a preexisting topic or domain.

Next steps

You can learn more about Event Grid by visiting the following links: