Sitecore Experience Manager

Considerations for using HADR

Abstract

Refer to these common Managed Cloud HADR considerations when building your Sitecore site.

When building a Sitecore solution using high availability disaster recovery (HADR), consider the following:

Note

Currently, the Managed Cloud disaster recovery options only include support for Azure Search.

Microsoft Azure data centers are organized into regions with a latency-defined perimeter because they are connected through a dedicated regional low latency network. To ensure fast backups and consistent customer delivery speed, you must choose a secondary data center that is in the same region as as the primary data center.

If your Sitecore solution uses any third-party service APIs that limit access based on IP, then you must register the IPs of your secondary data center with the Managed Cloud service.

Note

If you do not register your secondary data center IPs with the Managed Cloud Operations team, this could result in a delay in bringing your secondary Sitecore environment online.

Managed Cloud uses Azure Storage to host a page notifying customers of an outage, if a disaster occurs. One of the limitations of using Azure Storage is that it requires that your outage page be static (that is, written in pure HTML). This means that you cannot execute custom code in the backend. Keep your outage page simple and mention only the following:

  • That your site is temporarily unavailable.

  • That your support staff are aware of the outage and are working on it.

  • The approximate recovery time.

Another limitation of using Azure Storage or Azure WebApps to host your outage page is that you cannot return a 503 (Service unavailable) HTTP code. Usually, if a Search engine crawls your site and discovers a 503 code, it is aware that your site is temporarily down and will return later to re-index. However, if you do not have a 503 HTTP code then this can lead to undesirable SEO results. If it is important that you have a 503 HTTP code, speak with your salesperson to explore other Managed Cloud options that can return a 503 HTTP code.

Any certificates that you use for your Sitecore site are stored in the Key Vault of your secondary data center. In the event of a failover, the Managed Cloud Operations team uses these certificates, as part of the new setup.

Important

To help meet the hot-warm recovery time objective (RTO) it is essential to keep the certificates in the Key Vault of the secondary database up to date. Notify the Managed Cloud Operations team if there have been any changes to the certificates (for example, replacing a certificate that is due to expire).

The Managed Cloud Operations team will update the secondary datacenter, you must update the primary datacenter at the same time.