Liferay manages automatic zonal disaster recovery for our customers. This means that if a zone goes unavailable in the region where the customer’s data is hosted, the SaaS service would be still available from a different zone within the same region.
Some customers might require Liferay to ensure that their site is available even when the whole region goes down. With the Cross Region Disaster Recovery add-on, customers can have Liferay manage a disaster recovery (DR) site in another region in the case of such a disaster event. Due to data privacy rules, Liferay offers these DR cluster regions currently:
Live Cluster Region |
DR Cluster Region |
---|---|
Europe |
East Asia (Japan) |
US Central |
Europe |
South America |
Europe |
East Asia (Japan) |
Europe |
Australia |
Europe |
The DR procedures are agreed upon during the customer onboarding. The Customer can specify the conditions under which Liferay executes the failover procedures, including pre-authorization for Liferay to automatically execute DR steps when certain criteria and conditions are met, without needing to wait for customer confirmation. Liferay will monitor the customer’s site and execute these incident management workflows according to the agreement with the customer.
Liferay can execute the DNS changes for the customer if their DNS supports CNAME or ALIAS records. In other cases, the customer has to take care of redirecting the DNS.
Deploying and keeping the client extensions up to date on the DR environment is the customers' responsibility. So, whenever a customer deploys a new version of a client extension to their production environment, the deployment should be made to the DR environment as well.
The RPO* depends on the customer’s data set. Liferay targets a four-hour RPO, to migrate the data over to the customer’s DR site. The RTO** is 15 minutes once the failover decision has been made.
*RPO (Recovery Point Objective): the maximum data loss in case of an incident. It is defined by the maximum duration from the time when the last backup has been taken.
**RTO (Recovery Time Objective): the maximum time the service is down after the failover switch was decided.