Published: September 08, 2022

Update: Migrations can now be performed via restart. Please do so before Nov 10, 2022.

Summary

At Liferay, we take our Cloud customer's uptime and availability very seriously. This is why we are committed to a Monthly Infrastructure Availability (“MIA”) of 99.95% or greater. Liferay works with cloud providers to serve as our Cloud back-end and leverage their tools to ensure High Availability. In our conversations with our cloud provider, we've been informed that our current High Availability setup (previously handled by failover replicas) has been deprecated and will soon be shut down. In its place, our provider offers High Availability via regional persistent disks that synchronously replicate data between zones. Because of this change, Liferay is working to migrate our Cloud customer's databases to the new system.

Benefits

Regional persistent disk provides synchronous replication of data between two zones in a region while failover replica provides semi synchronous replication. In the case of a failover event, the semi synchronous failover replica has additional latency because Cloud SQL waits for the failover replica to catch up to the primary instance’s state. According to our cloud provider, the regional HA mode is 2.5 times faster than on average when compared to the replica mode strategy.

An added benefit of persistent disks is the high-performance block storage for workload. Persistent disk performance scales with the size of the disk and with the number of vCPUs on your VM instance. 

Aside from minimizing latency, a key factor for upgrading from the failover replica mode is that our cloud provider will eventually deprecate the failover replica mode. We are in close contact with their team and they will support us during this transition.

Details

For many customers, this process will be seamless. But for others, some additional steps will be required. All cloud environments will fall under one of the following three categories and will have corresponding actions that must be taken for the migration.

Category 1: Database Service Image Version ≥ 4.2.1

Environments that have a database service image version of 4.2.1 or higher have the fewest steps and the process will be relatively seamless. Our team will be initiating the migration in the background. Only one step will be required to complete the process after our team has initiated the migration.

Steps:

1. Restart the database service

Category 2: Database Service Image Version ≥ 4.0.0 but < 4.2.1

Environments that have a database service image version between 4.0.0 and 4.2.1 have one additional step then the process is the same as above.

Steps:

1. Update the database service version to 4.2.1 or higher
2. Restart the database service

Category 3: Database Service Image Version < 4.0.0

Environments that have a database service image version below 4.0.0 have one further step then it will look the same as above.

Steps:

1. Upgrade all services from stack 3.x to stack 4.x
2. Update the database service version to 4.2.1 or higher
3. Restart the database service

Timeline and Additional Instructions

This means that your database service must be on version 4.2.1+ and you must have performed a database restart by then.

For customer environments that fall under Category 1 (≥4.2.1), the restart must be performed by the above date and no interaction with Liferay will be necessary.

For customers whose environments fall under the other two categories, Liferay will reach out to the appropriate customer contact in order to make sure the requirements are clear and provide assistance where possible.

Additional Information

Liferay will be communicating this information via multiple channels leading up to the dates mentioned above to ensure that the relevant parties are notified of the changes in advance.

As always, if you encounter any issues or have questions about these changes, please submit a ticket and Liferay Support will be happy to assist.

Incidents and Uptime Status