Use the new Asynchronous Replication technology to safeguard data from calamities. Corporations need data integrity and availability. Natural or man-made disasters can disrupt operations and compromise critical data.
Asynchronous Replication Meaning
Google Cloud are introducing Persistent Disk Asynchronous Replication, which provides a sub-1 min Recovery Point Objective (RPO) and low Recovery Time Objective (RTO) by duplicating data between Google Cloud regions, enabling disaster recovery for Compute Engine workloads.
This solution’s design places a strong emphasis on simplicity. A few API calls are used to manage replication; no VM agents or dedicated replication virtual machines are needed, and there are no restrictions on the guest operating systems that may be used or workload performance issues. PD Async Replication provides a quick, shared infrastructure basis to guard against disasters by operating at the block infrastructure level.
It may be used in conjunction with other data protection methods, such as disk clones, regional (synchronous replication) disks, and snapshots, and is designed to be easy to onboard, run, and monitor. Direct replication is carried out between a main disk in the area where your workload is executed and a recovery disk that you construct in a secondary area.
In addition to helping us satisfy regulatory requirements for low RPO/RTO data protection, persistent disk asynchronous replication provides the strong foundation Google Cloud require for infrastructure-based disaster recovery. Google Cloud’s whole DR testing, failover, and failback lifecycle is supported by PD Async Replication.
Configuring replication
Just two calls in the Google Cloud dashboard, gcloud, or API will enable PD Asynchronous Replication on your current PD disks. Using a reference to the primary disk you wish to safeguard, first make a new blank disk in the secondary area. Next, use a reference to the secondary disk to begin replication from the primary disk. Depending on the disk’s change rate, data is then automatically copied across disks, usually with an RPO of less than a minute. Before any data is exchanged, this setup procedure serves to guarantee that a certain action is executed in both locations. PD Async Replication does not require network reconfiguration.
You may see the network bytes delivered and the time since the last replication in Cloud Monitoring once PD Asynchronous Replication is operational. The operations teams in charge of a workload are in charge of determining when a disaster has struck the primary area and when a failover should be started. You may start the failover in a matter of minutes by attaching the secondary disk to a virtual machine (VM) in the secondary region and stopping disk replication.
After a prior failover to the secondary region, construct a new replication pair back to the primary region so that the workload can failback and return to the primary region. You may build resilient data copies that protect against localised interruptions brought on by natural disasters or other localised occurrences by transferring data across data centres located in various locations.
For complicated stateful workloads, use consistency groups
Consistency groups enable the coordinated management of dependent data in workloads with dispersed, dependent data across disks and virtual machine instances. By automatically synchronising the replication time across all disks in the group, PD Asynchronous Replication makes simultaneous and atomic data replication possible with consistency groups. In the case of a disaster, this aids in ensuring that data is consistent between primary and backup disks for effective workload recovery.
Disaster recovery testing
Google Cloud advise conducting testing in the secondary zone on a regular basis to assist guarantee that recovery protocols will function in the event of a true disaster. By bulk-cloning the secondary disks with a consistency group applied, even as they are getting fresh data, you may accomplish this without interfering with or disconnecting PD Asynchronous Replication.
Implement disaster recovery and high availability
Workloads requiring both disaster recovery and high availability (HA) are intended to employ PD Async Replication in conjunction with Regional Persistent Disk (Regional PD). A zonal disk in the primary or secondary region may be used in conjunction with a Regional PD configured as the primary or secondary async disk. The disk will continue to replicate from the remaining healthy zone to the secondary region even if there is only one zone in the primary region where Regional PD is setup experiencing an outage. It should be noted that each Regional PD is a single disk that stores data in two zones and is attachable in two zones, whereas PD Async Replication is configured between two separate disks.
Make your HA and DR posture better
Businesses may benefit from strong data protection and recovery capabilities by utilising PD Async Replication and consistency groups. This method helps you improve fault tolerance, reduce downtime, ensure data consistency, and protect important data from calamities. You may contribute to the development of a robust platform that offers constant access to data even in the event of unanticipated interruptions by using the strength of asynchronous replication and consistency groups.
The console, Compute Engine API, gcloud tool, Terraform, and Cloud Monitoring are all ways to access PD Async Replication.
