Home > Storage > PowerScale (Isilon) > Product Documentation > Storage (general) > Dell PowerScale: Considerations and Best Practices for Large Clusters > Replication recommendations
Data protection and disaster recovery at the large cluster level is another area that relies upon a well-designed architecture. For example, many is the cluster that has started out with NDMP backup as the DR strategy and grown organically into the multi-petabyte size. At this point, realizing that traditional backup is not scaling, replication has had to be hurriedly introduced – often involving a lengthy initial replication job for a huge dataset over a WAN link.
OneFS SyncIQ provides high-performance, asynchronous replication of unstructured data to address a broad range of recovery objectives which scale linearly up into the large and extra-large cluster ranges. Utilizing a highly-parallel, policy-based replication architecture, SyncIQ works across either LAN or WAN connectivity, providing protection from both site-specific and regional disasters.
Synchronization policies may be configured at the file, directory, or entire file system-level and have a priority setting to allow favored policies to preempt others. In addition to chronological scheduling, replication policies can also be configured to start whenever the source is modified (change based replication).
Leveraging SnapshotIQ, the Linear Restore functionality of SyncIQ can detect and restore (commit) consistent, point in time, block-level changes between cluster replication sets, with a minimal impact on operations and a granular RPO. In the event that a primary cluster becomes unavailable, SyncIQ provides the ability to failover to a mirrored, DR cluster. During such a scenario, the administrator makes the decision to redirect client I/O to the mirror and initiates SyncIQ failover on the DR cluster. Users will continue to read and write to the DR cluster while the primary cluster is repaired. As such, disaster recovery sites for large clusters typically contain SyncIQ target cluster(s) which store replicated data from the primary site(s).
Further information is available in the Dell PowerScale SyncIQ: Architecture, Configuration, and Considerations white paper.