You are here
SiteSync™ – The Scale-Out Answer to Big Data Replication
Panasas® SiteSync™ high-speed parallel replication and file transfer software solves the challenge of moving massive data sets across the Local Area Network (LAN) quickly and with ease. Designed for use with the Panasas ActiveStor® high-performance hybrid storage appliance, SiteSync software can replicate data up to 10x faster than conventional file transfer utilities. SiteSync provides customers with extremely fast and flexible parallel data migration and synchronization for disaster recovery, disk-to-disk backup and remote archive applications.
SiteSync – True Scale-Out Replication
SiteSync performance is the result of a true scale-out architecture. Unlike traditional approaches, SiteSync uses parallelism to scale out to as many as 64 simultaneous Linux clients to speed the effort of data replication between any two high-performance storage systems.
Licensed and distributed with the PanFS® storage operating system, SiteSync is client-side software optimized for data movement between scale-out storage systems like the Panasas ActiveStor high-performance hybrid storage appliance. However, SiteSync can also be used to migrate data to or from third party storage systems as needed. SiteSync provides a familiar rsync-like command line interface, speeding the learning curve for storage administrators.
With SiteSync, replication cycles complete much faster and file synchronization delays are minimized. This leads to: Faster completion and availability of replicated copies Enhanced data protection driven by fast and scalable performance to target file systems.
SiteSync can replicate data up to 10x faster than common single-threaded file transfer utilities. As additional SiteSync clients are added, replication performance scales far beyond conventional software such as rsync.
SiteSyncTM Parallel Replication
Why Parallel Data Replication?
Technical computing and enterprise environments often struggle with rapidly moving data from source to destination. The reasons for moving data vary, but delays from slow serial data transfers impact workflows while magnifying data risks. Workflows and projects associated with a replicated volume must often wait until replication is complete and suffer increased exposure to data loss when data replication is slow.
Most replication and data movement tools were designed for legacy and scale-up storage architectures. These tools have not kept up with the performance requirements of parallel scale-out storage architectures. High-performance large capacity storage systems create massive data sets with hundreds of terabytes and petabytes. Serial data transfer becomes impractical as this can takes weeks to complete. Multiple serial replication streams can accelerate data movement, but at the expense of high management effort to partition data and manage each serial replication instance.
Scale-out storage systems scale out all in all respects: capacity, performance, and network bandwidth. Without the ability to scale replication bandwidth to match the scale of modern scale-out parallel storage systems, data replication, data migration, and data protection all take a big hit. The goal is for replication to scale out in the same manner as the storage systems they support. When this goal is achieved through a scalable parallel approach, performance is assured resulting in greater data reliability and increased business productivity.