By Kevin Komiega
-- EMC took a step towards global data deduplication with today's debut of the Global Deduplication Array, a high-end inline deduplication storage system based on an extension of the Data Domain architecture.
The Global Deduplication Array (GDA) combines two EMC Data Domain DD880 controllers into a single deduplication system with a global namespace. Billed as a massive deduplication repository for large data centers, the GDA can accommodate up to 270 concurrent backup jobs and boasts throughput of up to 12.8TB per hour and up to 14.2PB of logical backup capacity.
The Data Domain deduplication architecture is CPU-centric, scaling in performance as Intel's multi-core processors increase in speed. This time, however, multiple controllers provide the performance boost.
"We usually ride the CPU curve to increase the performance of our systems, but this time we're combining to controllers to advance performance," says Shane Jackson, senior director of product marketing.
Jackson says the speed and capacity of the GDA make it easier for data center administrators to manage multi-terabyte data sets and multiple backup jobs in a single system.
The GDA will be generally available in the second quarter of 2010. The system will initially support Symantec NetBackup and Backup Exec through a software plug-in based on Symantec's OpenStorage (OST) API. EMC plans to add support for EMC NetWorker later this year.
In conjunction with the debut of the GDA, EMC doubled the capacity of the DD880 system to 7.1PB of logical capacity. EMC began shipping the DD880 controllers last year (see "Data Domain debuts new dedupe system").
For the security-conscious, EMC added an optional security feature called the EMC Data Domain Encryption software, which provides encryption of data at rest for the Data Domain deduplication appliances.
Data Domain Encryption integrates with the Data Domain inline deduplication process and provides 128-bit or 256-bit Advanced Encryption Standard (AES) algorithms for encrypting all data before it is written to disk. The encryption software works with all Data Domain protocols and software, and can be used with DD Replicator to send encrypted data over the WAN. Files that are retained on disk in a non-rewriteable and non-erasable format, as a long-term archive using EMC Data Domain Retention Lock software, can also be encrypted.
Like other Data Domain systems, the GDA runs the optional EMC Data Domain Replicator software, which now supports ‘one-to-many' directory replication that allows users to send multiple copies of source system data to different DR sites.
Data Domain Replicator facilitates WAN vaulting for use in disaster recovery (DR), remote office backup, or multi-site tape consolidation. Jackson says a single GDA system can support a replication fan-in of up to 270 remote offices using smaller Data Domain deduplication appliances. The GDA provides up to 54TB/hour of replication throughput.
The expanded DD880 capacity, DD Replicator enhancements and DD Encryption will be generally available in the second quarter of 2010.
EMC refreshes Data Domain dedupe systems