"The combination of disk-based change journals that are pulled by the replication targets instead of pushed by the source, makes it extremely resilient, capable of automatically recovering from elongated disruptions," said Christophe Bertrand, senior director, solutions and product marketing business continuity at Hitachi Data Systems. "Because changes are pulled by replication target arrays, valuable processing cycles are offloaded from primary arrays to secondary target arrays."

Type of replication

Array based

Host based

Network based

Support of heterogeneous environments

Low; only works between similar arrays

High; storage-agnostic and works with network- and direct-attached storage

High; storage array- and platform-agnostic

Performance and scalability

Depends on the storage array; very good in high-end arrays

Good; workload is spread across servers; limited scalability because of management challenges

Very good

Cost

Requires similar arrays; high entry cost; expensive for a large number of locations

No hardware required; low entry cost; cost rises proportionally to the number of servers

Requires intelligent switches or inline appliances; high entry cost; expensive for a large number of locations

Complexity

Medium to high

Low

Medium to high

Replication Modes

Synchronous and asynchronous

Asynchronous

Synchronous and asynchronous

Predominant Replication Type

Logical ...

File-system based

LUN or volume block-level based

[IMAGE]

< PREV PAGE   |   1  |   2  |   3  |   4  |   5  |   6  |   7  |   NEXT PAGE  >