Brocade Communications Systems StorageWorks 4400 - Enterprise Virtual Array Uživatelský manuál stáhnout pdf (Strana 3)

HP StorageWorks Continuous Access EVA

Continuous Access EVA is a feature of the Enterprise Virtual Array (EVA) that allows data replication

between two or more EVAs. This section describes some basic Continuous Access EVA terminology,

concepts, and features. The following topics are discussed.

 Data replication

 Copy sets

 Data replication (DR) groups

 Log disk

 Failover

Data replication

Data replication with Continuous Access EVA is host-independent and supports various interconnection

technologies, such as Fibre Channel over IP (FCIP) and Fibre Channel. Additionally, the EVA also supports

bidirectional replication. When a storage system contains both source virtual disks (Vdisks) and destination

Vdisks, it is bidirectional. A given storage system can have a bidirectional data replication relationship with

only one other storage system, and an individual Vdisk can have a unidirectional-replicating relationship

with only one other Vdisk. Continuous Access EVA enables data replication between all models of the EVA

family. Continuous Access EVA can replicate data synchronously and asynchronously between source and

destination arrays. Data replication between sites is most widely used when creating a true disaster-tolerant

data center as described in this document. It can also migrate data between two EVA arrays or provide an

alternative method for performing backups.

DR groups and copy sets

A copy set is a generic term for a replicated Vdisk. A data replication (DR) group is comprised of copy sets

(replicated Vdisks). Each DR group acts as a consistency group—all of its member Vdisks replicate to the

same destination EVA, failover together, preserve write order across members, and share a write history

log. Therefore, a DR group is the primary level of Continuous Access management.

Write history log (log disk)

The DR group has storage allocated on demand for its write history log (WHL). The WHL collects hosts write

commands and data if access to the destination storage system is unavailable. When the connection is re-

established, the content of the WHL is written to the destination Vdisk for quick resynchronization. Log

writes occur in the same order that the host writes occurred, this process is called merging.

Continuous Access failover

The recovery process whereby one or more DR group switches over to its backup is called a failover. The

process can be planned or unplanned. A planned failover allows an orderly shutdown of the attached hosts

and the EVA controllers before the redundant system takes over. An unplanned failover occurs when a

failure or outage occurs that may not allow an orderly transition of roles.

1 2 3 4 5 6 7 8 ... 37 38

Žádné komentáře

Brocade Communications Systems StorageWorks 4400 - Enterprise Virtual Array Uživatelský manuál Strana 3