RecoverPoint replication for VMware
Some businesses have a challenge in maintaining high availability of their virtualized
servers. In the event of a failure, the business may need to re-install the host
operating system and recover the data. This works if your recovery point objective
(RPO) is measured in multiple hours or days, but if you have a smaller RPO you need a
more realtime solution.
RecoverPoint provides a realtime replication solution for VMware ESX and virtual
machines. For asynchronous replication it uses a unique small-aperture data capture
technology. For synchronous replication every change is sent to the remote site. This
ensures that the VMware ESX platform is protected from data corruption and
guarantees recoverability with minimum to no data loss with point-in-time crash-
consistent images. For local replication in the same SAN, and synchronous remote
replication between two separate SANs, every write is captured and stored in the
RecoverPoint journal. For remote asynchronous replication between two separate
SANs, the user specifies policies for the replication that controls the aperture size to
ensure that specific RPOs are met for each consistency group. Additionally the user
can establish policies that enforce synchronous remote replication or utilize the
dynamic synchronous capability of RecoverPoint, which automates changes in
replication state from synchronous to asynchronous.
RecoverPoint supports the replication of iSCSI and SAN-attached volumes.
RecoverPoint supports RDM/P and VMFS volumes and if the VMware ESX server is
configured for boot from SAN (BFS), then the boot volumes can also be replicated to
the remote site. RecoverPoint captures changes to data by intercepting every write
(either to an RDM/P volume or to a VMFS volume) that reaches the SAN though the
intelligent fabric or CLARiiON-based splitters. Only the CLARiiON-based splitter
supports iSCSI volumes. If intelligent fabric or a CLARiiON-based splitter is not used,
then RecoverPoint captures the changes by using a Windows-based KDriver; however
the KDriver can only capture changes written to an RDM/P volume.
To ensure that the images are consistent for each virtual machine, it is recommended
that you create frequent RecoverPoint bookmarks while the ESX server is in a
quiesced state. To quiesce the ESX server, first power off all guest virtual machines
that reside on replication LUNs or VMFS volumes. Once the virtual machines are
powered off create a RecoverPoint bookmark for the appropriate consistency groups
using either the RecoverPoint GUI or CLI. Alternatively, it is possible to use the
VMware Tools SYNC driver (LGTO_SYNC) to flush pending writes to a VMFS before
creating the bookmark.
1
Most of today’s applications and databases have a built-in
resiliency allowing them to deal with crash-consistent images without the need to
flush pending writes or shut down virtual machines.
RecoverPoint’s image access technology allows administrators to access any image in
seconds and to mount it directly as either a VMFS volume or as a RDM/P volume.
Once an image is mounted it can be repurposed for backup or DR testing, or used for
immediate recovery of files, folders, volumes, or entire virtual machines. When using
1
While it is possible to use the SYNC driver, a description of this procedure is beyond the scope of this document.
(40 strony)
Manymanuals.com
Manymanuals.de
Manymanuals.fr
Manymanuals.it
Manymanuals.pl
Manymanuals.cz
Manymanuals.es
Manymanuals-pt.com
Komentarze do niniejszej Instrukcji