Administrator Guide

Synchronous replication use cases
28 Dell EMC SC Series: Synchronous Replication and Live Volume | CML1064
going to satisfy a 24-hour RTO. Data growth on tapes means that there is a growing number of sequential-
access, long-seek time tapes for restoration. This diminishes the chances of meeting RTO, and increases the
chances that one bad tape will cause data recovery to fail. Data replication is a major player in meeting RTO.
Intra-volume consistency is extremely important in a distributed virtual machine disk or database volume
architecture. Comparing the synchronous replication modes, high consistency guarantees data consistency
between sites across all high consistency replicated volumes. Unfortunately, this is at the cost of destination
site latency, or worse, downtime of the production application if the destination volume becomes unavailable
or exceeds latency thresholds.
Outside of use cases that require the textbook definition of synchronous replication, high availability mode (or
asynchronous) may be a lot more attractive for DR purposes. This mode offers data consistency in the proper
conditions, as well as some allowance for latency while in sync. However, consistency is not guaranteed if
production application uptime is jeopardized should the destination volume become unavailable.
Because consistency cannot be guaranteed in high availability mode, it is important to implement VSS-
integrated Replay Manager snapshots or consistency groups with high availability synchronous replication
where a multiple volume relationship exists (this is commonly found in both SQL Server and Oracle
environments). While this will not guarantee active snapshot consistency across volumes, the next set of
frozen snapshots that have been replicated to the remote array should be consistent across volumes.
4.5.3 Preparing and executing volume and data recovery
With the appropriate hypervisor, tools, and automation in the DR plan, powering on a virtual machine is
relatively simple. Likewise, preparing the volumes for use and getting the database servers up and running
requires a quick process compared to legacy methods. This is especially true when replicating database
server boot from SAN (BFS) volumes with similar hardware at the DR site.
When choosing to access volumes at the DR site, it is important to consider the purpose. Is this a validation
test of the DR plan, or is this an actual declared disaster? As an active replication destination target
(regardless of asynchronous, synchronous, mode, or topology), destination volumes cannot be mounted for
read/write use to a storage host at the DR site. (For circumstances involving Live Volume, refer to section 5).
To perform a test of the DR plan, present view volumes from the snapshots of each test volume to the storage
hosts. Snapshots and view volumes are available for both asynchronous and synchronous replications in
either high consistency or high availability mode. Snapshots and view volumes are beneficial during DR
testing because replication continues between the source and destination volumes to maintain RPO in case
an actual disaster occurs during the test. Conversely, if a disaster is being declared and the Activate Disaster
Recovery feature is invoked, then replication from source to destination needs to be halted (if it has not been
already by the disaster) if the active volume at the destination site is intended for data recovery in the DR
plan.
Unisphere Central for SC Series in Dell Storage Manager is a unified management suite available to SC
Series and Dell FS8600 customers. Unisphere Central has disaster recovery features built into it that can
create, manage, and monitor replications, as well as automate the testing and execution of a predefined DR
plan.