Administrator Guide

Synchronous replication features
13 Dell EMC SC Series: Synchronous Replication and Live Volume | CML1064
3 Synchronous replication features
SC Series storage supports a wide variety of replication features. Each feature is outlined in the following
sections.
3.1 Modes of operation
A number of evolutionary improvements have been made to enhance synchronous replication with SC Series
arrays. Among these improvements are choice in replication mode on a per-volume basis. Synchronous
replication can be configured in one of two modes: high consistency or high availability.
3.1.1 Legacy
Synchronous replications created prior to SCOS 6.3 are identified as legacy after upgrading to SCOS 6.3 and
newer. Legacy synchronous replications cannot be created in SCOS 6.3 or newer and do not possess the
newer synchronous replication features currently available. To upgrade a legacy synchronous replication to
synchronous high consistency or synchronous high availability replication, a legacy synchronous replication
must be deleted and recreated after both source and destination SC Series arrays have SCOS 6.3 or newer
installed. Deleting and recreating a synchronous replication will result in data inconsistency between the
replication source and destination volumes until 100% of the initial and journaled replication is completed.
3.1.2 High consistency
Synchronous high consistency mode rigidly follows the storage industry specification of synchronous
replication outlined earlier and shown in Figure 1. The mechanisms involved with this method of replication
will guarantee data consistency between the replication source and destination volumes unless an
administrator pauses the replication for maintenance or other reasons. Latency can impact applications at the
source volume if the replication link or replication destination volume is unable to absorb the amount of data
being replicated or the rate of change. Furthermore, if write transaction data cannot be committed to the
destination volume, the write will not be committed on the source volume and in effect, a transaction involving
a write fails. An accumulation of write failures will likely result in an application failure or outage when a
tolerance threshold is crossed. For these reasons, application latency and high availability are important
points to consider in a storage design proposing synchronous replication in high consistency mode.
3.1.3 High availability
Synchronous high availability mode bends the rules of synchronous replication by relaxing the requirements
associated with high consistency mode. While the replication link and the replica destination storage are able
to absorb the write throughput, high availability mode performs like high consistency mode (described in
section 3.1.2 and illustrated in Figure 1). Data is consistently committed at both source and destination
volumes and excess latency in the replication link or destination volume will be observed as application
latency at the source volume.
The difference between high consistency and high availability mode is that data availability will not be
sacrificed for data consistency. What this means is that if the replication link or the destination storage either
becomes unavailable or exceeds a latency threshold, the SC Series array will automatically remove the dual
write committal requirement at the destination volume. This allows application write transactions at the source
volume to continue, with no downstream latency impacts, instead of write I/O being halted or slowed, which is
the case with high consistency mode and legacy synchronous replication. This relaxed state is referred to as
being out of date. If and when an SC Series array enters the out-of-date state, inconsistent write I/O will be
journaled at the source volume. When the destination volume becomes available within a tolerable latency