5.8

Interpreting Replication Statistics for a Site
You can use the reports that vSphere Replication compiles to optimize your environment for replication,
identify problems in your environment, and reveal their most probable cause.
Server and site connectivity, number of RPO violations, and other metrics give you, as an administrator,
the information you need to diagnose replication issues.
The following sections contain examples of interpreting the data displayed under Reports on the
vSphere Replication tab under Monitor.
RPO Violations
The large number of RPO violations can be caused by various problems in the environment, on both the
source and the target site. With more details on historical replication jobs, you can make educated
decisions on how to manage the replication environment.
Table 10‑2. Analysing RPO Violations
Probable Cause Solution
n
The network bandwidth cannot accommodate all
replications.
n
The replication traffic might have increased.
n
The initial full sync for a large virtual machine is taking
longer than the configured RPO for the virtual machine.
n
Disable the replication on some virtual machines with high
change rate to allow lower change rate virtual machines to
meet their RPO objectives.
n
Increase the network bandwidth for the selected host.
n
Check if the replication traffic has increased. If the traffic has
increased, investigate possible causes, for example the
usage of an application might have changed without you
being informed.
n
Check the historical data for average of transferred bytes for
a notable and sustained increase. If an increase exists,
contact application owners to identify recent events that
could be related to this increase.
n
Adjust to a less aggressive RPO or look at other ways to
increase bandwidth to accommodate the current RPO
requirements.
n
A connectivity problem exists between the source and the
target site.
n
An infrastructure change might have occurred on the target
site.
n
Check the site connectivity data to verify the connection
between the source and target site.
n
Check if the infrastructure on the target site has changed or
is experiencing problems that prevent vSphere Replication
from writing on the target datastores. For example, storage
bandwidth management changes made to target hosts
might result in storage delays during the replication process.
n
Check on the vSphere Replication Management Server
appliance and the vSphere Replication Server appliance.
Someone might have shut down the appliance or it might
have lost connection.
VMware vSphere Replication Administration
VMware, Inc. 72