Alarm
JetStream DR manages VM replication performance by adjusting its operations to accommodate varying network conditions. This alarm is triggered when “backpressure” is applied to throttle replication network traffic of a protected domain.
Possible Cause
Backpressure is used when the rate of incoming data from protected VMs exceeds the available outgoing network bandwidth to the destination site. This can lead to the replication log filling up or delays processing incoming data.
- If the incoming data exceeds the outgoing bandwidth, JetStream DR uses a replication log to temporarily cache data for replication at a later point in time after network traffic (i.e., caused by a spike in VM usage) decreases.
- If the incoming data remains excessively high or the replication log hits capacity, JetStream DR applies backpressure to slow down incoming data to prevent overloading the replication log and maintain manageable data rates for reliable operation.
- If backpressure limits are exceeded, JetStream DR switches to network bitmap protocol mode. This allows protected VMs to continue operation by reducing the load on outgoing replication.
- JetStream DR will periodically exit network bitmap protocol mode to check network performance conditions and will return to regular replication if possible.
Resolution
There are no specific actions required to address this alarm. However, best practices should be used to ensure optimal replication traffic performance of the system which minimizes the application of backpressure.