You have implemented synchronous mirroring for your data warehouse. You are experiencing performance issues during peak workloads. Which of the following techniques can help improve performance?

Increasing dataflow activities relates to ETL/ELT processing capacity, not the performance of synchronous database mirroring.

Synchronous mirroring, by definition, requires immediate mirroring of all updates, so 'reducing the frequency of mirroring updates' is not a viable option for this type of mirroring.

In synchronous mirroring, transactions on the primary database commit only after the transaction log records are hardened on the secondary database, meaning network latency and bandwidth between the two databases directly impact primary database write performance. Using a faster network connection reduces this bottleneck, allowing transactions to commit more quickly and improving overall performance during peak workloads.

Increasing the number of workers in a dataflow cluster is related to data processing capacity, not the underlying database mirroring performance.