Degraded CN performance with BN

7 CNs with 7BNs where each CN talks to its own BN exclusively, NFT transfer reported **TPS=9383** that is below 10k:
https://perf.analytics.eng.hashgraph.io/ephemeral/test_runs/main_AdHoc7_578/report/NftTransferLoadTest/client.log
`03-03 06:24:34:355 [Client-T0] INFO com.hedera.benchmark.NftTransferJob - Finished NftTransferLoadTest: 185797334 transferred in 19800 sec, TPS: 9383`

At same time, neither BN nor CN had any connection issues/errors, saturation=0%
https://perf.analytics.eng.hashgraph.io/ephemeral/test_runs/main_AdHoc7_578/report/NftTransferLoadTest/podlog_solo-sdpt-n7/

All logs:
https://perf.analytics.eng.hashgraph.io/ephemeral/test_runs/main_AdHoc7_578/report/NftTransferLoadTest/

Start with the linked NftTransferLoadTest client.log and the CN/BN pod logs for test run main_AdHoc7_578, comparing the reported throughput with connection errors and saturation. Done means identifying the cause of the sub-10k TPS result and documenting a reproducible fix or actionable finding; no source file or test is named in the issue.

Written by the indexing model from the issue text.