Pool stops writing during load testing against domain and pool ledgers together

Build Info:
indy-node 1.6.648

Steps to Reproduce:
1. Run writing load test with 10 nyms/sec.
2. Run writing load test with 1 config txn/sec from another client and stop it after 10..15 minutes.
3. Run writing load test with 1 pool txn/sec from Step 2 client.
4. Check load script output/logs/metrics.

Actual Results:
Both load scripts start throwing Timeout errors after ~250 pool txns written, ~10:25 in metrics and logs (1) pool has stopped to write any txns, ~10:45 load script was stopped manually (2), after that the whole pool was restarted.