In many jstestfuzz suites, it can often be difficult to determine exactly what command the client sent to the server, since there is often much obfuscation that occurs in the generated fuzzer files. In the case of certain failures, like a command that trips an invariant, seeing the full command object as it was received by the server would be very helpful. By increasing the verbosity of the "COMMAND" log component to level 2, we would be able to see all command requests and responses in the server logs. To start, we could consider adding this for all single threaded fuzzer suites, so as not to flood the logs too aggressively in concurrency suites that run many more operations. Eventually it might make sense to increase verbosity for concurrent fuzzer suites as well. These would include the following suites:

• jstestfuzz.yml
• jstestfuzz_interrupt.yml
• jstestfuzz_interrupt_replication.yml
• jstestfuzz_replication.yml
• jstestfuzz_replication_continuous_stepdown.yml
• jstestfuzz_replication_initsync.yml
• jstestfuzz_replication_session.yml
• jstestfuzz_sharded.yml
• jstestfuzz_sharded_causal_consistency.yml
• jstestfuzz_sharded_continuous_stepdown.yml
• jstestfuzz_sharded_session.yml