Improve txn-emitter stability and visibility on Forge

[igor-aptos]

• visibility - add few more log lines, and print txn-emitter stats (every 1m) as well
• make txn-emitter work more precisely. With large number of accounts/submission_workes (and graceful test has 180000 of them), there were a few issues. First creating workers is slow, and so they start with some divergence to begin with - fixing by creating them first before starting, and passing the same "start" instant to all. second divergence wasn't visible because it logged only after 240s divergence (changed to log on 5s divergence)
• fix db flag that got changed recently for test tuned for throughput

Description

Test Plan

[github-actions]

✅ Forge suite land_blocking success on 1c59c35762e6f19f53dec1496717d26fb133eea5

performance benchmark : committed: 5661 txn/s, latency: 7012 ms, (p50: 6300 ms, p90: 9000 ms, p99: 26000 ms), latency samples: 2417327
Max round gap was 1 [limit 4] at version 1404300. Max no progress secs was 3.4734979 [limit 10] at version 1531484.
Test Ok

• Grafana dashboard
• Humio Logs
• Test runner output
• Test run is land-blocking

[github-actions]

✅ Forge suite compat success on testnet_2d8b1b57553d869190f61df1aaf7f31a8fc19a7b ==> 1c59c35762e6f19f53dec1496717d26fb133eea5

Compatibility test results for testnet_2d8b1b57553d869190f61df1aaf7f31a8fc19a7b ==> 1c59c35762e6f19f53dec1496717d26fb133eea5 (PR)
1. Check liveness of validators at old version: testnet_2d8b1b57553d869190f61df1aaf7f31a8fc19a7b
compatibility::simple-validator-upgrade::liveness-check : committed: 9211 txn/s, latency: 3611 ms, (p50: 3500 ms, p90: 5000 ms, p99: 5800 ms), latency samples: 313180
2. Upgrading first Validator to new version: 1c59c35762e6f19f53dec1496717d26fb133eea5
compatibility::simple-validator-upgrade::single-validator-upgrade : committed: 5217 txn/s, latency: 6128 ms, (p50: 6500 ms, p90: 8100 ms, p99: 8700 ms), latency samples: 198260
3. Upgrading rest of first batch to new version: 1c59c35762e6f19f53dec1496717d26fb133eea5
compatibility::simple-validator-upgrade::half-validator-upgrade : committed: 4160 txn/s, latency: 7558 ms, (p50: 8300 ms, p90: 9700 ms, p99: 10400 ms), latency samples: 158100
4. upgrading second batch to new version: 1c59c35762e6f19f53dec1496717d26fb133eea5
compatibility::simple-validator-upgrade::rest-validator-upgrade : committed: 6912 txn/s, latency: 4767 ms, (p50: 4600 ms, p90: 6900 ms, p99: 8400 ms), latency samples: 241920
5. check swarm health
Compatibility test for testnet_2d8b1b57553d869190f61df1aaf7f31a8fc19a7b ==> 1c59c35762e6f19f53dec1496717d26fb133eea5 passed
Test Ok

• Grafana dashboard
• Humio Logs
• Test runner output
• Test run is land-blocking

[github-actions]

✅ Forge suite framework_upgrade success on aptos-node-v1.3.0_3fc3d42b6cfe27460004f9a0326451bcda840a60 ==> 1c59c35762e6f19f53dec1496717d26fb133eea5

Compatibility test results for aptos-node-v1.3.0_3fc3d42b6cfe27460004f9a0326451bcda840a60 ==> 1c59c35762e6f19f53dec1496717d26fb133eea5 (PR)
Upgrade the nodes to version: 1c59c35762e6f19f53dec1496717d26fb133eea5
framework_upgrade::framework-upgrade::full-framework-upgrade : committed: 4747 txn/s, latency: 6810 ms, (p50: 7200 ms, p90: 9300 ms, p99: 9800 ms), latency samples: 175640
5. check swarm health
Compatibility test for aptos-node-v1.3.0_3fc3d42b6cfe27460004f9a0326451bcda840a60 ==> 1c59c35762e6f19f53dec1496717d26fb133eea5 passed
Test Ok

• Grafana dashboard
• Humio Logs
• Test runner output
• Test run is land-blocking

[igor-aptos]

it takes 5 minutes to go through the loop below, in the graceful overload case (https://aptoslabs.grafana.net/goto/GaWX1nu4g?orgId=1) with 180k workers to create, which makes no sense - but at least with this change it doesn't affect the measurements.

[bchocho]

fix db flag that got changed recently for test tuned for throughput

what was this change?