No flush for lower ranges.

[gusinacio]

I'm trying to execute the app by setting the following range: 17893158:17893162. The substreams finishes with the following logs and no flush:

2023-09-13T12:13:46.122-0300 INFO (sink-postgres) sinker from CLI {"endpoint": "mainnet.eth.streamingfast.io:443", "manifest_path": "substreams.spkg", "params": [], "output_module_name": "db_out", "expected_module_type": "sf.substreams.sink.database.v1.DatabaseChanges,sf.substreams.database.v1.DatabaseChanges", "block_range": "17893158:17893162", "development_mode": false, "infinite_retry": false, "final_blocks_only": false, "skip_package_validation": false, "live_block_time_delta": "5m0s", "undo_buffer_size": 12, "extra_headers": []}
2023-09-13T12:13:46.122-0300 INFO (sink-postgres) reading substreams manifest {"manifest_path": "substreams.spkg"}
2023-09-13T12:13:46.124-0300 INFO (sink-postgres) finding output module {"module_name": "db_out"}
2023-09-13T12:13:46.124-0300 INFO (sink-postgres) validating output module type {"module_name": "db_out", "module_type": "proto:sf.substreams.sink.database.v1.DatabaseChanges"}
2023-09-13T12:13:46.124-0300 INFO (sink-postgres) sinker configured {"mode": "Production", "module_count": 1, "output_module_name": "db_out", "output_module_type": "proto:sf.substreams.sink.database.v1.DatabaseChanges", "output_module_hash": "cb430a7f31a05e6057d087f6e62410865cd0a1b9", "client_config": "mainnet.eth.streamingfast.io:443 (insecure: false, plaintext: false, JWT present: true)", "buffer": "Buffering (12 blocks)", "block_range": "[17893158, 17893162)", "infinite_retry": false, "final_blocks_only": false, "liveness_checker": true}
2023-09-13T12:13:46.126-0300 INFO (sink-postgres) starting postgres sink {"stats_refresh_each": "15s", "restarting_at": "None", "database": "uniswap_v2_ethereum", "schema": "uniswap_v2_ethereum"}
2023-09-13T12:13:46.126-0300 INFO (sink-postgres) starting sinker {"stats_refresh_each": "15s", "end_at": "#17893173"}
2023-09-13T12:13:46.530-0300 INFO (sink-postgres) session initialized with remote endpoint {"max_parallel_workers": 10, "linear_handoff_block": 17893174, "resolved_start_block": 17893158, "trace_id": "0a483664a47a455929a13b21420c891f"}
2023-09-13T12:13:47.097-0300 INFO (sink-postgres) substreams ended correctly, reached your stop block {"last_block_seen": "#17893173 (8d6a8f90dce5f10e27b6763fc2c3bf267a0f5549e09dee0f9024b0833bdbf12a)"}
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) postgres sink stats {"db_flush_rate": "NaN flush/s (0 total)", "flushed_entries": 0, "last_block": "<Unset>"}
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) postgres sinker terminating {"last_block_written": "<Unset>"}
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) sinker terminating
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) run terminating {"from_signal": false, "with_error": false}
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) waiting for run termination
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) run terminated gracefully
2023-09-13T12:13:47.098-0300 INFO (sink-postgres) substreams stream stats {"data_msg_rate": "16.000 msg/s (16 total)", "progress_last_block": {}, "progress_running_jobs": {}, "progress_total_processed_blocks": 0, "progress_last_contiguous_block": {}, "undo_msg_rate": "0.000 msg/s (0 total)", "last_block": "#17893173 (8d6a8f90dce5f10e27b6763fc2c3bf267a0f5549e09dee0f9024b0833bdbf12a)"}

The same happens for both Clickhouse and Postgres.

I thought it could be something related to the undo signal, but trying with ranges bigger than 12 blocks but within the 1000 blocks did the same result.