Improve DB streaming perf for cracking and statistics

This commit is contained in:
Jack Kingsman
2026-03-30 21:31:59 -07:00
parent 5c60559cb8
commit 43abcd07b2
8 changed files with 203 additions and 59 deletions
+23
View File
@@ -190,6 +190,29 @@ class TestDebugEndpoint:
assert payload["database"]["total_channel_messages"] == 1
assert payload["database"]["total_outgoing"] == 1
@pytest.mark.asyncio
async def test_support_snapshot_uses_lightweight_message_totals(self, test_db, client):
"""Debug snapshot should not call the full statistics aggregation."""
with (
patch(
"app.routers.debug.StatisticsRepository.get_all",
new=AsyncMock(side_effect=AssertionError("get_all should not be called")),
),
patch(
"app.routers.debug.StatisticsRepository.get_database_message_totals",
new=AsyncMock(
return_value={
"total_dms": 0,
"total_channel_messages": 0,
"total_outgoing": 0,
}
),
),
):
response = await client.get("/api/debug")
assert response.status_code == 200
class TestRadioDisconnectedHandler:
"""Test that RadioDisconnectedError maps to 503."""
+38 -1
View File
@@ -5,7 +5,7 @@ undecrypted count endpoint, and the maintenance endpoint.
"""
import time
from unittest.mock import patch
from unittest.mock import AsyncMock, patch
import pytest
@@ -305,6 +305,43 @@ class TestDecryptHistoricalPackets:
assert "key_type" in data["detail"].lower()
class TestUndecryptedTextPacketStreaming:
@pytest.mark.asyncio
async def test_count_undecrypted_text_messages_uses_batched_streaming(self, test_db):
"""Counting undecrypted DM packets should stream batches and filter by payload type."""
class FakeCursor:
def __init__(self):
self._batches = [
[
{"id": 1, "data": b"\x09\x00dm", "timestamp": 1000},
{"id": 2, "data": b"\x15\x00chan", "timestamp": 1001},
],
[{"id": 3, "data": b"\x09\x00dm2", "timestamp": 1002}],
[],
]
self.fetchall_called = False
async def fetchmany(self, size):
assert size > 0
return self._batches.pop(0)
async def close(self):
return None
async def fetchall(self):
self.fetchall_called = True
raise AssertionError("fetchall() should not be used")
fake_cursor = FakeCursor()
with patch.object(test_db.conn, "execute", new=AsyncMock(return_value=fake_cursor)):
count = await RawPacketRepository.count_undecrypted_text_messages(batch_size=2)
assert fake_cursor.fetchall_called is False
assert count == 2
class TestRunHistoricalChannelDecryption:
"""Test the _run_historical_channel_decryption background task."""
+47
View File
@@ -1,6 +1,7 @@
"""Tests for the statistics repository and endpoint."""
import time
from types import SimpleNamespace
from unittest.mock import AsyncMock, patch
import pytest
@@ -349,6 +350,52 @@ class TestPathHashWidthStats:
assert breakdown["double_byte_pct"] == pytest.approx(100 / 3, rel=1e-3)
assert breakdown["triple_byte_pct"] == pytest.approx(100 / 3, rel=1e-3)
@pytest.mark.asyncio
async def test_path_hash_width_scan_uses_batched_fetchmany(self, test_db):
"""Hash-width stats should stream batches instead of calling fetchall()."""
class FakeCursor:
def __init__(self):
self._batches = [
[{"data": b"a"}, {"data": b"b"}],
[{"data": b"c"}],
[],
]
self.fetchall_called = False
async def fetchmany(self, size):
assert size > 0
return self._batches.pop(0)
async def fetchall(self):
self.fetchall_called = True
raise AssertionError("fetchall() should not be used")
fake_cursor = FakeCursor()
def fake_parse(raw_packet: bytes):
hash_sizes = {
b"a": 1,
b"b": 2,
b"c": 3,
}
hash_size = hash_sizes.get(raw_packet)
if hash_size is None:
return None
return SimpleNamespace(hash_size=hash_size)
with (
patch.object(test_db.conn, "execute", new=AsyncMock(return_value=fake_cursor)),
patch("app.repository.settings.parse_packet_envelope", side_effect=fake_parse),
):
breakdown = await StatisticsRepository._path_hash_width_24h()
assert fake_cursor.fetchall_called is False
assert breakdown["total_packets"] == 3
assert breakdown["single_byte"] == 1
assert breakdown["double_byte"] == 1
assert breakdown["triple_byte"] == 1
class TestStatisticsEndpoint:
@pytest.mark.asyncio