mirror of
https://github.com/blakeblackshear/frigate.git
synced 2024-11-26 19:06:11 +01:00
ab50d0b006
* Add isort and ruff linter Both linters are pretty common among modern python code bases. The isort tool provides stable sorting and grouping, as well as pruning of unused imports. Ruff is a modern linter, that is very fast due to being written in rust. It can detect many common issues in a python codebase. Removes the pylint dev requirement, since ruff replaces it. * treewide: fix issues detected by ruff * treewide: fix bare except clauses * .devcontainer: Set up isort * treewide: optimize imports * treewide: apply black * treewide: make regex patterns raw strings This is necessary for escape sequences to be properly recognized.
192 lines
7.2 KiB
Python
192 lines
7.2 KiB
Python
"""Handle storage retention and usage."""
|
|
|
|
import logging
|
|
import shutil
|
|
import threading
|
|
from pathlib import Path
|
|
|
|
from peewee import fn
|
|
|
|
from frigate.config import FrigateConfig
|
|
from frigate.const import RECORD_DIR
|
|
from frigate.models import Event, Recordings
|
|
|
|
logger = logging.getLogger(__name__)
|
|
bandwidth_equation = Recordings.segment_size / (
|
|
Recordings.end_time - Recordings.start_time
|
|
)
|
|
|
|
|
|
class StorageMaintainer(threading.Thread):
|
|
"""Maintain frigates recording storage."""
|
|
|
|
def __init__(self, config: FrigateConfig, stop_event) -> None:
|
|
threading.Thread.__init__(self)
|
|
self.name = "storage_maintainer"
|
|
self.config = config
|
|
self.stop_event = stop_event
|
|
self.camera_storage_stats: dict[str, dict] = {}
|
|
|
|
def calculate_camera_bandwidth(self) -> None:
|
|
"""Calculate an average MB/hr for each camera."""
|
|
for camera in self.config.cameras.keys():
|
|
# cameras with < 50 segments should be refreshed to keep size accurate
|
|
# when few segments are available
|
|
if self.camera_storage_stats.get(camera, {}).get("needs_refresh", True):
|
|
self.camera_storage_stats[camera] = {
|
|
"needs_refresh": (
|
|
Recordings.select(fn.COUNT(Recordings.id))
|
|
.where(
|
|
Recordings.camera == camera, Recordings.segment_size != 0
|
|
)
|
|
.scalar()
|
|
< 50
|
|
)
|
|
}
|
|
|
|
# calculate MB/hr
|
|
try:
|
|
bandwidth = round(
|
|
Recordings.select(fn.AVG(bandwidth_equation))
|
|
.where(Recordings.camera == camera, Recordings.segment_size != 0)
|
|
.limit(100)
|
|
.scalar()
|
|
* 3600,
|
|
2,
|
|
)
|
|
except TypeError:
|
|
bandwidth = 0
|
|
|
|
self.camera_storage_stats[camera]["bandwidth"] = bandwidth
|
|
logger.debug(f"{camera} has a bandwidth of {bandwidth} MB/hr.")
|
|
|
|
def calculate_camera_usages(self) -> dict[str, dict]:
|
|
"""Calculate the storage usage of each camera."""
|
|
usages: dict[str, dict] = {}
|
|
|
|
for camera in self.config.cameras.keys():
|
|
camera_storage = (
|
|
Recordings.select(fn.SUM(Recordings.segment_size))
|
|
.where(Recordings.camera == camera, Recordings.segment_size != 0)
|
|
.scalar()
|
|
)
|
|
|
|
usages[camera] = {
|
|
"usage": camera_storage,
|
|
"bandwidth": self.camera_storage_stats.get(camera, {}).get(
|
|
"bandwidth", 0
|
|
),
|
|
}
|
|
|
|
return usages
|
|
|
|
def check_storage_needs_cleanup(self) -> bool:
|
|
"""Return if storage needs cleanup."""
|
|
# currently runs cleanup if less than 1 hour of space is left
|
|
# disk_usage should not spin up disks
|
|
hourly_bandwidth = sum(
|
|
[b["bandwidth"] for b in self.camera_storage_stats.values()]
|
|
)
|
|
remaining_storage = round(shutil.disk_usage(RECORD_DIR).free / 1000000, 1)
|
|
logger.debug(
|
|
f"Storage cleanup check: {hourly_bandwidth} hourly with remaining storage: {remaining_storage}."
|
|
)
|
|
return remaining_storage < hourly_bandwidth
|
|
|
|
def reduce_storage_consumption(self) -> None:
|
|
"""Remove oldest hour of recordings."""
|
|
logger.debug("Starting storage cleanup.")
|
|
deleted_segments_size = 0
|
|
hourly_bandwidth = sum(
|
|
[b["bandwidth"] for b in self.camera_storage_stats.values()]
|
|
)
|
|
|
|
recordings: Recordings = Recordings.select().order_by(
|
|
Recordings.start_time.asc()
|
|
)
|
|
retained_events: Event = (
|
|
Event.select()
|
|
.where(
|
|
Event.retain_indefinitely is True,
|
|
Event.has_clip,
|
|
)
|
|
.order_by(Event.start_time.asc())
|
|
.objects()
|
|
)
|
|
|
|
event_start = 0
|
|
deleted_recordings = set()
|
|
for recording in recordings.objects().iterator():
|
|
# check if 1 hour of storage has been reclaimed
|
|
if deleted_segments_size > hourly_bandwidth:
|
|
break
|
|
|
|
keep = False
|
|
|
|
# Now look for a reason to keep this recording segment
|
|
for idx in range(event_start, len(retained_events)):
|
|
event = retained_events[idx]
|
|
|
|
# if the event starts in the future, stop checking events
|
|
# and let this recording segment expire
|
|
if event.start_time > recording.end_time:
|
|
keep = False
|
|
break
|
|
|
|
# if the event is in progress or ends after the recording starts, keep it
|
|
# and stop looking at events
|
|
if event.end_time is None or event.end_time >= recording.start_time:
|
|
keep = True
|
|
break
|
|
|
|
# if the event ends before this recording segment starts, skip
|
|
# this event and check the next event for an overlap.
|
|
# since the events and recordings are sorted, we can skip events
|
|
# that end before the previous recording segment started on future segments
|
|
if event.end_time < recording.start_time:
|
|
event_start = idx
|
|
|
|
# Delete recordings not retained indefinitely
|
|
if not keep:
|
|
deleted_segments_size += recording.segment_size
|
|
Path(recording.path).unlink(missing_ok=True)
|
|
deleted_recordings.add(recording.id)
|
|
|
|
# check if need to delete retained segments
|
|
if deleted_segments_size < hourly_bandwidth:
|
|
logger.error(
|
|
f"Could not clear {hourly_bandwidth} currently {deleted_segments_size}, retained recordings must be deleted."
|
|
)
|
|
recordings = Recordings.select().order_by(Recordings.start_time.asc())
|
|
|
|
for recording in recordings.objects().iterator():
|
|
if deleted_segments_size > hourly_bandwidth:
|
|
break
|
|
|
|
deleted_segments_size += recording.segment_size
|
|
Path(recording.path).unlink(missing_ok=True)
|
|
deleted_recordings.add(recording.id)
|
|
|
|
logger.debug(f"Expiring {len(deleted_recordings)} recordings")
|
|
# delete up to 100,000 at a time
|
|
max_deletes = 100000
|
|
deleted_recordings_list = list(deleted_recordings)
|
|
for i in range(0, len(deleted_recordings_list), max_deletes):
|
|
Recordings.delete().where(
|
|
Recordings.id << deleted_recordings_list[i : i + max_deletes]
|
|
).execute()
|
|
|
|
def run(self):
|
|
"""Check every 5 minutes if storage needs to be cleaned up."""
|
|
while not self.stop_event.wait(300):
|
|
if not self.camera_storage_stats or True in [
|
|
r["needs_refresh"] for r in self.camera_storage_stats.values()
|
|
]:
|
|
self.calculate_camera_bandwidth()
|
|
logger.debug(f"Default camera bandwidths: {self.camera_storage_stats}.")
|
|
|
|
if self.check_storage_needs_cleanup():
|
|
self.reduce_storage_consumption()
|
|
|
|
logger.info("Exiting storage maintainer...")
|