core/homeassistant/components/recorder/util.py

689 lines
23 KiB
Python

"""SQLAlchemy util functions."""
from __future__ import annotations
from collections.abc import Callable, Generator
from contextlib import contextmanager
from datetime import date, datetime, timedelta
import functools
import logging
import os
import time
from typing import TYPE_CHECKING, Any, TypeVar
from awesomeversion import (
AwesomeVersion,
AwesomeVersionException,
AwesomeVersionStrategy,
)
import ciso8601
from sqlalchemy import text
from sqlalchemy.engine.cursor import CursorFetchStrategy
from sqlalchemy.engine.row import Row
from sqlalchemy.exc import OperationalError, SQLAlchemyError
from sqlalchemy.orm.query import Query
from sqlalchemy.orm.session import Session
from sqlalchemy.sql.lambdas import StatementLambdaElement
from typing_extensions import Concatenate, ParamSpec
import voluptuous as vol
from homeassistant.core import HomeAssistant
from homeassistant.helpers import config_validation as cv
import homeassistant.util.dt as dt_util
from .const import DATA_INSTANCE, SQLITE_URL_PREFIX, SupportedDialect
from .db_schema import (
TABLE_RECORDER_RUNS,
TABLE_SCHEMA_CHANGES,
TABLES_TO_CHECK,
RecorderRuns,
)
from .models import StatisticPeriod, UnsupportedDialect, process_timestamp
if TYPE_CHECKING:
from . import Recorder
_RecorderT = TypeVar("_RecorderT", bound="Recorder")
_P = ParamSpec("_P")
_LOGGER = logging.getLogger(__name__)
RETRIES = 3
QUERY_RETRY_WAIT = 0.1
SQLITE3_POSTFIXES = ["", "-wal", "-shm"]
DEFAULT_YIELD_STATES_ROWS = 32768
MIN_VERSION_MARIA_DB = AwesomeVersion(
"10.3.0", ensure_strategy=AwesomeVersionStrategy.SIMPLEVER
)
MIN_VERSION_MYSQL = AwesomeVersion(
"8.0.0", ensure_strategy=AwesomeVersionStrategy.SIMPLEVER
)
MIN_VERSION_PGSQL = AwesomeVersion(
"12.0", ensure_strategy=AwesomeVersionStrategy.SIMPLEVER
)
MIN_VERSION_SQLITE = AwesomeVersion(
"3.31.0", ensure_strategy=AwesomeVersionStrategy.SIMPLEVER
)
# This is the maximum time after the recorder ends the session
# before we no longer consider startup to be a "restart" and we
# should do a check on the sqlite3 database.
MAX_RESTART_TIME = timedelta(minutes=10)
# Retry when one of the following MySQL errors occurred:
RETRYABLE_MYSQL_ERRORS = (1205, 1206, 1213)
# 1205: Lock wait timeout exceeded; try restarting transaction
# 1206: The total number of locks exceeds the lock table size
# 1213: Deadlock found when trying to get lock; try restarting transaction
FIRST_POSSIBLE_SUNDAY = 8
SUNDAY_WEEKDAY = 6
DAYS_IN_WEEK = 7
@contextmanager
def session_scope(
*,
hass: HomeAssistant | None = None,
session: Session | None = None,
exception_filter: Callable[[Exception], bool] | None = None,
) -> Generator[Session, None, None]:
"""Provide a transactional scope around a series of operations."""
if session is None and hass is not None:
session = get_instance(hass).get_session()
if session is None:
raise RuntimeError("Session required")
need_rollback = False
try:
yield session
if session.get_transaction():
need_rollback = True
session.commit()
except Exception as err: # pylint: disable=broad-except
_LOGGER.error("Error executing query: %s", err)
if need_rollback:
session.rollback()
if not exception_filter or not exception_filter(err):
raise
finally:
session.close()
def commit(session: Session, work: Any) -> bool:
"""Commit & retry work: Either a model or in a function."""
for _ in range(0, RETRIES):
try:
if callable(work):
work(session)
else:
session.add(work)
session.commit()
return True
except OperationalError as err:
_LOGGER.error("Error executing query: %s", err)
session.rollback()
time.sleep(QUERY_RETRY_WAIT)
return False
def execute(
qry: Query, to_native: bool = False, validate_entity_ids: bool = True
) -> list[Row]:
"""Query the database and convert the objects to HA native form.
This method also retries a few times in the case of stale connections.
"""
for tryno in range(0, RETRIES):
try:
timer_start = time.perf_counter()
if to_native:
result = [
row
for row in (
row.to_native(validate_entity_id=validate_entity_ids)
for row in qry
)
if row is not None
]
else:
result = qry.all()
if _LOGGER.isEnabledFor(logging.DEBUG):
elapsed = time.perf_counter() - timer_start
if to_native:
_LOGGER.debug(
"converting %d rows to native objects took %fs",
len(result),
elapsed,
)
else:
_LOGGER.debug(
"querying %d rows took %fs",
len(result),
elapsed,
)
return result
except SQLAlchemyError as err:
_LOGGER.error("Error executing query: %s", err)
if tryno == RETRIES - 1:
raise
time.sleep(QUERY_RETRY_WAIT)
assert False # unreachable # pragma: no cover
def execute_stmt_lambda_element(
session: Session,
stmt: StatementLambdaElement,
start_time: datetime | None = None,
end_time: datetime | None = None,
yield_per: int | None = DEFAULT_YIELD_STATES_ROWS,
) -> list[Row]:
"""Execute a StatementLambdaElement.
If the time window passed is greater than one day
the execution method will switch to yield_per to
reduce memory pressure.
It is not recommended to pass a time window
when selecting non-ranged rows (ie selecting
specific entities) since they are usually faster
with .all().
"""
executed = session.execute(stmt)
use_all = not start_time or ((end_time or dt_util.utcnow()) - start_time).days <= 1
for tryno in range(0, RETRIES):
try:
return executed.all() if use_all else executed.yield_per(yield_per) # type: ignore[no-any-return]
except SQLAlchemyError as err:
_LOGGER.error("Error executing query: %s", err)
if tryno == RETRIES - 1:
raise
time.sleep(QUERY_RETRY_WAIT)
assert False # unreachable # pragma: no cover
def validate_or_move_away_sqlite_database(dburl: str) -> bool:
"""Ensure that the database is valid or move it away."""
dbpath = dburl_to_path(dburl)
if not os.path.exists(dbpath):
# Database does not exist yet, this is OK
return True
if not validate_sqlite_database(dbpath):
move_away_broken_database(dbpath)
return False
return True
def dburl_to_path(dburl: str) -> str:
"""Convert the db url into a filesystem path."""
return dburl[len(SQLITE_URL_PREFIX) :]
def last_run_was_recently_clean(cursor: CursorFetchStrategy) -> bool:
"""Verify the last recorder run was recently clean."""
cursor.execute("SELECT end FROM recorder_runs ORDER BY start DESC LIMIT 1;")
end_time = cursor.fetchone()
if not end_time or not end_time[0]:
return False
last_run_end_time = process_timestamp(dt_util.parse_datetime(end_time[0]))
assert last_run_end_time is not None
now = dt_util.utcnow()
_LOGGER.debug("The last run ended at: %s (now: %s)", last_run_end_time, now)
if last_run_end_time + MAX_RESTART_TIME < now:
return False
return True
def basic_sanity_check(cursor: CursorFetchStrategy) -> bool:
"""Check tables to make sure select does not fail."""
for table in TABLES_TO_CHECK:
if table in (TABLE_RECORDER_RUNS, TABLE_SCHEMA_CHANGES):
cursor.execute(f"SELECT * FROM {table};") # nosec # not injection
else:
cursor.execute(f"SELECT * FROM {table} LIMIT 1;") # nosec # not injection
return True
def validate_sqlite_database(dbpath: str) -> bool:
"""Run a quick check on an sqlite database to see if it is corrupt."""
import sqlite3 # pylint: disable=import-outside-toplevel
try:
conn = sqlite3.connect(dbpath)
run_checks_on_open_db(dbpath, conn.cursor())
conn.close()
except sqlite3.DatabaseError:
_LOGGER.exception("The database at %s is corrupt or malformed", dbpath)
return False
return True
def run_checks_on_open_db(dbpath: str, cursor: CursorFetchStrategy) -> None:
"""Run checks that will generate a sqlite3 exception if there is corruption."""
sanity_check_passed = basic_sanity_check(cursor)
last_run_was_clean = last_run_was_recently_clean(cursor)
if sanity_check_passed and last_run_was_clean:
_LOGGER.debug(
"The system was restarted cleanly and passed the basic sanity check"
)
return
if not sanity_check_passed:
_LOGGER.warning(
"The database sanity check failed to validate the sqlite3 database at %s",
dbpath,
)
if not last_run_was_clean:
_LOGGER.warning(
"The system could not validate that the sqlite3 database at %s was shutdown cleanly",
dbpath,
)
def move_away_broken_database(dbfile: str) -> None:
"""Move away a broken sqlite3 database."""
isotime = dt_util.utcnow().isoformat()
corrupt_postfix = f".corrupt.{isotime}"
_LOGGER.error(
"The system will rename the corrupt database file %s to %s in order to allow startup to proceed",
dbfile,
f"{dbfile}{corrupt_postfix}",
)
for postfix in SQLITE3_POSTFIXES:
path = f"{dbfile}{postfix}"
if not os.path.exists(path):
continue
os.rename(path, f"{path}{corrupt_postfix}")
def execute_on_connection(dbapi_connection: Any, statement: str) -> None:
"""Execute a single statement with a dbapi connection."""
cursor = dbapi_connection.cursor()
cursor.execute(statement)
cursor.close()
def query_on_connection(dbapi_connection: Any, statement: str) -> Any:
"""Execute a single statement with a dbapi connection and return the result."""
cursor = dbapi_connection.cursor()
cursor.execute(statement)
result = cursor.fetchall()
cursor.close()
return result
def _fail_unsupported_dialect(dialect_name: str) -> None:
"""Warn about unsupported database version."""
_LOGGER.error(
"Database %s is not supported; Home Assistant supports %s. "
"Starting with Home Assistant 2022.6 this prevents the recorder from "
"starting. Please migrate your database to a supported software",
dialect_name,
"MariaDB ≥ 10.3, MySQL ≥ 8.0, PostgreSQL ≥ 12, SQLite ≥ 3.31.0",
)
raise UnsupportedDialect
def _fail_unsupported_version(
server_version: str, dialect_name: str, minimum_version: str
) -> None:
"""Warn about unsupported database version."""
_LOGGER.error(
"Version %s of %s is not supported; minimum supported version is %s. "
"Starting with Home Assistant 2022.6 this prevents the recorder from "
"starting. Please upgrade your database software",
server_version,
dialect_name,
minimum_version,
)
raise UnsupportedDialect
def _extract_version_from_server_response(
server_response: str,
) -> AwesomeVersion | None:
"""Attempt to extract version from server response."""
try:
return AwesomeVersion(
server_response,
ensure_strategy=AwesomeVersionStrategy.SIMPLEVER,
find_first_match=True,
)
except AwesomeVersionException:
return None
def _datetime_or_none(value: str) -> datetime | None:
"""Fast version of mysqldb DateTime_or_None.
https://github.com/PyMySQL/mysqlclient/blob/v2.1.0/MySQLdb/times.py#L66
"""
try:
return ciso8601.parse_datetime(value)
except ValueError:
return None
def build_mysqldb_conv() -> dict:
"""Build a MySQLDB conv dict that uses cisco8601 to parse datetimes."""
# Late imports since we only call this if they are using mysqldb
from MySQLdb.constants import ( # pylint: disable=import-outside-toplevel,import-error
FIELD_TYPE,
)
from MySQLdb.converters import ( # pylint: disable=import-outside-toplevel,import-error
conversions,
)
return {**conversions, FIELD_TYPE.DATETIME: _datetime_or_none}
def setup_connection_for_dialect(
instance: Recorder,
dialect_name: str,
dbapi_connection: Any,
first_connection: bool,
) -> AwesomeVersion | None:
"""Execute statements needed for dialect connection."""
version: AwesomeVersion | None = None
if dialect_name == SupportedDialect.SQLITE:
if first_connection:
old_isolation = dbapi_connection.isolation_level
dbapi_connection.isolation_level = None
execute_on_connection(dbapi_connection, "PRAGMA journal_mode=WAL")
dbapi_connection.isolation_level = old_isolation
# WAL mode only needs to be setup once
# instead of every time we open the sqlite connection
# as its persistent and isn't free to call every time.
result = query_on_connection(dbapi_connection, "SELECT sqlite_version()")
version_string = result[0][0]
version = _extract_version_from_server_response(version_string)
if not version or version < MIN_VERSION_SQLITE:
_fail_unsupported_version(
version or version_string, "SQLite", MIN_VERSION_SQLITE
)
# The upper bound on the cache size is approximately 16MiB of memory
execute_on_connection(dbapi_connection, "PRAGMA cache_size = -16384")
#
# Enable FULL synchronous if they have a commit interval of 0
# or NORMAL if they do not.
#
# https://sqlite.org/pragma.html#pragma_synchronous
# The synchronous=NORMAL setting is a good choice for most applications running in WAL mode.
#
synchronous = "NORMAL" if instance.commit_interval else "FULL"
execute_on_connection(dbapi_connection, f"PRAGMA synchronous={synchronous}")
# enable support for foreign keys
execute_on_connection(dbapi_connection, "PRAGMA foreign_keys=ON")
elif dialect_name == SupportedDialect.MYSQL:
execute_on_connection(dbapi_connection, "SET session wait_timeout=28800")
if first_connection:
result = query_on_connection(dbapi_connection, "SELECT VERSION()")
version_string = result[0][0]
version = _extract_version_from_server_response(version_string)
is_maria_db = "mariadb" in version_string.lower()
if is_maria_db:
if not version or version < MIN_VERSION_MARIA_DB:
_fail_unsupported_version(
version or version_string, "MariaDB", MIN_VERSION_MARIA_DB
)
else:
if not version or version < MIN_VERSION_MYSQL:
_fail_unsupported_version(
version or version_string, "MySQL", MIN_VERSION_MYSQL
)
elif dialect_name == SupportedDialect.POSTGRESQL:
if first_connection:
# server_version_num was added in 2006
result = query_on_connection(dbapi_connection, "SHOW server_version")
version_string = result[0][0]
version = _extract_version_from_server_response(version_string)
if not version or version < MIN_VERSION_PGSQL:
_fail_unsupported_version(
version or version_string, "PostgreSQL", MIN_VERSION_PGSQL
)
else:
_fail_unsupported_dialect(dialect_name)
return version
def end_incomplete_runs(session: Session, start_time: datetime) -> None:
"""End any incomplete recorder runs."""
for run in session.query(RecorderRuns).filter_by(end=None):
run.closed_incorrect = True
run.end = start_time
_LOGGER.warning(
"Ended unfinished session (id=%s from %s)", run.run_id, run.start
)
session.add(run)
def retryable_database_job(
description: str,
) -> Callable[
[Callable[Concatenate[_RecorderT, _P], bool]],
Callable[Concatenate[_RecorderT, _P], bool],
]:
"""Try to execute a database job.
The job should return True if it finished, and False if it needs to be rescheduled.
"""
def decorator(
job: Callable[Concatenate[_RecorderT, _P], bool]
) -> Callable[Concatenate[_RecorderT, _P], bool]:
@functools.wraps(job)
def wrapper(instance: _RecorderT, *args: _P.args, **kwargs: _P.kwargs) -> bool:
try:
return job(instance, *args, **kwargs)
except OperationalError as err:
assert instance.engine is not None
if (
instance.engine.dialect.name == SupportedDialect.MYSQL
and err.orig
and err.orig.args[0] in RETRYABLE_MYSQL_ERRORS
):
_LOGGER.info(
"%s; %s not completed, retrying", err.orig.args[1], description
)
time.sleep(instance.db_retry_wait)
# Failed with retryable error
return False
_LOGGER.warning("Error executing %s: %s", description, err)
# Failed with permanent error
return True
return wrapper
return decorator
def periodic_db_cleanups(instance: Recorder) -> None:
"""Run any database cleanups that need to happen periodically.
These cleanups will happen nightly or after any purge.
"""
assert instance.engine is not None
if instance.engine.dialect.name == SupportedDialect.SQLITE:
# Execute sqlite to create a wal checkpoint and free up disk space
_LOGGER.debug("WAL checkpoint")
with instance.engine.connect() as connection:
connection.execute(text("PRAGMA wal_checkpoint(TRUNCATE);"))
@contextmanager
def write_lock_db_sqlite(instance: Recorder) -> Generator[None, None, None]:
"""Lock database for writes."""
assert instance.engine is not None
with instance.engine.connect() as connection:
# Execute sqlite to create a wal checkpoint
# This is optional but makes sure the backup is going to be minimal
connection.execute(text("PRAGMA wal_checkpoint(TRUNCATE)"))
# Create write lock
_LOGGER.debug("Lock database")
connection.execute(text("BEGIN IMMEDIATE;"))
try:
yield
finally:
_LOGGER.debug("Unlock database")
connection.execute(text("END;"))
def async_migration_in_progress(hass: HomeAssistant) -> bool:
"""Determine if a migration is in progress.
This is a thin wrapper that allows us to change
out the implementation later.
"""
if DATA_INSTANCE not in hass.data:
return False
instance = get_instance(hass)
return instance.migration_in_progress
def async_migration_is_live(hass: HomeAssistant) -> bool:
"""Determine if a migration is live.
This is a thin wrapper that allows us to change
out the implementation later.
"""
if DATA_INSTANCE not in hass.data:
return False
instance: Recorder = hass.data[DATA_INSTANCE]
return instance.migration_is_live
def second_sunday(year: int, month: int) -> date:
"""Return the datetime.date for the second sunday of a month."""
second = date(year, month, FIRST_POSSIBLE_SUNDAY)
day_of_week = second.weekday()
if day_of_week == SUNDAY_WEEKDAY:
return second
return second.replace(
day=(FIRST_POSSIBLE_SUNDAY + (SUNDAY_WEEKDAY - day_of_week) % DAYS_IN_WEEK)
)
def is_second_sunday(date_time: datetime) -> bool:
"""Check if a time is the second sunday of the month."""
return bool(second_sunday(date_time.year, date_time.month).day == date_time.day)
def get_instance(hass: HomeAssistant) -> Recorder:
"""Get the recorder instance."""
instance: Recorder = hass.data[DATA_INSTANCE]
return instance
PERIOD_SCHEMA = vol.Schema(
{
vol.Exclusive("calendar", "period"): vol.Schema(
{
vol.Required("period"): vol.Any("hour", "day", "week", "month", "year"),
vol.Optional("offset"): int,
}
),
vol.Exclusive("fixed_period", "period"): vol.Schema(
{
vol.Optional("start_time"): vol.All(cv.datetime, dt_util.as_utc),
vol.Optional("end_time"): vol.All(cv.datetime, dt_util.as_utc),
}
),
vol.Exclusive("rolling_window", "period"): vol.Schema(
{
vol.Required("duration"): cv.time_period_dict,
vol.Optional("offset"): cv.time_period_dict,
}
),
}
)
def resolve_period(
period_def: StatisticPeriod,
) -> tuple[datetime | None, datetime | None]:
"""Return start and end datetimes for a statistic period definition."""
start_time = None
end_time = None
if "calendar" in period_def:
calendar_period = period_def["calendar"]["period"]
start_of_day = dt_util.start_of_local_day()
cal_offset = period_def["calendar"].get("offset", 0)
if calendar_period == "hour":
start_time = dt_util.now().replace(minute=0, second=0, microsecond=0)
start_time += timedelta(hours=cal_offset)
end_time = start_time + timedelta(hours=1)
elif calendar_period == "day":
start_time = start_of_day
start_time += timedelta(days=cal_offset)
end_time = start_time + timedelta(days=1)
elif calendar_period == "week":
start_time = start_of_day - timedelta(days=start_of_day.weekday())
start_time += timedelta(days=cal_offset * 7)
end_time = start_time + timedelta(weeks=1)
elif calendar_period == "month":
start_time = start_of_day.replace(day=28)
# This works for up to 48 months of offset
start_time = (start_time + timedelta(days=cal_offset * 31)).replace(day=1)
end_time = (start_time + timedelta(days=31)).replace(day=1)
else: # calendar_period = "year"
start_time = start_of_day.replace(month=12, day=31)
# This works for 100+ years of offset
start_time = (start_time + timedelta(days=cal_offset * 366)).replace(
month=1, day=1
)
end_time = (start_time + timedelta(days=365)).replace(day=1)
start_time = dt_util.as_utc(start_time)
end_time = dt_util.as_utc(end_time)
elif "fixed_period" in period_def:
start_time = period_def["fixed_period"].get("start_time")
end_time = period_def["fixed_period"].get("end_time")
elif "rolling_window" in period_def:
duration = period_def["rolling_window"]["duration"]
now = dt_util.utcnow()
start_time = now - duration
end_time = start_time + duration
if offset := period_def["rolling_window"].get("offset"):
start_time += offset
end_time += offset
return (start_time, end_time)