Merge pull request #6135 from matrix-org/erikj/fixup_devices_last_seen_query

Fix `devices_last_seen` background update.
This commit is contained in:
Erik Johnston 2019-10-01 10:16:57 +01:00 committed by GitHub
commit 479fbac96f
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 55 additions and 7 deletions

1
changelog.d/6135.bugfix Normal file
View File

@ -0,0 +1 @@
Fix bug in background update that adds last seen information to the `devices` table, and improve its performance on Postgres.

View File

@ -463,14 +463,46 @@ class ClientIpStore(background_updates.BackgroundUpdateStore):
last_device_id = progress.get("last_device_id", "") last_device_id = progress.get("last_device_id", "")
def _devices_last_seen_update_txn(txn): def _devices_last_seen_update_txn(txn):
# This consists of two queries:
#
# 1. The sub-query searches for the next N devices and joins
# against user_ips to find the max last_seen associated with
# that device.
# 2. The outer query then joins again against user_ips on
# user/device/last_seen. This *should* hopefully only
# return one row, but if it does return more than one then
# we'll just end up updating the same device row multiple
# times, which is fine.
if self.database_engine.supports_tuple_comparison:
where_clause = "(user_id, device_id) > (?, ?)"
where_args = [last_user_id, last_device_id]
else:
# We explicitly do a `user_id >= ? AND (...)` here to ensure
# that an index is used, as doing `user_id > ? OR (user_id = ? AND ...)`
# makes it hard for query optimiser to tell that it can use the
# index on user_id
where_clause = "user_id >= ? AND (user_id > ? OR device_id > ?)"
where_args = [last_user_id, last_user_id, last_device_id]
sql = """ sql = """
SELECT u.last_seen, u.ip, u.user_agent, user_id, device_id FROM devices SELECT
last_seen, ip, user_agent, user_id, device_id
FROM (
SELECT
user_id, device_id, MAX(u.last_seen) AS last_seen
FROM devices
INNER JOIN user_ips AS u USING (user_id, device_id) INNER JOIN user_ips AS u USING (user_id, device_id)
WHERE user_id > ? OR (user_id = ? AND device_id > ?) WHERE %(where_clause)s
GROUP BY user_id, device_id
ORDER BY user_id ASC, device_id ASC ORDER BY user_id ASC, device_id ASC
LIMIT ? LIMIT ?
""" ) c
txn.execute(sql, (last_user_id, last_user_id, last_device_id, batch_size)) INNER JOIN user_ips AS u USING (user_id, device_id, last_seen)
""" % {
"where_clause": where_clause
}
txn.execute(sql, where_args + [batch_size])
rows = txn.fetchall() rows = txn.fetchall()
if not rows: if not rows:

View File

@ -72,6 +72,13 @@ class PostgresEngine(object):
""" """
return True return True
@property
def supports_tuple_comparison(self):
"""
Do we support comparing tuples, i.e. `(a, b) > (c, d)`?
"""
return True
def is_deadlock(self, error): def is_deadlock(self, error):
if isinstance(error, self.module.DatabaseError): if isinstance(error, self.module.DatabaseError):
# https://www.postgresql.org/docs/current/static/errcodes-appendix.html # https://www.postgresql.org/docs/current/static/errcodes-appendix.html

View File

@ -38,6 +38,14 @@ class Sqlite3Engine(object):
""" """
return self.module.sqlite_version_info >= (3, 24, 0) return self.module.sqlite_version_info >= (3, 24, 0)
@property
def supports_tuple_comparison(self):
"""
Do we support comparing tuples, i.e. `(a, b) > (c, d)`? This requires
SQLite 3.15+.
"""
return self.module.sqlite_version_info >= (3, 15, 0)
def check_database(self, txn): def check_database(self, txn):
pass pass