normalise reaping query

This commit is contained in:
Neil Johnson 2018-08-01 23:24:38 +01:00
parent ec716a35b2
commit c21d82bab3
1 changed files with 38 additions and 3 deletions

View File

@ -7,6 +7,7 @@ class MonthlyActiveUsersStore(SQLBaseStore):
def __init__(self, hs):
super(MonthlyActiveUsersStore, self).__init__(None, hs)
self._clock = hs.get_clock()
self.max_mau_value = hs.config.max_mau_value
def reap_monthly_active_users(self):
"""
@ -19,8 +20,42 @@ class MonthlyActiveUsersStore(SQLBaseStore):
thirty_days_ago = (
int(self._clock.time_msec()) - (1000 * 60 * 60 * 24 * 30)
)
sql = "DELETE FROM monthly_active_users WHERE timestamp < ?"
txn.execute(sql, (thirty_days_ago,))
# Query deletes the union of users that have either:
# * not visited in the last 30 days
# * exceeded the total max_mau_value threshold. Where there is
# an excess, more recent users are favoured - this is to cover
# the case where the limit has been step change reduced.
#
sql = """
DELETE FROM monthly_active_users
WHERE user_id
IN (
SELECT * FROM (
SELECT monthly_active_users.user_id
FROM monthly_active_users
LEFT JOIN (
SELECT user_id, max(last_seen) AS last_seen
FROM user_ips
GROUP BY user_id
) AS uip ON uip.user_id=monthly_active_users.user_id
ORDER BY uip.last_seen desc LIMIT -1 OFFSET ?
)
UNION
SELECT * FROM (
SELECT monthly_active_users.user_id
FROM monthly_active_users
LEFT JOIN (
SELECT user_id, max(last_seen) AS last_seen
FROM user_ips
GROUP BY user_id
) AS uip ON uip.user_id=monthly_active_users.user_id
WHERE uip.last_seen < ?
)
)
"""
txn.execute(sql, (self.max_mau_value, thirty_days_ago,))
return self.runInteraction("reap_monthly_active_users", _reap_users)
@ -45,7 +80,7 @@ class MonthlyActiveUsersStore(SQLBaseStore):
user_id (str): user to add/update
"""
return self._simple_insert(
desc="upsert_monthly_active_user",
desc="insert_monthly_active_user",
table="monthly_active_users",
values={
"user_id": user_id,