Hi,
I ran into a very strange situation: a cron job runs every night and is supposed to delete form old data from a table. It turns out that for a few weeks now the query runs, takes a lot of time (as the table kept growing), reports the correct number of affected rows, but it does not actually delete the rows.
Here is the slow-query log for such a query:
# Time: 2019-04-17T03:13:19.714146Z
# User@Host: root[root] @ [10.70.222.8] Id: 11005891
# Schema: live_indicators Last_errno: 0 Killed: 0
# Query_time: 11552.742740 Lock_time: 0.000092 Rows_sent: 0 Rows_examined: 16026170 Rows_affected: 10231900
# Bytes_sent: 14
SET timestamp=1555470799;
DELETE FROM `reports_hourly` WHERE (start_date < '2019-03-03 00:00:00');
If I connect run the query from the client (I only ran it with limit, as this is a master and it would make replication lag) and it works ok.
Any idea what might be causing this or where to dig further?
Thank you,
Alex
I ran into a very strange situation: a cron job runs every night and is supposed to delete form old data from a table. It turns out that for a few weeks now the query runs, takes a lot of time (as the table kept growing), reports the correct number of affected rows, but it does not actually delete the rows.
Here is the slow-query log for such a query:
# Time: 2019-04-17T03:13:19.714146Z
# User@Host: root[root] @ [10.70.222.8] Id: 11005891
# Schema: live_indicators Last_errno: 0 Killed: 0
# Query_time: 11552.742740 Lock_time: 0.000092 Rows_sent: 0 Rows_examined: 16026170 Rows_affected: 10231900
# Bytes_sent: 14
SET timestamp=1555470799;
DELETE FROM `reports_hourly` WHERE (start_date < '2019-03-03 00:00:00');
If I connect run the query from the client (I only ran it with limit, as this is a master and it would make replication lag) and it works ok.
Any idea what might be causing this or where to dig further?
Thank you,
Alex