Just a quick question...
Consider this scenario: a table with a timestamp column like this:
CREATE TABLE foo (
int key,
int foo,
timestamp last_update,
PRIMARY KEY (key),
KEY (last_update)
) TYPE=MyISAM;
and a monitoring script that does this every few seconds:
SELECT key, foo, last_update
FROM foo
WHERE last_update > ?
AND last_update < NOW() -- avoid edge case
to select recent changes (the placeholder is set to the maximum
value of last_update from the previous execution of the select).
Multiple monitor processes can be running and it all works well and
efficiently. No changes are missed. No problems at all. Great.
If you've used this technique with MySQL please email (just me) a
quick reply saying YES. I want to get a feel for how common it is
before I discuss some potential issues. I'll summarize to the list
in a few days.
Thanks.
Tim.
