4 Sep
2022
4 Sep
'22
1:12 p.m.
Kristian Nielsen <knielsen@knielsen-hq.org> writes:
BTW, the testcase only fails sporadically (without the fix), becase it depends on whether the SQL thread has had time to read ahead to a new FD when the CREATE TABLE t1 runs in the worker thread. There's a wait in the testcase which seems to be intended to ensure this condition:
# Wait for the last event recieve before to unlock t2 --let $wait_condition= SELECT COUNT(*) > 0 FROM information_schema.processlist WHERE state = "Waiting for worker threads to be idle"
But apparently it's ineffective, I had to run the test ~50 times for the error to trigger. Just mentioning it, maybe it's worth spending time on.
Erhm - i meant: "maybe it's *not* worth spending time on". - Kristian.