RingsC
commented
6 months ago It's not a deadlock.
the root cause is here: need to imp the records function to get the real rows count of a rapid table. and optimization does not work rightly.
/*
Try to optimize count(*), min() and max() to const fields if
there is implicit grouping (aggregate functions but no
group_list). In this case, the result set shall only contain one
row.
*/
if (tables_list && implicit_grouping &&
!(query_block->active_options() & OPTION_NO_CONST_TABLES)) {
aggregate_evaluated outcome;
if (optimize_aggregated_query(thd, query_block, *fields, where_cond,
&outcome)) {
!(query_block->active_options() & OPTION_NO_CONST_TABLES) return false. and in ShannonPrepareSecondaryEngine, it set to OPTION_NO_CONST_TABLES
Search before asking
Version
ver0.2.0-beta
What's Wrong?
session 1: alter table tt secondary_load;
session2 : insert into tt values(xxxx); batch insertion.
then in session 1: execute select statement, a dead lock occurs.
when pop the changes into rapid, and now, executing some queries on rapid. it could occur deadlock.
How to Reproduce?
1: session1: CALL insert_into_tt1(); //trying to insert 1000rows into a table.
2: session2: select count(1) from tt1;
Are you willing to submit PR?