1. fixed some error regressions (results error with big nstance_num due to incorrect order by). 2. if set parallel_fragment_exec_instance_num to 0, the concurrency in the Pipeline execution engine will automatically be set to half of the number of CPU cores. 3. add limit to parallel_fragment_exec_instance_num that it cannot be set to more than fe.conf::max_instance_num(Default: 128) ``` mysql [(none)]>set parallel_fragment_exec_instance_num = 514; ERROR 1231 (42000): errCode = 2, detailMessage = Variable 'parallel_fragment_exec_instance_num' can't be set to the value of '514(Should not be set to more than 128)' ```
7 lines
518 B
SQL
7 lines
518 B
SQL
SELECT count() FROM test_btc_json;
|
|
SELECT avg(fee) FROM test_btc_json;
|
|
SELECT avg(size(`inputs.prev_out.spending_outpoints.n`)) FROM test_btc_json;
|
|
SELECT avg(size(`inputs.prev_out.spending_outpoints.tx_index`)) FROM test_btc_json;
|
|
select `inputs.prev_out.spending_outpoints.tx_index`, fee from test_btc_json order by fee,hash;
|
|
select `out.tx_index`[-1] from test_btc_json order by hash,`out.tx_index`[-1];
|
|
select `out.tx_index`, fee, `out.value`[1] from test_btc_json where array_contains(`out.value`, 2450939412); |