Iceberg optimizer timeout about trino

[martint]

This should be fixed in master (to be released in 443). Are you able to build the latest snapshot and try it to see if it works for you now?

[findinpath]

Tried with 440 the following scenario:

trino> create table iceberg.default.t1(id integer, data varchar);
CREATE TABLE
trino> select * from iceberg.default.t1;
 id | data 
----+------
(0 rows)

trino> use iceberg.default;
USE
trino:default> insert into t1 values (1, 'A');
INSERT: 1 row

trino:default> insert into t1 values (2, 'B');
INSERT: 1 row

-- trying with a filter having 2K values
trino:default: select * from t1 where id in (1, 2, ...., 2000);

Query 20240320_115100_00007_d5q3m, FINISHED, 1 node
http://localhost:8080/ui/query.html?20240320_115100_00007_d5q3m
Splits: 2 total, 2 done (100.00%)
CPU Time: 0.0s total,    47 rows/s, 14.5KB/s, 18% active
Per Node: 0.0 parallelism,     1 rows/s,   314B/s
Parallelism: 0.0
Peak Memory: 320B
1.99 [2 rows, 624B] [1 rows/s, 314B/s]

-- trying with a filter having 20K values
trino:default: select * from t1 where id in (1, 2, ...., 20000);
 id | data 
----+------
  2 | B    
  1 | A    
(2 rows)

Query 20240320_115145_00008_d5q3m, FINISHED, 1 node
http://localhost:8080/ui/query.html?20240320_115145_00008_d5q3m
Splits: 2 total, 2 done (100.00%)
CPU Time: 0.0s total,   105 rows/s, 32.1KB/s, 12% active
Per Node: 0.0 parallelism,     0 rows/s,     5B/s
Parallelism: 0.0
Peak Memory: 320B
2:07 [2 rows, 624B] [0 rows/s, 5B/s]

Trying on 442 the same scenario

-- trying with a filter having 20K values
trino:default: select * from t1 where id in (1, 2, ...., 20000);
 id | data 
----+------
  2 | B    
  1 | A    
(2 rows)

Query 20240320_120535_00003_dz7bu, FINISHED, 1 node
http://localhost:8080/ui/query.html?20240320_120535_00003_dz7bu
Splits: 2 total, 2 done (100.00%)
CPU Time: 0.1s total,    18 rows/s, 5.64KB/s, 55% active
Per Node: 0.0 parallelism,     0 rows/s,     4B/s
Parallelism: 0.0
Peak Memory: 320B
2:37 [2 rows, 624B] [0 rows/s, 4B/s]

Trying on master the same scenario

-- trying with a filter having 20K values
trino:default: select * from t1 where id in (1, 2, ...., 20000);
 id | data 
----+------
  2 | B    
  1 | A    
(2 rows)
Query 20240320_120148_00003_g5krv, FINISHED, 1 node
http://localhost:8080/ui/query.html?20240320_120148_00003_g5krv
Splits: 2 total, 2 done (100.00%)
CPU Time: 0.2s total,    10 rows/s, 3.36KB/s, 61% active
Per Node: 0.2 parallelism,     1 rows/s,   572B/s
Parallelism: 0.2
Peak Memory: 320B
1.14 [2 rows, 650B] [1 rows/s, 572B/s]

[findinpath]

Fixed by #21146

[vinay1331]

Thanks @findinpath @martint