In the context of reorder join, when a new plan is generated, it may include a project operation. In this case, the newly generated join root and the original join root will no longer be in the same group. To avoid inconsistencies in the statistics between these two groups, we keep the child group's row count unchanged when the parent group expression is a project operation.
legacy planner will set error msg when throw AnalysisException.
However, in some place, we catch these exception and muted them.
So, we should reset back error msg and error code.
The given "length" of broker's pread() method is the buffer length, not the length required from file.
So it may larger than the file length.
So we should return all read data, instead of return EOF when `read()` method return -1
I will add regression test case later when the framework support broker process.
should check the pb's type is set, or the deserialize will core.
should not return unknown type because deserialize will core.
---------
Co-authored-by: yiguolei <yiguolei@gmail.com>
- sh start_fe/start_be --console is used to instruct the program to run in console mode.
- sh start_fe/start_be --daemon is used to instruct the program to run in daemon mode.
- sh start_fe/start_be used starts as a background execution, records output and error logs to the specified file
Fix two bugs:
1. Call `UserGroupInformation.doAs` when enable security authentication
2. `catalogId` is 0 when `IcebergExternalCatalog` is loaded from fe image
Sometimes the first run of query may be longer then former given threshold, which case test fail.
Also add a new session variable test_query_cache_hit
So that we can use it to test if cache is hit in regression test
* [improvement](scanner_schedule) reduce memory consumption of scanner
1. limit scanner by memory consumptin rather than blocks.
2. scheduler run correcty instread of at lest 1.
UNIX_TIMESTAMP function data format parameter supports 'yyyy-MM-dd HH:mm:ss'
The implementation is the same as the date_format function
before:
```sql
mysql> select UNIX_TIMESTAMP('2023-09-18 00:00:00','yyyy-MM-dd HH:mm:ss');
+--------------------------------------------------------------+
| unix_timestamp('2023-09-18 00:00:00', 'yyyy-MM-dd HH:mm:ss') |
+--------------------------------------------------------------+
| NULL |
+--------------------------------------------------------------+
1 row in set (0.04 sec)
```
now:
```sql
mysql> select UNIX_TIMESTAMP('2023-09-18 00:00:00','yyyy-MM-dd HH:mm:ss');
+------------+
| 1694966400 |
+------------+
| 1694966400 |
+------------+
1 row in set (0.01 sec)
```
runtime filter should not push down through topN
runtime filter should not push down through window if target slot is not partition key of all windowExpressions
a wide table
Users rarely set scan_queue_mem_limit, so it almost often works as 2G/20. However,
somecases we need set it to a larger value, especially for insrt into
select from a wide table.
generate more runtime filters
example:
lineitem join partsupp on l_partkey= ps_partkey join filter(part) on ps_partkey=p_partkey
we need two RFs:
RF1: p_partkey->ps_partkey
RF2: p_partkey->l_partkey
This pr will generate RF2, but current version will not.
merge runtime filters
current version, if one src could affect 2 targets, we will generate 2 runtime filters.
after this pr, the two rf will be merged.
refer to regression test: ds_rf2/ds_rf5/ds_rf54
