Function

get_meta_version(Oid)

Description: Obtains the metadata version information cached in a session on a DN. The input parameter is the OID of the primary table. The output parameter is the version information of all auxiliary tables related to the primary table, including the index, partition, and primary table.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5
6
7

SELECT * FROM get_meta_version(16972);
 obj_oid | obj_type | obj_parent_oid | meta_version 
---------+----------+----------------+--------------
   16972 |        r |          16972 |       267910
   16952 |        p |          16972 |       267910
   16958 |        p |          16972 |       267910
(3 row)

get_meta_version()

Description: Obtains all metadata version information cached in a session on a DN. The output is the version information of all primary tables and related auxiliary tables, including indexes, partitions, and primary tables.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5
6
7

SELECT * FROM get_meta_version();
 obj_oid | obj_type | obj_parent_oid | meta_version 
---------+----------+----------------+--------------
   16972 |        r |          16972 |       267910
   16952 |        p |          16972 |       267910
   16958 |        p |          16972 |       267910
(3 row)

pgxc_get_meta_version(schemaname, relname)

Description: Obtains the version information of specified metadata cached in sessions on all DNs. The input parameters are the schema name and table name of the primary table. The output parameters are the version information of all auxiliary tables related to the primary table, including the index, partition, and primary table.

Return type: record

The following table describes return columns.

pgxc_get_meta_version()

Description: Obtains all metadata version information cached in sessions on all DNs. The output is the version information of all auxiliary tables related to the primary table, including the index, partition, and primary table itself.

Return type: record

The following table describes return columns.

clean_dn_metadata(int)

Description: Clears all metadata cached in DN sessions. If this parameter is set to 1, the metadata cached in all sessions is cleared. If this parameter is set to other values, the metadata cached in the current session is cleared.

Return type: int

The following table describes return columns.

pgxc_clean_dn_metadata(int)

Description: Clears all metadata cached in all DN sessions. If this parameter is set to 1, the metadata cached in all sessions is cleared. If this parameter is set to other values, the metadata cached in the current session is cleared.

Return type: record

The following table describes return columns.

get_global_meta_cache(int)

Description: Obtains the global cache metadata on a DN. The input parameter is the bucket number, which ranges from 0 to 511.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5

EXECUTE DIRECT ON (datanode5) 'SELECT * FROM get_global_meta_cache(1)';
 bucket_idx | meta_seq | db_oid | meta_oid | meta_part_num | meta_idx_num |  meta_version 
------------+----------+--------+----------+---------------+--------------+----------------
          1 |        1 |  16852 |    18552 |             0 |            0 | 18552 2394152 , 
(1 row)

get_global_meta_cache()

Description: Obtains the global cache metadata on a DN.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5

EXECUTE DIRECT ON (datanode5) 'SELECT * FROM get_global_meta_cache()';
 bucket_idx | meta_seq | db_oid | meta_oid | meta_part_num | meta_idx_num |  meta_version 
------------+----------+--------+----------+---------------+--------------+----------------
          1 |        1 |  16852 |    18552 |             0 |            0 | 18552 2394152 , 
(1 row)

pgxc_get_global_meta_cache()

Description: Obtains the global cache metadata on all DNs.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5
6
7

SELECT * FROM pgxc_get_global_meta_cache();
 node_name | bucket_idx | meta_seq | db_oid | meta_oid | meta_part_num | meta_idx_num |  meta_version 
-----------+------------+----------+--------+----------+---------------+--------------+----------------
 datanode1 |          1 |        1 |  16852 |    18552 |             0 |            0 | 18552 2394152 , 
 datanode2 |          1 |        1 |  16852 |    18552 |             0 |            0 | 18552 2394152 , 
 datanode3 |          1 |        1 |  16852 |    18552 |             0 |            0 | 18552 2394152 , 
(3 row)

global_meta_cache_reset()

Description: Clears global cached metadata on DNs.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5

ELECT * FROM global_meta_cache_reset();
 meta_num 
----------
        1
(1 row)

pgxc_global_meta_cache_reset()

Description: Clears the global cache metadata on all DNs.

Return type: record

The following table describes return columns.

Example:

1
2
3
4
5
6
7

SELECT * FROM pgxc_global_meta_cache_reset();
 node_name | meta_num  
-----------+----------
 datanode1 |        1
 datanode2 |        1
 datanode3 |        1
(3 row)

pg_obs_file_size(scheme_name.tablename)

Description: Obtains the CU file name and size of a table or partition on OBS. This function is valid only for tables whose colversion is 3.

Return type: record

The function parameter fields are as follows:

Example:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

-- The input parameter is tablename.
elect pg_obs_file_size('t2_col_part_obs');
      pg_obs_file_size
----------------------------
 (C1_16777266462721.0,1024)
 (C1_16777266429953.0,1024)
 (C1_16777249734657.0,1024)
 (C1_16777249701889.0,1024)
(4 rows)
-- The input parameter is schema.tablename.
select pg_obs_file_size('public.t2_col_part_obs');
      pg_obs_file_size
----------------------------
 (C1_16777266462721.0,1024)
 (C1_16777266429953.0,1024)
 (C1_16777249734657.0,1024)
 (C1_16777249701889.0,1024)
(4 rows)
-- The input parameter is oid.
select pg_obs_file_size(16593);
      pg_obs_file_size
----------------------------
 (C1_16777266462721.0,1024)
 (C1_16777266429953.0,1024)
 (C1_16777249734657.0,1024)
 (C1_16777249701889.0,1024)
(4 rows)

pg_obs_file_size(scheme_name.tablename,partition_name)

Description: Obtains the column-store CU file name and file size of a partitioned table on OBS. This function is valid only for column-store tables whose colversion is 3.

Return type: record

The function parameter fields are as follows:

Example:

1
2
3
4
5
6
7
8

elect pg_obs_file_size('public.t2_col_part_obs','p1');
      pg_obs_file_size
----------------------------
 (C1_16777266462721.0,1024)
 (C1_16777266429953.0,1024)
 (C1_16777249734657.0,1024)
 (C1_16777249701889.0,1024)
(4 rows)

pg_scan_residualfiles()

Description: Scans all residual file records in the database where the current node resides. When it is executed on a CN, it scans the database of the CN and OBS for residual files. When it is executed on a CN, it scans the database of the DN for residual files. This function is a database-level function and applies only to the current database. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

pgxc_scan_residualfiles()

Description: Scans all nodes for the residual files of the current database. This function is a cluster-level function and can be executed only on a CN. It is related to the database where the CN is located. This function cannot be executed on the standby node.

Parameter type: none

Return type: record

The following table describes return columns.

pg_get_scan_residualfiles()

Description: Obtains all residual file records of the current node. This function is an instance-level function and is irrelevant to the current database. It can run on any instance. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

1
2
3
4
5
6
7

 select * from pg_get_scan_residualfiles();
 handled |  dbname  |                                                         residualfile                                                         | size | inode | atime | mtime | ctime |                filepath                |
notes
---------+----------+------------------------------------------------------------------------------------------------------------------------------+------+-------+-------+-------+-------+----------------------------------------+-
------
 f       | postgres | /test/obsview/cudesc_check/user1/obs.xxx.com/cu_obs_tbs/tablespace_secondary/15854/19865 |    0 |     0 |       |       |       | pgscrf_meta_15854_20231106095438240991 |
(1 row)

pgxc_get_scan_residualfiles()

Description: Obtains residual file records on all nodes. This function is a cluster-level function and can be executed only on CNs. It is irrelevant to the current database. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

1
2
3
4
5
6
7

 select * from pgxc_get_scan_residualfiles();
  node_name   | instance_id  | handled |  dbname  |                                                         residualfile                                                         | size |  inode  |         atime          |
mtime          |         ctime          |                filepath                | notes
--------------+--------------+---------+----------+------------------------------------------------------------------------------------------------------------------------------+------+---------+------------------------+------------------------+------------------------+----------------------------------------+-------
 datanode1    | datanode1    | f       | postgres | base/15854/19863                                                                                                             |    0 | 2939427 | 2023-11-06 09:54:15+08 | 2023-11-06 09:54:15+08 | 2023-11-06 09:54:15+08 | pgscrf_meta_15854_20231106095437555205 |
 coordinator1 | coordinator1 | f       | postgres | /test/obsview/cudesc_check/user1/obs.xxx.com/cu_obs_tbs/tablespace_secondary/15854/19865 |    0 |       0 |                            |             |                        | pgscrf_meta_15854_20231106095438240991 |
(2 rows)

pg_archive_scan_residualfiles()

Description: Archives all residual file records of the current node. This function is an instance-level function and is irrelevant to the current database. It can run on any instance. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

1
2
3
4
5

select * from pg_archive_scan_residualfiles();
                             archive                               |  count | size 
-------------------------------------------------------------------+--------+------
 pg_residualfiles/archive/pgscrf_archive_15842_20230912182934335330|      1 |    0
(1 row)

pgxc_archive_scan_residualfiles()

Description: Archives residual file records on all nodes. This function is a cluster-level function and can be executed only on CNs. It is irrelevant to the current database. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

1
2
3
4
5
6

select * from pgxc_archive_scan_residualfiles();
  node_name   | instance_id  |                           archive                            | count | size
--------------+--------------+--------------------------------------------------------------+-------+------
 datanode1    | datanode1    | pg_residualfiles/archive/pgscrf_archive_20231106103246489550 |     1 |    0
 coordinator1 | coordinator1 | pg_residualfiles/archive/pgscrf_archive_20231106103246592449 |     1 |    0
(2 rows)

pg_rm_scan_residualfiles_archive()

Description: Deletes files from the archived file list on the current instance. This function is an instance-level function and is irrelevant to the current database. It can run on any instance. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

1
2
3
4
5

select * from pg_rm_scan_residualfiles_archive();
 count | size
-------+------
     1 |    0 
(1 row)

pgxc_rm_scan_residualfiles_archive()

Description: Deletes files in the archive directory from all nodes. This function is a cluster-level function and can be executed only on CNs. It is irrelevant to the current database. This function cannot be executed on the standby node.

Return type: record

The following table describes return columns.

1
2
3
4
5
6

select * from pgxc_rm_scan_residualfiles_archive();
  node_name   | instance_id  | count | size
--------------+--------------+-------+------
 datanode1    | datanode1    |     1 |    0
 coordinator1 | coordinator1 |     1 |    0
(2 rows)

analyze_table(scheme_name, rel_name, sample_rate, random_rate default null, prarallel_degree default null)

Description: Samples data to a temporary table in parallel, performs Full Analyze on the temporary table, and updates statistics.

Return type: record

The function parameter columns are as follows:

Example:

1

call analyze_table('public','t1_col_obs',10,0,20);

pgxc_clear_disk_cache()

Description: Deletes all disk cache files.

Return type: void

Example:

postgres=# select pgxc_clear_disk_cache();
 pgxc_clear_disk_cache
-----------------------

(1 row)