feature: asynchronously update file statistics during vacuum

contrib/pax_storage/expected/cluster.out

@@ -1,4 +1,5 @@
 set pax.max_tuples_per_file to 131072;
+set pax.enable_sync_collect_stats = on;
 -- cluster table using index
 -- start_ignore
 drop table if EXISTS t_index_cluster;
@@ -295,3 +296,4 @@ select ptblockname,ptstatistics,ptisclustered from get_pax_aux_table('t_lexical_
 (6 rows)
 
 drop table t_lexical_cluster;
+reset pax.enable_sync_collect_stats;

contrib/pax_storage/expected/disable_autovacuum.out

@@ -0,0 +1,10 @@
+alter system set autovacuum = off;
+select gp_segment_id, pg_reload_conf() from gp_id union select gp_segment_id, pg_reload_conf() from gp_dist_random('gp_id');
+ gp_segment_id | pg_reload_conf 
+---------------+----------------
+             2 | t
+             1 | t
+             0 | t
+            -1 | t
+(4 rows)
+

contrib/pax_storage/expected/enable_autovacuum.out

@@ -0,0 +1,10 @@
+alter system set autovacuum = on;
+select gp_segment_id, pg_reload_conf() from gp_id union select gp_segment_id, pg_reload_conf() from gp_dist_random('gp_id');
+ gp_segment_id | pg_reload_conf 
+---------------+----------------
+             2 | t
+             1 | t
+             0 | t
+            -1 | t
+(4 rows)
+

contrib/pax_storage/expected/filter.out

@@ -1,5 +1,6 @@
 set pax.enable_debug to on;
 set pax.enable_sparse_filter = on;
+set pax.enable_sync_collect_stats = on;
 create table pax_test.null_test_t(a int, b int, c text) using pax;
 insert into pax_test.null_test_t(a) select null from generate_series(1,2)i;
 insert into pax_test.null_test_t select 1, i, 'cc_' || i from generate_series(1,2)i;
@@ -223,3 +224,4 @@ kind group, filter rate: 0 / 1
 reset client_min_messages;
 drop table pax_test.in_test_t;
 reset pax.enable_sparse_filter;
+reset pax.enable_sync_collect_stats;

contrib/pax_storage/expected/filter_1.out

@@ -1,5 +1,6 @@
 set pax.enable_debug to on;
 set pax.enable_sparse_filter = on;
+set pax.enable_sync_collect_stats = on;
 create table pax_test.null_test_t(a int, b int, c text) using pax;
 insert into pax_test.null_test_t(a) select null from generate_series(1,2)i;
 insert into pax_test.null_test_t select 1, i, 'cc_' || i from generate_series(1,2)i;
@@ -223,3 +224,4 @@ kind group, filter rate: 0 / 1
 reset client_min_messages;
 drop table pax_test.in_test_t;
 reset pax.enable_sparse_filter;
+reset pax.enable_sync_collect_stats;

contrib/pax_storage/expected/filter_tree.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create  or replace function intrc(iint int)
  returns int as $$ 
  begin return iint; end;
@@ -735,6 +736,7 @@ LOG:  statement: select count(*) from t1 where coalesce(v1, 2) != 1;
 
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
+reset pax.enable_sync_collect_stats;
 drop table t1;
 drop table t2;
 drop table t_allnull;

contrib/pax_storage/expected/filter_tree_1.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create  or replace function intrc(iint int)
  returns int as $$ 
  begin return iint; end;
@@ -742,6 +743,7 @@ LOG:  statement: select count(*) from t1 where coalesce(v1, 2) != 1;
 
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
+reset pax.enable_sync_collect_stats;
 drop table t1;
 drop table t2;
 drop table t_allnull;

contrib/pax_storage/expected/filter_tree_arithmetic.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create table t_arithmetic(same int, v1 int, v2 int, v3 int) using pax with (minmax_columns='v1,v2,v3');
 NOTICE:  Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'same' as the Apache Cloudberry data distribution key for this table.
 HINT:  The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
@@ -1063,5 +1064,6 @@ LOG:  kind file, filter rate: 0 / 1
 
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
+reset pax.enable_sync_collect_stats;
 drop table t_arithmetic;
 drop table ta_mul;

contrib/pax_storage/expected/filter_tree_arithmetic_1.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create table t_arithmetic(same int, v1 int, v2 int, v3 int) using pax with (minmax_columns='v1,v2,v3');
 NOTICE:  Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'same' as the Apache Cloudberry data distribution key for this table.
 HINT:  The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
@@ -938,5 +939,6 @@ LOG:  statement: select count(*) from ta_mul where v2 * v2 >= 10000;
 
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
+reset pax.enable_sync_collect_stats;
 drop table t_arithmetic;
 drop table ta_mul;

contrib/pax_storage/expected/filter_tree_optimizer.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create  or replace function intrc(iint int)
  returns int as $$ 
  begin return iint; end;
@@ -738,6 +739,7 @@ LOG:  statement: select count(*) from t1 where coalesce(v1, 2) != 1;
 
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
+reset pax.enable_sync_collect_stats;
 drop table t1;
 drop table t2;
 drop table t_allnull;

contrib/pax_storage/expected/filter_tree_optimizer_1.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create  or replace function intrc(iint int)
  returns int as $$ 
  begin return iint; end;
@@ -746,6 +747,7 @@ LOG:  statement: select count(*) from t1 where coalesce(v1, 2) != 1;
 
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
+reset pax.enable_sync_collect_stats;
 drop table t1;
 drop table t2;
 drop table t_allnull;

contrib/pax_storage/expected/filter_tree_root_quals.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create table t1(same int, v1 int, v2 int, v3 int, v4 int) using pax with (minmax_columns='v1,v2,v3,v4');
 NOTICE:  Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'same' as the Apache Cloudberry data distribution key for this table.
 HINT:  The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
@@ -55,5 +56,6 @@ LOG:  kind file, filter rate: 1 / 2
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
 reset optimizer;
+reset pax.enable_sync_collect_stats;
 drop table t1;
 drop table t2;

contrib/pax_storage/expected/filter_tree_root_quals_1.out

@@ -7,6 +7,7 @@
 set default_table_access_method to pax;
 set pax.enable_debug to on;
 set pax.enable_sparse_filter to on;
+set pax.enable_sync_collect_stats = on;
 create table t1(same int, v1 int, v2 int, v3 int, v4 int) using pax with (minmax_columns='v1,v2,v3,v4');
 NOTICE:  Table doesn't have 'DISTRIBUTED BY' clause -- Using column named 'same' as the Apache Cloudberry data distribution key for this table.
 HINT:  The 'DISTRIBUTED BY' clause determines the distribution of data. Make sure column(s) chosen are the optimal data distribution key to minimize skew.
@@ -55,5 +56,6 @@ LOG:  kind file, filter rate: 2 / 2
 reset client_min_messages;
 LOG:  statement: reset client_min_messages;
 reset optimizer;
+reset pax.enable_sync_collect_stats;
 drop table t1;
 drop table t2;