Merge pull request #2 from ongres/checksfor13plus

adding reviewed queries

sql/AutoVacuum/candidates-for-vacuum-freeze.md

@@ -18,3 +18,6 @@ On `percent_towards_emergency_autovacuum` in 100%, autovacuum will run in aggres
 but also, will force a block read to disk.
 
 The sizes are useful for estimate how long the execution will take, but this is also relative to the concurrency, indexes and other factors.
+
+The default value is too conservative, (10%) probably you can set autovacuum_freeze_max_age for this table in a higer value (~25%).
+https://postgresqlco.nf/doc/en/param/autovacuum_freeze_max_age/13/

sql/BackgroundWriter/10_secs_bgwriter.sql

@@ -1,6 +1,6 @@
 -- 10 second iteration for calculate bgwriter activity
-WITH tara AS (
-  SELECT 
+with tara as (
+  select 
     checkpoints_timed     ,
     checkpoints_req       ,
     checkpoint_write_time ,
@@ -13,7 +13,7 @@ WITH tara AS (
     buffers_alloc         
     from pg_stat_bgwriter, pg_sleep(10)
 )
-SELECT 
+select 
     pgb.checkpoints_timed     - tara.checkpoints_timed,
     pgb.checkpoints_req       - tara.checkpoints_req,
     pgb.checkpoint_write_time - tara.checkpoint_write_time,
@@ -24,5 +24,5 @@ SELECT
     pgb.buffers_backend       - tara.buffers_backend,
     pgb.buffers_backend_fsync - tara.buffers_backend_fsync,
     pgb.buffers_alloc      - tara.buffers_alloc
-    FROM pg_stat_bgwriter pgb,tara
+    from pg_stat_bgwriter pgb,tara
 ; 

sql/Cache/cache_effectivity_by_table.sql

@@ -1,2 +1,11 @@
 -- Shows the cache effectivity on each table (1 row per table)
 -- "_hits" and "_reads" includes "regular" blocks, indexes blocks, toast and tidx blocks
+
+select
+    schemaname, relname, 
+	heap_blks_read as heap_read,
+	heap_blks_hit  as heap_hit,
+	heap_blks_hit / (heap_blks_hit + heap_blks_read::float) as ratio
+from
+	pg_statio_user_tables
+where (heap_blks_hit + heap_blks_read) > 0;

sql/DatabaseCapacity/size_growth_interval.sql

@@ -4,11 +4,11 @@
 WITH timeset AS (
 select pg_database_size(datname) num, 
     pg_size_pretty(pg_database_size(datname)) size 
-    from pg_database where datname = 'db'
+    from pg_database where datname = current_database()
 UNION ALL
 select pg_database_size(datname) num, 
     pg_size_pretty(pg_database_size(datname)) size 
-    from pg_database, pg_sleep(10) where datname = 'db' 
+    from pg_database, pg_sleep(10) where datname = current_database()
 )
 SELECT pg_size_pretty((num - lag(num,1) 
             OVER (ORDER BY num))/10) transfer_per_second, 

sql/Indexes/unused_indexes.sql

@@ -1,13 +1,16 @@
-SELECT s.schemaname,
-       s.relname AS tablename,
-       s.indexrelname AS indexname,
-       pg_relation_size(s.indexrelid) AS index_size
-FROM pg_catalog.pg_stat_user_indexes s
-   JOIN pg_catalog.pg_index i ON s.indexrelid = i.indexrelid
-WHERE s.idx_scan = 0      -- has never been scanned
-  AND 0 <>ALL (i.indkey)  -- no index column is an expression
-  AND NOT i.indisunique   -- is not a UNIQUE index
-  AND NOT EXISTS          -- does not enforce a constraint
-         (SELECT 1 FROM pg_catalog.pg_constraint c
-          WHERE c.conindid = s.indexrelid)
-ORDER BY pg_relation_size(s.indexrelid) DESC;
+select s.schemaname,
+       s.relname as tablename,
+       s.indexrelname as indexname,
+       pg_relation_size(s.indexrelid) as index_size,
+       pg_size_pretty(pg_relation_size(s.indexrelid)) as index_size_human
+from pg_catalog.pg_stat_user_indexes s
+   join pg_catalog.pg_index i on s.indexrelid = i.indexrelid
+where --s.idx_scan = 0    -- has never been scanned
+      s.idx_scan < 50     -- almost never used
+  and pg_relation_size(relid) > 5 * 8192 -- skip empty objects
+  and 0 <>all (i.indkey)  -- no index column is an expression
+  and not i.indisunique   -- is not a unique index
+  and not exists          -- does not enforce a constraint
+         (select 1 from pg_catalog.pg_constraint c
+          where c.conindid = s.indexrelid)
+order by pg_relation_size(s.indexrelid) desc;

sql/Indexes/worst_index_hits.sql

@@ -16,5 +16,8 @@ WITH worst_index_hit_ratio AS (
   GROUP BY indexrelname, relid
   ORDER BY hit_ratio ASC
 )
-SELECT * FROM worst_index_hit_ratio WHERE idx_read = 0 and idx_hit = 0; --  < 0.9;
+SELECT * FROM worst_index_hit_ratio 
+  WHERE 
+    -- idx_read = 0 and idx_hit = 0; --  < 0.9;
+    hit_ratio < 0.8;
 

sql/MultiXact/get_oldest_by_active_conn.sql

@@ -0,0 +1,6 @@
+-- This query returns a list of active session holding multixact slots and their ages
+-- The age here is not tied to a time constraint, but rather to the transaction
+-- amount distance from the relminmxid and now.
+SELECT pid, datname, usename, state, txid_current(), backend_xmin, 
+txid_current()::text::int - backend_xmin::text::int difference
+, statement_timestamp() - query_start elapsed , pg_blocking_pids(pid) as blocked_by,  query::varchar(150) query FROM pg_stat_activity WHERE backend_xmin IS NOT NULL ORDER BY age(backend_xmin) ; 

sql/MultiXact/get_oldest_multixact.sql

@@ -5,4 +5,4 @@ SELECT relnamespace::regnamespace, relname, relkind, relminmxid,
       (select next_multixact_id::text::bigint FROM pg_control_checkpoint()) - relminmxid::TEXT::BIGINT AS age
 FROM pg_class
 WHERE relminmxid::TEXT::BIGINT <> 0
-ORDER BY relminmxid::TEXT::BIGINT ASC;    
+ORDER BY relminmxid::TEXT::BIGINT ASC;    

sql/PgStatStatements/large_result_sets.sql

@@ -3,7 +3,7 @@ SELECT
     substring(query,0,100),
     calls,
     rows/calls as rowspercall,
-    mean_time,
+    mean_exec_time,
     shared_blks_hit,
     shared_blks_read
 FROM pg_stat_statements

sql/PgStatStatements/most_frequent_queries.sql

@@ -2,8 +2,8 @@
 SELECT queryid,
     query,
     calls,
-    total_time/1000 as time_secs,
-    (total_time/1000)/calls as per_call,
+    total_exec_time/1000 as time_secs,
+    (total_exec_time/1000)/calls as per_call,
     CASE
       WHEN (shared_blks_hit > 0 AND shared_blks_read > 0) THEN round(shared_blks_hit / (shared_blks_hit + shared_blks_read)) 
       ELSE 0 END as hit_ratio,

sql/PgStatStatements/slowest_queries.sql

@@ -2,8 +2,8 @@
 SELECT queryid,
     query,
     calls,
-    total_time/1000 as time_secs,
-    (total_time/1000)/calls as per_call,
+    total_exec_time/1000 as time_secs,
+    (total_exec_time/1000)/calls as per_call,
     CASE
       WHEN (shared_blks_hit > 0 AND shared_blks_read > 0) THEN round(shared_blks_hit / (shared_blks_hit + shared_blks_read)) 
       ELSE 0 END as hit_ratio,
@@ -14,5 +14,5 @@ SELECT queryid,
       WHEN (shared_blks_written > 0) THEN round(shared_blks_written / calls )
       ELSE 0 END as blk_wrtn_per_call
 FROM pg_stat_statements
-ORDER BY total_time/calls DESC
+ORDER BY total_exec_time/calls DESC
 LIMIT 10;

sql/PgStatStatements/unstable_queries.sql

@@ -1,14 +1,14 @@
 -- Unstable queries, with larges deltas.
 
-SELECT queryid, calls,mean_time, max_time, query, 
+SELECT queryid, calls,mean_exec_time, max_exec_time, query, 
   rows, shared_blks_read, 
 
-  blk_read_time+blk_write_time,temp_blks_written+temp_blks_read,stddev_time 
+  blk_read_time+blk_write_time,temp_blks_written+temp_blks_read,stddev_exec_time 
 from pg_Stat_statements 
 
 ORDER BY 
-  mean_time DESC,
-  stddev_time DESC 
+  mean_exec_time DESC,
+  stddev_exec_time DESC 
 
 LIMIT 10;
 

sql/Settings/generic.sql

@@ -1,4 +1,5 @@
 -- Get a better output of pg_settings
+/*
 SELECT name, category,
         setting,
         short_desc::text , extra_desc::text, 
@@ -9,3 +10,6 @@ WHERE
     category ~ '^Replication'
     -- category ~ 'eplication'
 ORDER BY context;
+*/
+SELECT *
+FROM pg_settings;