query reboot pgsql 9.5.1

[Felipe de Jesús Molina Bravo <fjmolinabravo@xxxxxxxxx>]

Hi!!!

I try to explain my problem...sorry for my english :(


In pgsql 9.5.1 I have a two tables with the next structure:

1. Tabla unlogged «public._gc_cat»
   Columna         |     Tipo     | Modificadores 
   -----------------+--------------+---------------
   idppicat        | integer      | 
   idprodxintegrar | integer      | 
   tipo            | character(1) | 
   valor           | numeric      | 
   estado          | character(1) | 
   idsll           | text         | 
   idsfte          | text         | 
   arama           | text[]       | 
   ne_arama        | integer      | 
   rama            | text         | 
   rvar            | text         | 
   nodec           | integer      | 

   Índices:
   "_gc_cat_arama" btree (ne_arama)
   "_gc_cat_arama_gin" gin (arama)

2. Tabla unlogged «public._gc_tb»
   Columna  |  Tipo   | Modificadores 
   ----------+---------+---------------
   idb2     | integer | 
   idc1     | integer | 
   rama     | text    | 
   arama    | text[]  | 
   ne_arama | integer | 
   Índices:
   "_gc_tb_arama" btree (ne_arama)
   "_gc_tb_arama_gin" gin (arama)
   "_gc_tb_idb2idc1" btree (idb2, idc1)


the tabla _gc_cat have 91932 records an _gc_tb have 120130 records; when i run the
next query:

SELECT idprodxintegrar
  FROM _gc_tb a
  LEFT  join
       _gc_cat b
       on ( b.arama <@ a.arama  and a.arama < @ b.arama )

psql send the next message (after three minutes aprox.):
      Terminado (killed)

and i have to reboot my "guest server".

Now i execute the same in pgsql 9.4.5 and all is fine!!! 

The EXPLAINs are:

- pgsql 9.5.1:

  Nested Loop Left Join  (cost=0.03..492944.81 rows=276095 width=4)
    ->  Seq Scan on _gc_tb a  (cost=0.00..3321.30 rows=120130 width=66)
      ->  Bitmap Heap Scan on _gc_cat b  (cost=0.03..4.06 rows=2 width=70)
            Recheck Cond: ((arama <@ a.arama) AND (a.arama <@ arama))
                ->  Bitmap Index Scan on _gc_cat_arama_gin
                (cost=0.00..0.03 rows=2 width=0)
                               Index Cond: ((arama <@ a.arama) AND
                               (a.arama <@ arama))


- pgsql 9.4.5:
  Nested Loop Left Join  (cost=0.03..492944.81 rows=276095 width=4)
    ->  Seq Scan on _gc_tb a  (cost=0.00..3321.30 rows=120130 width=66)
       ->  Bitmap Heap Scan on _gc_cat b  (cost=0.03..4.06 rows=2 width=70)
              Recheck Cond: ((arama <@ a.arama) AND (a.arama <@ arama))
                       ->  Bitmap Index Scan on _gc_cat_arama_gin
                       (cost=0.00..0.03 rows=2 width=0)
                                      Index Cond: ((arama <@ a.arama) AND
                                      (a.arama <@ arama))

If i change the query as:
   SELECT idprodxintegrar
     FROM _gc_tb a
     LEFT  join
          _gc_cat b
          on ( a.ne_arama = b.ne_arama and a.arama <@ b.arama )

In pgsql 9.5.1 finished after  450708.112 ms

In pgsql 9.4.5 finished after   17996.756 ms (very fast!!!)

The EXPLAINs are:
- pgsql 9.5.1
  Nested Loop Left Join  (cost=3.49..1915550.34 rows=41825277 width=4)
    ->  Seq Scan on _gc_tb a  (cost=0.00..3321.30 rows=120130 width=70)
       ->  Bitmap Heap Scan on _gc_cat b  (cost=3.49..14.39 rows=153 width=74)
             Recheck Cond: (a.arama <@ arama)
                Filter: (a.ne_arama = ne_arama)
                         ->  Bitmap Index Scan on _gc_cat_arama_gin
                         (cost=0.00..3.45 rows=460 width=0)
                                        Index Cond: (a.arama <@ arama)

- pgsql 9.4.5
  Nested Loop Left Join  (cost=3.48..1868759.71 rows=42284738 width=4)
    ->  Seq Scan on _gc_tb a  (cost=0.00..3321.30 rows=120130 width=70)
       ->  Bitmap Heap Scan on _gc_cat b  (cost=3.48..14.38 rows=115 width=74)
             Recheck Cond: (a.arama <@ arama)
               Filter: (a.ne_arama = ne_arama)
                        ->  Bitmap Index Scan on _gc_cat_arama_gin
                        (cost=0.00..3.45 rows=460 width=0)
                                       Index Cond: (a.arama <@ arama)


The shared_buffers and work_mem are the same in both versions of pgsql (128MB and
4MB)

I am doing this test in a laptop with the next characteristics:

- hp probook with 8 Gb ram. SATA disk, AMD A8-5550M 
- OS Linux (fedora 23)
- lxc containers


I am sharing the dumper's database are in the next links:

  http://sipre.inegi.org.mx/pgsql/dump_pgsql_9_4_5.dump

  http://sipre.inegi.org.mx/pgsql/dump_pgsql_9_5_1.dump


similar post is found in:

http://www.postgresql.org/message-id/CALrs2KPMowV6juLdOkMRq_P3MA5VkUmhdM4Q1OD0vCf2qimFfA@mail.gmail.
com

thanks in advance!