Invent min_parallel_relation_size GUC to replace a hard-wired constant.

The main point of doing this is to allow the cutoff to be set very small,
even zero, to allow parallel-query behavior to be tested on relatively
small tables such as we typically use in the regression tests.  But it
might be of use to users too.  The number-of-workers scaling behavior in
create_plain_partial_paths() is pretty ad-hoc and subject to change, so
we won't expose anything about that, but the notion of not considering
parallel query at all for tables below size X seems reasonably stable.

Amit Kapila, per a suggestion from me

Discussion: <17170.1465830165@sss.pgh.pa.us>
This commit is contained in:
Tom Lane 2016-06-16 13:47:20 -04:00
parent 3b5a2a8856
commit 75be66464c
5 changed files with 37 additions and 5 deletions

View File

@ -3699,6 +3699,20 @@ include_dir 'conf.d'
</listitem>
</varlistentry>
<varlistentry id="guc-min-parallel-relation-size" xreflabel="min_parallel_relation_size">
<term><varname>min_parallel_relation_size</varname> (<type>integer</type>)
<indexterm>
<primary><varname>min_parallel_relation_size</> configuration parameter</primary>
</indexterm>
</term>
<listitem>
<para>
Sets the minimum size of relations to be considered for parallel scan.
The default is 8 megabytes (<literal>8MB</>).
</para>
</listitem>
</varlistentry>
<varlistentry id="guc-effective-cache-size" xreflabel="effective_cache_size">
<term><varname>effective_cache_size</varname> (<type>integer</type>)
<indexterm>

View File

@ -15,6 +15,7 @@
#include "postgres.h"
#include <limits.h>
#include <math.h>
#include "access/sysattr.h"
@ -56,6 +57,7 @@ typedef struct pushdown_safety_info
/* These parameters are set by GUC */
bool enable_geqo = false; /* just in case GUC doesn't set it */
int geqo_threshold;
int min_parallel_relation_size;
/* Hook for plugins to get control in set_rel_pathlist() */
set_rel_pathlist_hook_type set_rel_pathlist_hook = NULL;
@ -690,7 +692,7 @@ create_plain_partial_paths(PlannerInfo *root, RelOptInfo *rel)
parallel_workers = rel->rel_parallel_workers;
else
{
int parallel_threshold = 1000;
int parallel_threshold;
/*
* If this relation is too small to be worth a parallel scan, just
@ -699,21 +701,24 @@ create_plain_partial_paths(PlannerInfo *root, RelOptInfo *rel)
* might not be worthwhile just for this relation, but when combined
* with all of its inheritance siblings it may well pay off.
*/
if (rel->pages < parallel_threshold &&
if (rel->pages < (BlockNumber) min_parallel_relation_size &&
rel->reloptkind == RELOPT_BASEREL)
return;
/*
* Select the number of workers based on the log of the size of the
* relation. This probably needs to be a good deal more
* sophisticated, but we need something here for now.
* sophisticated, but we need something here for now. Note that the
* upper limit of the min_parallel_relation_size GUC is chosen to
* prevent overflow here.
*/
parallel_workers = 1;
while (rel->pages > parallel_threshold * 3)
parallel_threshold = Max(min_parallel_relation_size, 1);
while (rel->pages >= (BlockNumber) (parallel_threshold * 3))
{
parallel_workers++;
parallel_threshold *= 3;
if (parallel_threshold >= PG_INT32_MAX / 3)
if (parallel_threshold > INT_MAX / 3)
break; /* avoid overflow */
}
}

View File

@ -2747,6 +2747,17 @@ static struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
{
{"min_parallel_relation_size", PGC_USERSET, QUERY_TUNING_COST,
gettext_noop("Sets the minimum size of relations to be considered for parallel scan."),
NULL,
GUC_UNIT_BLOCKS,
},
&min_parallel_relation_size,
1024, 0, INT_MAX / 3,
NULL, NULL, NULL
},
{
/* Can't be set in postgresql.conf */
{"server_version_num", PGC_INTERNAL, PRESET_OPTIONS,

View File

@ -304,6 +304,7 @@
#cpu_operator_cost = 0.0025 # same scale as above
#parallel_tuple_cost = 0.1 # same scale as above
#parallel_setup_cost = 1000.0 # same scale as above
#min_parallel_relation_size = 8MB
#effective_cache_size = 4GB
# - Genetic Query Optimizer -

View File

@ -22,6 +22,7 @@
*/
extern bool enable_geqo;
extern int geqo_threshold;
extern int min_parallel_relation_size;
/* Hook for plugins to get control in set_rel_pathlist() */
typedef void (*set_rel_pathlist_hook_type) (PlannerInfo *root,