背景
绑定SQL执行计划,大家一定会想到SQL HINT,通过HINT告诉优化器你要用什么访问方法,用什么JOIN方法,JOIN的顺序,驱动表等等。
但是SQL HINT有一定的弊端,它需要修改应用程序中的SQL语句,把SQL加上HINT。
对pg_hint_plan感兴趣的同学,可以参考我写到文档
https://yq.aliyun.com/articles/57945
但是,有什么方法可以在不修改应用程序,不修改SQL的情况下,锁定SQL的执行计划呢?
锁定执行计划
要锁定执行计划,同样要用到PostgreSQL的 HOOK,在生成plan时动手。
源码
https://github.com/postgrespro/sr_plan
原理如下图
parser的hook,用于获取query text。
planner的hook,干了两间事情,
.1. 在开启了writer模式时,判断表中是否有该sql的执行计划,没有的话就将planner记录在表中。
.2. 从表中读取到对应的执行计划,传给executor,从而达到锁定执行计划的目的。
用法介绍
安装略
在需要锁定执行计划的数据库中创建扩展
CREATE EXTENSION sr_plan;
生成正确的执行计划
连接到对应数据库,并打开一个会话
使用一些开关,使得接下来要记录的执行计划
set enable_bitmapscan=off;
......
打开记录模式
set sr_plan.write_mode = true;
执行需要生成执行计划的QUERY,例如
select a,b,c from tbl where id=1 and c1=2;
检查执行计划是否正确
select * from sr_plans;
如果发现执行计划不正确,删除即可,重新生成。
delett from sr_plans where query_hash=...;
记得关闭记录模式
set sr_plan.write_mode = false;
如果要生成下一条QUERY的执行计划,继续前面的步骤即可。
让优化器使用已生成的执行计划
生成后的执行计划,并不会马上生效,需要更新sr_plans.enable字段
select query_hash, enable, valid, query, explain_jsonb_plan(plan) from sr_plans;
query_hash | enable | valid | query | explain_jsonb_plan
------------+--------+-------+------------------------------------------------------+----------------------------------------------------
1783086253 | f | t | select query_hash from sr_plans where query_hash=10; | Bitmap Heap Scan on sr_plans +
| | | | Recheck Cond: (query_hash = 10) +
| | | | -> Bitmap Index Scan on sr_plans_query_hash_idx+
| | | | Index Cond: (query_hash = 10) +
| | | |
explain_jsonb_plan function allows you to display explain execute the plan of which lies in jsonb.
By default, all the plans are off, you need enable it:
update sr_plans set enable=true where query_hash=1783086253;
(1783086253 for example only) After that, the plan for the query will be taken from the sr_plans.
对于使用参数的QUERY,同样适用。
In addition sr plan allows you to save a parameterized query plan.
In this case, we have some constants in the query are not essential.
For the parameters we use a special function _p (anyelement) example:
select query_hash from sr_plans where query_hash=1000+_p(10);
if we keep the plan for the query and enable it to be used also for the following queries:
select query_hash from sr_plans where query_hash=1000+_p(11);
select query_hash from sr_plans where query_hash=1000+_p(-5);
源码
.1. 获取query text
src/backend/optimizer/plan/planner.c
/*****************************************************************************
*
* Query optimizer entry point
*
* To support loadable plugins that monitor or modify planner behavior,
* we provide a hook variable that lets a plugin get control before and
* after the standard planning process. The plugin would normally call
* standard_planner().
*
* Note to plugin authors: standard_planner() scribbles on its Query input,
* so you'd better copy that data structure if you want to plan more than once.
*
*****************************************************************************/
PlannedStmt *
planner(Query *parse, int cursorOptions, ParamListInfo boundParams)
{
PlannedStmt *result;
if (planner_hook)
result = (*planner_hook) (parse, cursorOptions, boundParams);
else
result = standard_planner(parse, cursorOptions, boundParams);
return result;
}
.2.
src/backend/parser/analyze.c
/*
* parse_analyze
* Analyze a raw parse tree and transform it to Query form.
*
* Optionally, information about $n parameter types can be supplied.
* References to $n indexes not defined by paramTypes[] are disallowed.
*
* The result is a Query node. Optimizable statements require considerable
* transformation, while utility-type statements are simply hung off
* a dummy CMD_UTILITY Query node.
*/
Query *
parse_analyze(Node *parseTree, const char *sourceText,
Oid *paramTypes, int numParams)
{
ParseState *pstate = make_parsestate(NULL);
Query *query;
Assert(sourceText != NULL); /* required as of 8.4 */
pstate->p_sourcetext = sourceText;
if (numParams > 0)
parse_fixed_parameters(pstate, paramTypes, numParams);
query = transformTopLevelStmt(pstate, parseTree);
if (post_parse_analyze_hook)
(*post_parse_analyze_hook) (pstate, query);
free_parsestate(pstate);
return query;
}
/*
* parse_analyze_varparams
*
* This variant is used when it's okay to deduce information about $n
* symbol datatypes from context. The passed-in paramTypes[] array can
* be modified or enlarged (via repalloc).
*/
Query *
parse_analyze_varparams(Node *parseTree, const char *sourceText,
Oid **paramTypes, int *numParams)
{
ParseState *pstate = make_parsestate(NULL);
Query *query;
Assert(sourceText != NULL); /* required as of 8.4 */
pstate->p_sourcetext = sourceText;
parse_variable_parameters(pstate, paramTypes, numParams);
query = transformTopLevelStmt(pstate, parseTree);
/* make sure all is well with parameter types */
check_variable_parameters(pstate, query);
if (post_parse_analyze_hook)
(*post_parse_analyze_hook) (pstate, query);
free_parsestate(pstate);
return query;
}
小结
- 对相应的表执行DDL之后,记得清理在sr_plans中对应表的执行计划。
祝大家玩得开心,欢迎随时来 阿里云促膝长谈 业务需求 ,恭候光临。
阿里云的小伙伴们加油,努力做 最贴地气的云数据库 。