[英]pg_stat_activity doesn't update within a procedure or transaction
I query pg_stat_activity
but the contents seem to stay the same after the first time I query it within a plpgsql function body, SERIALIZABLE
or REPEATABLE READ
transaction. 我查询
pg_stat_activity
但是在我第一次在plpgsql函数体, SERIALIZABLE
或REPEATABLE READ
事务中查询后,内容似乎保持不变。
Why does pg_stat_activity
not follow the rules for READ COMMITTED
in a plpgsql procedure? 为什么
pg_stat_activity
不遵循plpgsql过程中的READ COMMITTED
规则?
How do I get the current status? 我如何获得当前状态? I want to loop over a query against
pg_stat_activity
in plpgsql until another query running on another backend finishes. 我想在plpgsql中循环查询
pg_stat_activity
的查询,直到在另一个后端运行的另一个查询完成。
PostgreSQL makes a per-backend (per-connection, effectively) cache of the data used by the pg_stat_get_activity()
function used by both pg_stat_activity
and pg_stat_replication
. PostgreSQL对
pg_stat_activity
和pg_stat_replication
使用的pg_stat_get_activity()
函数使用的数据进行每个后端(每个连接,有效)缓存。
This cache is cleared on commit/rollback, but not at the end of each statement within a transaction in READ COMMITTED
like usual. 此缓存在提交/回滚时清除,但不会像往常一样在
READ COMMITTED
的事务中的每个语句的末尾清除。
You can explicitly clear it with SELECT pg_stat_clear_snapshot()
. 您可以使用
SELECT pg_stat_clear_snapshot()
显式清除它。 Call it within the body of a PL/PgSQL LOOP
to refresh. 在PL / PgSQL
LOOP
的主体内调用它来刷新。
There is AFAIK no way to ask PostgreSQL to auto-refresh after each statement when using repeatable read
or higher isolation. 当使用
repeatable read
或更高隔离时,AFAIK无法在每个语句后要求PostgreSQL自动刷新。
In the source code, see pgstat_read_current_status(void)
and pgstat_clear_snapshot(void)
. 在源代码中,请参阅
pgstat_read_current_status(void)
和pgstat_clear_snapshot(void)
。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.