Please post in plain text if possible.
2010/8/5 노현석
>
> hi..
>
> we test locking PostgreSQL 8.4.4 on x86_64.
> $
> $ kill -9 5143
What you want is plain old
kill 5143
or
kill -SIGTERM 5143
Kill -SIGTERM is like using a fly swatter to kill a fly. SIGKILL (-9)
is like lobbing a grena
Hi,
On Thu, 05 Aug 2010 17:52:44 +0900 노현석 wrote:
> we test locking PostgreSQL 8.4.4 on x86_64.
>
[...]
> begin;
> delete from citytest;
> <--- no commit/rollback
> <--- this will be Blocker
>
> 2> session 2,3,4
> ###
> $ psql mydb
> delete from citytest;
2010/8/5 노현석
> Could you teach me,Is this expected behavior ?? (disapper blocking
> process not only Blocker process)
> and
> Could you teach me, how can we eliminate just Blocker session...
>
> instead of going to the OS to kill the process, just type ctrl-c into the
psql terminal ses
=?EUC-KR?B?s+vH9ryu?= writes:
> $ kill -9 5143
[ and all the other backends disappear too ]
> Could you teach me, Is this expected behavior ?
Yes. That is not the recommended way to kill a backend.
Try pg_terminate_backend().
regards, tom lane
--
Sent via pgsql-gener
Le 05/08/2010 10:52, 노현석 a écrit :
> [...]
> we test locking PostgreSQL 8.4.4 on x86_64.
>
> 1> session 1
> ###
> $ psql mydb
> drop table citytest;
> CREATE TABLE citytest (
> i_id integer ,
> city varchar(80)
> ) ;
> ALTER TABLE citytest ADD CONSTRAINT citytest_pkey P
hi..
we test locking PostgreSQL 8.4.4 on x86_64.
1> session 1
###
$ psql mydb
drop table citytest;
CREATE TABLE citytest (
i_id integer ,
city varchar(80)
) ;
ALTER TABLE citytest ADD CONSTRAINT citytest_pkey PRIMARY KEY (i_id);
insert into citytest values (1,'aaa');
i
