Re: How to get the statment sql that caused the lock

Hi Jonathan,

I understand that , if we have and index in the column id, the first and the third line will lock the line, according the bind variable.

it is running in session S1

update t1 set n1 = 99 where id = :b1;
update t2 set a1 = 66
update t1 set n4 = 33 where id1 = :b2

Imagine another session (S2) using the same ID for update , it will be stopped until the first session apply the commit. This is a basic concept of the lock.

The issue is that when we have millions of updates, inserts, and deletes in the production environment, the sql that triggered the transaction may no longer be in the SGA, even though it triggered a transaction.

Regards
Eriovaldo

Em seg, 29 de out de 2018 às 08:45, Jonathan Lewis < jonathan_at_jlcomp.demon.co.uk> escreveu:

>
> Eriovaldo,
>
> I am surprised that anyone even bothered to create that enhancement
> request, but someone show have marked it is "ignore" by now.
>
> Consider the following transaction:
>
> update t1 set n1 = 99 where id = :b1;
> update t2 set a1 = 66
> update t1 set n4 = 33 where id1 = :b2
>
> Your ER asked for the sql_id of the first statement to be stored in
> v$transaction.
> How are you, while trouble-shooting, going to behave it it's actually the
> third statement that is blocking you ?
>
>
> Regards
> Jonathan Lewis
>
>
>
> ________________________________________
> From: oracle-l-bounce_at_freelists.org <oracle-l-bounce_at_freelists.org> on
> behalf of Eriovaldo Andrietta <ecandrietta_at_gmail.com>
> Sent: 29 October 2018 11:39:03
> To: neil_chandler_at_hotmail.com
> Cc: Mladen Gogala; ORACLE-L
> Subject: Re: How to get the statment sql that caused the lock
>
> Hi Gogala,
>
> Thanks for your comments.
> When I wrote "query" I want mean : sql statements executing : INSERT,
> UPDATE, DELETE, TRUNCATE and LOCK (select for update).
>
> I agree, we can get the objetct locked, and for it I used this query:
> select
> sb.sid sid_blocker
> ,sb.inst_id inst_id_blocker
> ,sb.event
> ,substr(sb.program,1,25) program
> ,o.OBJECT_NAME
> ,CASE WHEN
> o.object_type = 'TABLE' THEN
> dbms_rowid.rowid_create ( 1, o.DATA_OBJECT_ID,
> sb.ROW_WAIT_FILE#, sb.ROW_WAIT_BLOCK#, s.ROW_WAIT_ROW# )
> ELSE
> NULL
> END rowid_bloocked
> from gv$session s
> ,gv$session sb
> ,dba_objects o
> where s.seconds_in_wait > 60
> and s.BLOCKING_SESSION = sb.sid
> and s.BLOCKING_INSTANCE = sb.INST_ID
> and sb.ROW_WAIT_OBJ# = o.OBJECT_ID;
>
> My purpose is to retrieve the sql_id related to the transaction line in
> the view v$transaction that is responsible for the execution of the :
> INSERT, UPDATE, DELETE ... in order to show it to the developer and
> validate the application if need to add commit in the code, if does not
> exists.
>
> I received a message from a member of the group , like this :
>
> I asked Oracle (SR 3-12200129251 : Request for ID of SQL responsible for
> transaction to be added to v$transaction) and they created an enhancement
> request
> Bug 24920354 : ADD SQL_ID COLUMN TO V$TRANSACTION OF THE SQL THAT STARTED
> THE TRANSACTION
>
> But two years have passed. They did nothing about it.
>
> This is what I am locking for.
>
> Thanks and Regards
>
> Eriovaldo
>
>
> Em seg, 29 de out de 2018 às 05:51, Neil Chandler <
> neil_chandler_at_hotmail.com<mailto:neil_chandler_at_hotmail.com>> escreveu:
> Queries can lock tables:
>
> SELECT... FOR UPDATE;
>
> I find this is a common cause of blocking and is easy to overlook as you
> might miss the FOR UPDATE.
>
> Table DDL can also lock the resource by dropping a TM lock on there.
>
> Neil.
> sent from my phone
>
> On 29 Oct 2018, at 03:36, Mladen Gogala <gogala.mladen_at_gmail.com<mailto:
> gogala.mladen_at_gmail.com>> wrote:
>
>
> Hi Eriovaldo!
>
> Queries do not lock tables, unless the queries are distributed, as in
> "query over a database link". For everything else, queries to not lock
> tables.
>
> Second, the fact that the sessions are idle is irrelevant. If I execute
> "LOCK TABLE SCOTT.EMP IN EXCLUSIVE MODE" and go to lunch, my session will
> be idle. It will also keep the lock until the transaction ends. Locks are
> caused by 5 statements: INSERT, UPDATE, DELETE, TRUNCATE and LOCK. For the
> purpose of this consideration, we will consider MERGE to be an update &
> insert combination.
>
> So, if you have a SID, you should be able to find out what objects does
> the session have locked and in what mode. Sometimes, that can be a load of
> fun. While I was a DBA, I worked with a developer who learned that /*+
> APPEND */ hint for inserts "makes things go faster". So she has put the /*+
> APPEND */ hint to each and every insert statement in her multi-user web
> application. For some inexplicable reason, things did not go faster, but I
> did have some fun.
>
> * Idle sessions can hold locks, for a very long time. That was the
> primary reason for inventing resource limits.
> * Queries, except in very rare cases, DO NOT LOCK anything.
> * V$LOCK will tell what objects are locked and in what mode. Unless
> you are using 18.3 or newer, you cannot kill a statement (actually, not
> true: you can kill the current statement by using kill -URG, but very few
> people know that and even fewer do that).
> * Once you figure out that an idle session is blocking you by holding
> a lock, you kill it without further ado and ask the application developer
> to fix the bug. There is absolutely no reason whatsoever for an idle
> session to hold locks. Locks are means for preserving consistency. You want
> to prevent someone else from modifying a critical resource until your
> transaction is finished. You don't hold locks on resources until an upgrade
> to a new Oracle version takes you apart. TRANSACTION != MARRIAGE.
> * Just for completeness, the /*+ APPEND */ hint causes the insert to
> allocate a batch of empty blocks below the high water mark and effectively
> extend the table. That requires an exclusive lock on the table. And
> exclusive locks can be bad for concurrency.
>
> On 10/27/18 6:49 PM, Eriovaldo Andrietta wrote:
> Hello,
>
> I using Oracle 12c R2.
> I have this query that shows blocked sid:
>
> Query 1 :
> select * from gv$lock where sid in
> (select sid from gv$session where blocking_session is not null);
>
> and here , the query shows the blocker sid. These sessions are IDLE.
> Query 2
> select * from gv$lock where sid in
> (select blocking_session from gv$session where blocking_session is not
> null);
>
> My doubt is : what is the query that caused the lock that ran in the
> blocker sessions?
>
> I tried the query below, but did not get any query using the locked table.
>
> select * from
> (
> select sql_id
> ,to_char(last_active_time,'dd-hh24:mi:ss') last_active
> ,executions
> ,elapsed_time/1000000 elap_tot
> ,decode(executions,0,elapsed_time,(elapsed_time/executions))/1000000
> elap_exec
> ,decode(executions,0,disk_reads,(disk_reads/executions)) disk_exec
> ,decode(executions,0,buffer_gets,(buffer_gets/executions))
> buffer_exec
> ,tot
> ,sql_text
> from (select s.sql_id
> ,substr(s.sql_text,1,225) sql_text
> ,max(last_active_time) last_active_time
> ,sum(executions) executions
> ,sum(elapsed_time) elapsed_time
> ,sum(disk_reads) disk_reads
> ,sum(buffer_gets) buffer_gets
> ,count(*) tot
> from gv$sql s
> ,gv$open_cursor o
> where s.inst_id = o.inst_id
> and s.sql_id = o.sql_id
> and o.user_name = s.parsing_schema_name
> and o.sid=&vSid
> and o.inst_id = nvl('&vInstId',1)
> and s.parsing_schema_name <> 'SYS'
> group by s.sql_id
> ,substr(s.sql_text,1,225))
> order by to_char(last_active_time,'yyyymmddhh24mi'), elap_exec
> )
> where elap_exec > &vElap;
>
> What is the way to find out the query that caused the lock (in sessions
> showed in the Query 2 above)
>
> Regards
> Eriovaldo
>
>
>
>
> --
> Mladen Gogala
> Database Consultant
> Tel: (347) 321-1217
>
>

--
http://www.freelists.org/webpage/oracle-l