Scribd Logo
Upload

Welcome to Scribd!

  • 17 views

    Practice 18 - Handling Enqueue Waits

    Uploaded by

    Abdo Mohamed
    This document describes steps to simulate and diagnose enqueue wait events in an Oracle database: 1. Simulate an enqueue wait by having two sessions update the same row simultaneously. 2. Diagnose current waits by identifying blocking/waiting sessions and details on the waiting session. 3. Diagnose recent waits using ASH reports and V$ views after the lock is released. 4. Diagnose past waits using AWR reports to identify historical enqueue wait issues. Regular monitoring is recommended to detect frequent enqueue waits that impact performance.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Practice 18 - Handling Enqueue Waits

    Uploaded by

    Abdo Mohamed
    17 views7 pages
    This document describes steps to simulate and diagnose enqueue wait events in an Oracle database: 1. Simulate an enqueue wait by having two sessions update the same row simultaneously. 2. Diagnose current waits by identifying blocking/waiting sessions and details on the waiting session. 3. Diagnose recent waits using ASH reports and V$ views after the lock is released. 4. Diagnose past waits using AWR reports to identify historical enqueue wait issues. Regular monitoring is recommended to detect frequent enqueue waits that impact performance.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document describes steps to simulate and diagnose enqueue wait events in an Oracle database: 1. Simulate an enqueue wait by having two sessions update the same row simultaneously. 2. Diagnose current waits by identifying blocking/waiting sessions and details on the waiting session. 3. Diagnose recent waits using ASH reports and V$ views after the lock is released. 4. Diagnose past waits using AWR reports to identify historical enqueue wait issues. Regular monitoring is recommended to detect frequent enqueue waits that impact performance.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    17 views7 pages

    Practice 18 - Handling Enqueue Waits

    Uploaded by

    Abdo Mohamed
    This document describes steps to simulate and diagnose enqueue wait events in an Oracle database: 1. Simulate an enqueue wait by having two sessions update the same row simultaneously. 2. Diagnose current waits by identifying blocking/waiting sessions and details on the waiting session. 3. Diagnose recent waits using ASH reports and V$ views after the lock is released. 4. Diagnose past waits using AWR reports to identify historical enqueue wait issues. Regular monitoring is recommended to detect frequent enqueue waits that impact performance.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 7

    Practice 18 - Handling Enqueue Waits P a g e |1

    Practice 18

    Handling Enqueue Waits

    Practice Target
    In this practice, you will simulate an enqueue wait event and use multiple tools to diagnose it.

    Oracle Database Performance Tuning, a course by Ahmed Baraka


    Practice 18 - Handling Enqueue Waits P a g e |2

    Handling 'enq: TX - row lock contention'


    In this section of the practice, you will simulate and handle the 'enq: TX - row lock contention' wait
    event.

    Simulating the Enqueue

    1. Make sure srv1 is started. If it is not, start it up.

    2. Start Putty, login to srv1 as oracle. Invoke SQL*Plus and login to ORADB as sysdba. In the rest
    of this practice, this window is referred to as the admin window.
    sqlplus / as sysdba

    3. Start another two Putty sessions. In each of them login to srv1 as oracle, invoke SQL*Plus and
    login to ORADB as soe. In the rest of this practice, these windows are referred to as the client
    windows.
    sqlplus soe/soe

    4. In one of the client windows, obtain the current time then issue the update statement that
    follows. Take a note of the retrieved output time.
    SELECT TO_CHAR(SYSDATE, 'MM/DD/YY HH24:MI:SS') ctime FROM DUAL;

    var V_CUSTOMER_ID number


    exec :V_CUSTOMER_ID := 100
    UPDATE CUSTOMERS
    SET CUST_EMAIL = CUST_EMAIL || '' WHERE CUSTOMER_ID = :V_CUSTOMER_ID;

    5. In the other client window, issue the following update statement.


    The command hangs because the row is exclusively locked by the other session.
    UPDATE CUSTOMERS SET CUST_EMAIL = CUST_EMAIL || '' WHERE CUSTOMER_ID =100;

    Oracle Database Performance Tuning, a course by Ahmed Baraka


    Practice 18 - Handling Enqueue Waits P a g e |3

    Diagnosing Current Enqueue Events

    In the following steps, you will diagnose the enqueue wait events that are currently held in the
    system. At this stage, the user is currently blocked.

    6. In the admin window, issue the following query to detect the sessions involved in the locking
    issue.
    The query returns the SID of the blocking session and the waiting session.
    col SESSIONS format A20
    SELECT DECODE(REQUEST,0,'Holder SID: ','Waiter SID: ') ||
    SID SESSIONS, ID1, ID2, LMODE, REQUEST, TYPE
    FROM V$LOCK
    WHERE (ID1, ID2, TYPE) IN (SELECT ID1, ID2, TYPE FROM V$LOCK WHERE REQUEST > 0)
    ORDER BY ID1, REQUEST;

    7. Run the following query to display more information about the waiting session.
    set pagesize 20
    col INFO for a200
    SELECT
    'SID: '|| S.SID || CHR(10) ||
    'USERNAME: '|| S.USERNAME || CHR(10) ||
    'EVENT: ' || S.EVENT || CHR(10) ||
    'DESCRIPTION: ' || T.DESCRIPTION || CHR(10) ||
    'CURRENT STATEMENT: ' || Q.SQL_TEXT || CHR(10) ||
    'WAITING TIME (s): ' || S.SECONDS_IN_WAIT || CHR(10) ||
    'P1TEXT: ' || S.P1TEXT || CHR(10) ||
    'P1: ' || S.P1 || CHR(10) ||
    'P2TEXT: ' || S.P2TEXT || CHR(10) ||
    'P2: ' || S.P2 || CHR(10) ||
    'P3TEXT: ' || S.P3TEXT || CHR(10) ||
    'P3: ' || S.P3
    AS INFO
    FROM V$SESSION S, V$LOCK L, V$LOCK_TYPE T, V$SQL Q
    WHERE S.SID = L.SID AND T.TYPE=L.TYPE AND S.SQL_ID = Q.SQL_ID
    AND L.REQUEST>0;

    8. Run the following query to construct a query that retrieves the row affected by the locked row.
    Obtain the SID of waiting session from the output of the preceding step.
    SELECT 'SELECT * FROM "' || O.OWNER || '"."' || O.OBJECT_NAME || '"
    WHERE ROWID = DBMS_ROWID.ROWID_CREATE(1, ' || S.ROW_WAIT_OBJ# || ', ' ||
    S.ROW_WAIT_FILE# || ', ' || ROW_WAIT_BLOCK# || ', ' ||
    ROW_WAIT_ROW# || ');'
    FROM DBA_OBJECTS O, V$SESSION S
    WHERE S.ROW_WAIT_OBJ# = O.OBJECT_ID AND S.SID = &V_WSID;

    9. Run the query produced by the preceding step.

    Oracle Database Performance Tuning, a course by Ahmed Baraka


    Practice 18 - Handling Enqueue Waits P a g e |4

    Diagnosing Recent Enqueue Events

    In the preceding steps, you diagnosed the wait events of currently locked sessions. In the following
    steps, you will diagnose the issue after the lock is released but the lock information is still in the
    memory (information about them can be obtained from the V$ views).

    As the enqueue wait events have been recently released, we can obtain information about them from
    ASH and from the V$ views.

    10. In the blocking session window, release the lock (by issuing a rollback or commit statement).
    ROLLBACK;

    11. Verify that the lock has been released.


    SELECT DECODE(REQUEST,0,'Holder SID: ','Waiter SID: ') ||
    SID SESSIONS, ID1, ID2, LMODE, REQUEST, TYPE
    FROM V$LOCK
    WHERE (ID1, ID2, TYPE) IN (SELECT ID1, ID2, TYPE FROM V$LOCK WHERE REQUEST > 0)
    ORDER BY ID1, REQUEST;

    12. Generate ASH report for the locking period.


    The code assumes the issue happened within the past ten minutes. If this is not the case in your
    environment, fix the begin_time and duration values accordingly.
    define dbid = '';
    define inst_num = '';
    define report_type = 'html';
    define begin time
    define begin_time ='-10'
    define duration = 10;
    define report_name = '/media/sf_extdisk/ash_enqtx.html';
    define slot_width = '';
    define target_session_id = '';
    define target_sql_id = '';
    define target_wait_class = '';
    define target_service_hash = '';
    define target_module_name = '';
    define target_action_name = '';
    define target_client_id = '';
    define target_plsql_entry = '';
    define target_container = '';
    @ $ORACLE_HOME/rdbms/admin/ashrpti.sql

    13. Open the generated report and study it.

    Oracle Database Performance Tuning, a course by Ahmed Baraka


    Practice 18 - Handling Enqueue Waits P a g e |5

    In the preceding steps, we dealt with an issue that has been reported to us.

    But sometimes, enqueue wait events happen in brief periods. Sometimes the users do not report
    about the delay cause by them because they believe the system is just like that.

    Part of the DBA proactive monitoring routine should be to monitor the 'enq: TX - row lock contention'.
    If they frequently take place in the system, we need to perform an investigation.

    14. Run the following query to retrieve the enqueue wait events ever recorded in the database since
    the last time the instance was started.
    col EVENT format a40
    col WAIT_CLASS format a11
    SELECT EVENT, AVERAGE_WAIT,
    TO_CHAR(ROUND(TIME_WAITED/100),'999,999,999') TIME_SECONDS, WAIT_CLASS
    FROM V$SYSTEM_EVENT
    WHERE EVENT LIKE 'enq%'
    ORDER BY TIME_WAITED;

    15. Run the following query to return the objects that were waited for their locks to be released.
    The query returns the number of waits, not the wait time.
    col OBJECT_NAME for a10
    col stats for a20
    SELECT OBJECT_NAME, SUBSTR(STATISTIC_NAME, 1, 30) STATS, VALUE
    FROM V$SEGMENT_STATISTICS
    WHERE ( STATISTIC_NAME IN( 'ITL waits' , 'row lock waits' ))
    AND VALUE>0 AND OBJECT_NAME NOT LIKE 'BIN$%'
    ORDER BY VALUE DESC;

    16. Retrieve the SQL statements that waited the most for wait events of type "Application".
    In many cases, this corresponds to the TX enqueue waits.
    col SQL_TEXT for a30
    SELECT ROUND(APPLICATION_WAIT_TIME / 1000000) WAIT_TIME_S,
    SQL_ID, SUBSTR(SQL_TEXT,1,30) SQL_TEXT
    FROM V$SQLSTATS
    WHERE APPLICATION_WAIT_TIME >0
    ORDER BY APPLICATION_WAIT_TIME DESC FETCH FIRST 10 ROW ONLY;

    Oracle Database Performance Tuning, a course by Ahmed Baraka


    Practice 18 - Handling Enqueue Waits P a g e |6

    Diagnosing Past Enqueue Events

    In the following steps, you will diagnose the enqueue wait events that happened in the past
    (information about them is not available in the V$ views anymore).

    As the enqueue wait events happened in the past, we only can obtain information about it from AWR
    or ASH history. If the issue is older than the AWR retention period, we cannot obtain information
    about it.

    17. Create an AWR snapshot.


    exec DBMS_WORKLOAD_REPOSITORY.CREATE_SNAPSHOT;

    18. Produce an AWR report with HTML format using the most recent snapshots that cover the
    incident time. Give the produced report the name awr_enq_tx.html

    @ ?/rdbms/admin/awrrpt.sql

    19. Move the report to the staging folder.


    host mv awr_enq_tx.html /media/sf_extdisk/

    20. Open the AWR report and diagnose the issue.

    Oracle Database Performance Tuning, a course by Ahmed Baraka


    Practice 18 - Handling Enqueue Waits P a g e |7

    Summary

    • Enqueue wait events are events that should not frequently occur in healthy systems. They can
    be diagnosed using various performance tools.

    Oracle Database Performance Tuning, a course by Ahmed Baraka

    You might also like