Thursday, 17 May 2012

Oracle SQL Tuning 101


Oracle Tuning And Diagnostics Script

--Active Session Info
SELECT b.sid, b.serial#, a.spid, b.sql_id, b.program, b.osuser, b.machine,
b.TYPE, b.event, b.action, b.p1text, b.p2text, b.p3text, b.state, c.sql_text,b.logon_time
FROM v$process a, v$session b, v$sqltext c
WHERE a.addr=b.paddr
AND b.sql_hash_value = c.hash_value
AND b.STATUS = 'ACTIVE'
ORDER BY a.spid, c.piece
 
--Trace SQL Query Average Execution Time Using SQL ID
SELECT sql_id, child_number, plan_hash_value plan_hash, executions execs,
(elapsed_time/1000000)/decode(nvl(executions,0),0,1,executions) avg_etime,
buffer_gets/decode(nvl(executions,0),0,1,executions) avg_lio, sql_text
FROM v$sql s
WHERE s.sql_id='4n01r8z5hgfru'
 
--Get The Detail Explain Plan Using SQL ID
SELECT plan_table_output FROM TABLE(dbms_xplan.display_cursor('dtdqt19kfv6yx'))
 
--Session Elapsed Processing Time 
SELECT s.sid, s.username, s.module,
round(t.VALUE/1000000,2) "Elapsed Processing Time (Sec)"
FROM v$sess_time_model t, v$session s
WHERE t.sid = s.sid
AND t.stat_name = 'DB time'
AND s.username IS NOT NULL
AND t.VALUE/1000000 >= '1' --running more than 1 second
ORDER BY round(t.VALUE/1000000,2) DESC
 
--Session Elapsed Processing Time Statistic By SID
SELECT  a.sid, b.username, a.stat_name, ROUND((a.VALUE/1000000),2) "Time (Sec)"
FROM v$sess_time_model a, v$session b
WHERE a.sid = b.sid
AND b.sid = '194'  
ORDER BY ROUND((a.VALUE/1000000),2) DESC
 
--Use Longops To Check The Estimation Query Runtime
SELECT sid, serial#, opname, target, sofar, totalwork, units, start_time, 
last_update_time, time_remaining "REMAIN SEC", round(time_remaining/60,2) "REMAIN MINS", 
elapsed_seconds "ELAPSED SEC", round(elapsed_seconds/60,2) "ELAPSED MINS", 
round((time_remaining+elapsed_seconds)/60,2)"TOTAL MINS", message TIME 
FROM v$session_longops
WHERE sofar<>totalwork
AND time_remaining <> '0'
 
--Detect Blocking Session
SELECT sid, serial#, username, STATUS, state, event, 
blocking_session, seconds_in_wait, wait_time, action, logon_time
FROM gv$session 
WHERE state IN ('WAITING') 
AND wait_class != 'Idle'
AND event LIKE '%enq%'
AND TYPE='USER'
 
--Active Table Locking
SELECT b.sid, b.serial#, b.program, b.osuser, b.machine, b.TYPE, b.action,
c.sql_text,b.logon_time, e.owner, e.object_name "Table Lock"
FROM v$session b, v$sqltext c, v$locked_object d, dba_objects e
WHERE b.sql_address = c.address
AND b.sid = d.session_id
AND d.object_id = e.object_id
AND b.STATUS = 'ACTIVE'
ORDER BY b.sid, c.piece
 
--RAC Active Table Locking
SELECT b.sid, b.serial#, a.spid, b.program, b.osuser, b.machine, 
b.TYPE, b.event, b.action, b.p1text, b.p2text, b.p3text, 
b.state, c.sql_text,b.logon_time, 
b.STATUS,  e.owner, e.object_name "Table Lock"
FROM gv$process a, gv$session b, gv$sqltext c, gv$locked_object d, dba_objects e
WHERE a.addr=b.paddr
AND b.sql_address = c.address
AND b.sid = d.session_id
AND d.object_id = e.object_id
AND b.STATUS = 'ACTIVE'
ORDER BY a.spid, c.piece
 
--Monitor Top Waiting Event Using Active Session History (ASH)
SELECT h.event,
SUM(h.wait_time + h.time_waited) "Total Wait Time (ms)"
FROM v$active_session_history h, v$sqlarea SQL, dba_users u, v$event_name e 
WHERE h.sample_time BETWEEN sysdate - 1/24 AND sysdate --event in the last hour
AND h.sql_id = SQL.sql_id
AND h.user_id = u.user_id
AND h.event# = e.event#
GROUP BY h.event
ORDER BY SUM(h.wait_time + h.time_waited) DESC
 
--Monitor Highest SQL Wait Time Using Active Session History (ASH)
SELECT h.session_id, h.session_serial#, h.sql_id, h.session_state, 
h.blocking_session_status, h.event, e.wait_class, h.module, u.username, SQL.sql_text,
SUM(h.wait_time + h.time_waited) "Total Wait Time (ms)"
FROM v$active_session_history h, v$sqlarea SQL, dba_users u, v$event_name e 
WHERE h.sample_time BETWEEN sysdate - 1/24 AND sysdate --event in the last hour
AND h.sql_id = SQL.sql_id
AND h.user_id = u.user_id
AND h.event# = e.event#
GROUP BY h.session_id, h.session_serial#, h.sql_id, h.session_state, 
h.blocking_session_status, h.event, e.wait_class, h.module, u.username, SQL.sql_text
ORDER BY SUM(h.wait_time + h.time_waited) DESC
 
--Monitor Highest Object Wait Time Using Active Session History (ASH)
SELECT o.owner, o.object_name, o.object_type, h.session_id, h.session_serial#, 
h.sql_id, h.module, SUM(h.wait_time + h.time_waited) "Total Wait Time (ms)"
FROM v$active_session_history h, dba_objects o, v$event_name e 
WHERE h.sample_time BETWEEN sysdate - 1/24 AND sysdate --event in the last hour
AND h.current_obj# = o.object_id
AND e.event_id = h.event_id
GROUP BY  o.owner, o.object_name, o.object_type, h.session_id, h.session_serial#,
h.sql_id, h.module
ORDER BY SUM(h.wait_time + h.time_waited) DESC
 
--Monitor Highest Event Wait Time Using Active Session History (ASH)
SELECT h.event "Wait Event", SUM(h.wait_time + h.time_waited) "Total Wait Time (ms)"
FROM v$active_session_history h, v$event_name e
WHERE h.sample_time BETWEEN sysdate - 1/24 AND sysdate --event in the last hour
AND h.event_id = e.event_id
AND e.wait_class <> 'Idle'
GROUP BY h.event
ORDER BY SUM(h.wait_time + h.time_waited) DESC
 
--Database Time Model Statistic
SELECT wait_class, NAME, ROUND (time_secs, 2) "Time (Sec)",
ROUND (time_secs * 100 / SUM (time_secs) OVER (), 2) pct
FROM 
(SELECT n.wait_class, e.event NAME, e.time_waited / 100 time_secs
FROM v$system_event e, v$event_name n
WHERE n.NAME = e.event 
AND n.wait_class <> 'Idle'
AND time_waited > 0
UNION
SELECT 
'CPU', 
'Server CPU', 
SUM (VALUE / 1000000) time_secs
FROM v$sys_time_model
WHERE stat_name IN ('background cpu time', 'DB CPU'))
ORDER BY time_secs DESC;
 
--Monitor I/O On Data Files
SELECT vfs.file#, dbf.file_name, dbf.tablespace_name, dbf.bytes, vfs.phyrds/vfs.phywrts,
vfs.phyblkrd/vfs.phyblkwrt, vfs.readtim, vfs.writetim 
FROM v$filestat vfs, dba_data_files dbf
WHERE vfs.file# = dbf.file_id
 
--I/O Stats For Data Files & Temp Files
SELECT file_no,
filetype_name, 
small_sync_read_reqs "Synch Single Block Read Reqs",
small_read_reqs "Single Block Read Requests",
small_write_reqs "Single Block Write Requests",
round(small_sync_read_latency/1000,2) "Single Block Read Latency (s)",
large_read_reqs "Multiblock Read Requests",
large_write_reqs "Multiblock Write Requests",
async_io "Asynch I/O Availability"
FROM v$iostat_file
WHERE filetype_id IN (2,6) --data file and temp file
 
--I/O Stats By Functionality
SELECT function_name,
small_read_reqs "Single Block Read Requests",
small_write_reqs "Single Block Write Requests",
large_read_reqs "Multiblock Read Requests",
large_write_reqs "Multiblock Write Requests",
number_of_wait "I/O Waits",
round(wait_time/1000,2) "Total Wait Time (ms)"
FROM v$iostat_function
ORDER BY function_name
 
--Temporary Tablespace Usage By SID
SELECT tu.username, s.sid, s.serial#, s.sql_id, s.sql_address, tu.segtype, 
tu.extents, tu.blocks, SQL.sql_text
FROM v$tempseg_usage tu, v$session s, v$sql SQL
WHERE tu.session_addr = s.addr
AND tu.session_num = s.serial#
AND s.sql_id = SQL.sql_id
AND s.sql_address = SQL.address
 
--Monitor Overall Oracle Tablespace 
SELECT d.STATUS "Status",
d.tablespace_name "Name",
d.contents "Type",
d.extent_management "Extent Management",
d.initial_extent "Initial Extent",
TO_CHAR(NVL(a.bytes / 1024 / 1024, 0),'99,999,990.900') "Size (M)",
TO_CHAR(NVL(a.bytes - NVL(f.bytes, 0), 0)/1024/1024,'99,999,999.999') "Used (M)",
TO_CHAR(NVL((a.bytes - NVL(f.bytes, 0)) / a.bytes * 100, 0), '990.00') "Used %",
TO_CHAR(NVL(a.maxbytes / 1024 / 1024, 0),'99,999,990.900') "MaxSize (M)",
TO_CHAR(NVL((a.bytes - NVL(f.bytes, 0)) / a.maxbytes * 100, 0), '990.00') "Used % of Max"
FROM sys.dba_tablespaces d,
(SELECT tablespace_name, 
SUM(bytes) bytes, 
SUM(decode(autoextensible,'NO',bytes,'YES',maxbytes))
maxbytes FROM dba_data_files GROUP BY tablespace_name) a,
(SELECT tablespace_name, SUM(bytes) bytes FROM dba_free_space 
GROUP BY tablespace_name) f
WHERE d.tablespace_name = a.tablespace_name(+)
AND d.tablespace_name = f.tablespace_name(+)
ORDER BY 10 DESC;
 
--Cache Hit Ratio
SELECT ROUND(((1-(SUM(DECODE(name,
'physical reads', VALUE,0))/
(SUM(DECODE(name, 'db block gets', VALUE,0))+
(SUM(DECODE(name, 'consistent gets', VALUE, 0))))))*100),2)
|| '%' "Buffer Cache Hit Ratio"
FROM v$sysstat --Use gv$sysstat if running on RAC environment
 
--Library Cache Hit Ratio
SELECT SUM(pins) "Total Pins", SUM(reloads) "Total Reloads",
SUM(reloads)/SUM(pins) *100 libcache
FROM v$librarycache --Use v$librarycache if running on RAC environment
 
--DB Session Memory Usage
SELECT se.sid,n.name, MAX(se.VALUE) maxmem
FROM v$sesstat se, v$statname n
WHERE n.statistic# = se.statistic#
AND n.name IN ('session pga memory','session pga memory max',
'session uga memory','session uga memory max')
GROUP BY n.name, se.sid
ORDER BY MAX(se.VALUE) DESC

Useful Oracle Hint For Tuning – Gather_Plan_Statistics

The hint /*+ gather_plan_statistics */ enables collection of rowsource execution statistics for the statement.
The hint is useful if you want to have clear overview of the execution plan chosen by Oracle together with estimated rows or actual rows fetch in the SQL execution.
SET serveroutput off;
SET linesize 300;
 
SELECT /*+  gather_plan_statistics */ SUM(nvl(period_net_dr,0) - nvl(period_net_cr,0)) 
FROM gl_balances
WHERE code_combination_id = '133110'
GROUP BY code_combination_id;
 
SELECT * FROM TABLE(dbms_xplan.display_cursor(NULL, NULL, 'last allstats'));

Oracle Wait Events FAQ

Buffer Busy Waits

Area: Buffer Cache, DBWR
Possible Causes: The block is being read into the buffer by another session, so the waiting session must wait for the block read to complete.
Examine: Examine V$SESSION while the problem is occurring to determine the type of block in contention. Tune SQL to access rows with fewer block reads by adding indexes, adjusting the database writer or adding freelists to tables and indexes. Even if there is a huge db_cache_size, the DBA may still see buffer busy waits and increasing the buffer size will not help in this case.

Free Buffer Waits

Area: Buffer cache, DBWR, I/O
Possible Causes: Slow DBWR (possibly due to I/O?). Cache too small
Examine: Examine write time using operating system statistics. Check buffer cache statistics for evidence of too small cache.

DB File Scattered Read

Area: I/O, SQL statement tuning
Possible Causes: Poorly written SQL. Slow I/O system
Examine: Investigate V$SQLAREA to see whether there are SQL statements performing many disk reads. Cross-check I/O system and V$FILESTAT for poor read time.

DB File Sequential Read

Area: I/O, SQL statement tuning
Possible Causes: Poorly written SQL. Slow I/O system
Examine: Investigate V$SQLAREA to see whether there are SQL statements performing many disk reads. Cross-check I/O system and V$FILESTAT for poor read time.

Enqueue Waits (enq:)

Area: Locks
Possible Causes: Depends on type of enqueue
Examine: Look at V$ENQUEUE_STAT

Library Cache Latch Waits (librarycache, library cache pin and library cache lock)

Area: Latch contention
Possible Causes: SQL parsing or sharing
Examine: Check V$SQLAREA to see whether there are SQL statements with a relatively high number of parse calls or a high number of child cursors (column VERSION_COUNT). Check parse statistics inV$SYSSTAT and their corresponding rate for each second.

Log Buffer Space

Area: Log buffer, I/O
Possible Causes: Log buffer small. Slow I/O system
Examine: Check the statistic redo buffer allocation retries in V$SYSSTAT. Check configuring log buffer section in configuring memory chapter. Check the disks that house the online redo logs for resource contention.

Log File Sync

Area: I/O, over-commit
Possible Causes: Slow disks that store the online logs. Unbatched commits
Examine: Check the disks that house the online redo logs for resource contention. Check the number of transactions (commits + rollbacks) each second, from V$SYSSTAT.

Instance Tuning Using Performance Views – Oracle

V$ACTIVE_SESSION_HISTORY – view displays active database session activity, sampled once every second.
V$SESS_TIME_MODEL and V$SYS_TIME_MODEL – views contain time model statistics, including DB time which is the total time spent in database calls
V$SESSION_WAIT – view displays the resources or events for which active sessions are waiting.
V$SESSION – view contains the same wait statistics that are contained in the V$SESSION_WAIT view. If applicable, this view also contains detailed information on the object that the session is currently waiting for (object number, block number, file number, and row number), plus the blocking session responsible for the current wait.
V$SESSION_EVENT – view provides summary of all the events the session has waited for since it started.
V$SESSION_WAIT_CLASS – view provides the number of waits and the time spent in each class of wait events for each session.
V$SESSION_WAIT_HISTORY – view provides the last ten wait events for each active session.
V$SYSTEM_EVENT – view provides a summary of all the event waits on the instance since it started.
V$EVENT_HISTOGRAM – view displays a histogram of the number of waits, the maximum wait, and total wait time on a per-child cursor basis.
V$FILE_HISTOGRAM – view displays a histogram of times waited during single block reads for each file.
V$SYSTEM_WAIT_CLASS – view provides the instance wide time totals for the number of waits and the time spent in each class of wait events. This view also shows the object number for which the session is waiting.
V$TEMP_HISTOGRAM – view displays a histogram of times waited during single block reads for each temporary file.

Tuning Oracle Shared Pool

Sometimes your database might hit the error ORA-04031: unable to allocate 4048 bytes of shared memory (“shared pool”,”unknown object”,”sga heap(1,0)”,”library cache”).
You can use the dbms_shared_pool package to pin large packages, reduce your use of shared memory or increase the amount of available shared memory by increasing the value of the INIT.ORA parameters “shared_pool_reserved_size” and “shared_pool_size”.
Increasing the shared pool size might resolve the problem but it is not necessary will improve the performance of library caching.
Always check the advisory with v$shared_pool_advice view to see the estimated time improvement statistic if you increase the shared pool area.
--Free Memory Available In Shared Pool
SELECT *
FROM v$sgastat
WHERE name = 'free memory'
AND pool = 'shared pool'
 
--Database Memory Setting
SELECT component, current_size/1024/1024 "Current Size (MB)",
min_size/1024/1024 "Min Size (MB)", max_size/1024/1024 "Max Size (MB)",
user_specified_size/1024/1024 "User Specified Size (MB)", 
last_oper_type "Type" 
FROM v$sga_dynamic_components
 
--Estimation Of Shared Memory Pool Size vs. Time Saved
SELECT shared_pool_size_for_estimate "Pool Size (MB)",
estd_lc_size "Lib Cache Size (MB)", 
estd_lc_time_saved/1000000 "Lib Cache Time Saved (Sec)"
FROM v$shared_pool_advice

Tuning Oracle Buffer Cache

v$db_cache_advice contains rows that predict the number of physical reads for the cache size corresponding to each row. This view is populated when the DB_CACHE_ADVICE initialization parameter is set to ON.
When the advisory is enabled, there is a small increase in CPU usage as overhead associated with this advisory.
The rows also compute a “physical read factor” which is the ratio of the number of estimated reads to the number of reads actually performed by the real buffer cache during the measurement interval.
--Estimation Of Buffer Cache Size vs. Physical Reads
SELECT size_for_estimate "Cache Size (MB)", 
buffers_for_estimate "Buffers", 
estd_physical_read_factor "Estd Phys|Read Factor", 
estd_physical_reads "Estd Phys| Reads"
FROM v$db_cache_advice
WHERE name = 'DEFAULT'
AND block_size = (SELECT VALUE FROM V$PARAMETER WHERE name = 'db_block_size')
AND advice_status = 'ON'
 
                                Estd Phys    Estd Phys
 Cache SIZE (MB)      Buffers READ Factor        Reads
---------------- ------------ ----------- ------------
              30        3,802       18.70  192,317,943      10% OF CURRENT SIZE 
              60        7,604       12.83  131,949,536
              91       11,406        7.38   75,865,861
             121       15,208        4.97   51,111,658
             152       19,010        3.64   37,460,786
             182       22,812        2.50   25,668,196
             212       26,614        1.74   17,850,847
             243       30,416        1.33   13,720,149
             273       34,218        1.13   11,583,180
             304       38,020        1.00   10,282,475      CURRENT SIZE 
             334       41,822         .93    9,515,878
             364       45,624         .87    8,909,026
             395       49,426         .83    8,495,039
             424       53,228         .79    8,116,496
             456       57,030         .76    7,824,764
             486       60,832         .74    7,563,180
             517       64,634         .71    7,311,729
             547       68,436         .69    7,104,280
             577       72,238         .67    6,895,122
             608       76,040         .66    6,739,731      200% OF CURRENT SIZE
The following output shows that if the cache was 212 MB, rather than the current size of 304 MB, the estimated number of physical reads would increase by a factor of 1.74 or 74%. This means it would not be advisable to decrease the cache size to 212MB.
However, increasing the cache size to 334MB would potentially decrease reads by a factor of .93 or 7%. If an additional 30MB memory is available on the host machine and the SGA_MAX_SIZE setting allows the increment, it would be advisable to increase the default buffer cache pool size to 334MB.

Related Posts

No comments:

Post a Comment

Number of Visitors