Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Mailing Lists -> Oracle-L -> High Parse Calls(--> high latch free events)

High Parse Calls(--> high latch free events)

From: Valiveru, Siva <SValiveru_at_ea.com>
Date: Tue, 22 May 2001 10:07:35 -0700
Message-ID: <F001.0030B09F.20010522100113@fatcity.com>

Hi All,
Oracle Version: 8162
OS: Solaris ( SunOS 5.6)

We are doing load test on our application and db. During the load test we ran Statspack ( before and after) and took the snap report fo the start and end timestamps, during the load we say very high parse calls per sec( 900 parse req/ sec peak, and 285 parse req/sec average), In the statspack report, in the top 5 wait events. event ' Latch Free' has waits for 9000, here is the partial statspack report. CPU idil time is around 30% during the load test.

Question is even if the statements being parsed are found in the shared pool and shared ( developers are using bind variables..identical statements ...etc etc)the load of a parse call is high? What areas I should be looking and candidates for the tuning.

Thanks for any help,

Siva

partial Stats Pack Report:

cache Sizes


           db_block_buffers: 131072

              db_block_size: 8192

                 log_buffer:     5242880

           shared_pool_size:   314572800

 


Load Profile


                                       Per Second      Per Transaction

                                  ---------------      ---------------

                  Redo size:            43,522.47               597.57

              Logical reads:             1,164.26                15.99

              Block changes:               292.49                 4.02

             Physical reads:                 1.01                 0.01

            Physical writes:                 2.50                 0.03

                 User calls:             1,524.96                20.94

                     Parses:               285.49                 3.92

                Hard parses:                 0.00                 0.00

                      Sorts:                22.77                 0.31

               Transactions:                72.83

 

              Rows per Sort:       10.44

  Pct Blocks changed / Read:       25.12

         Recursive Call Pct:       15.98

 Rollback / transaction Pct:        0.70

 


Instance Efficiency Percentages (Target 100%)


        Buffer Nowait Ratio: 99.44

        Buffer Hit Ratio: 99.91

        Library Hit Ratio: 100.01

        Redo NoWait Ratio: 100.00

       In-memory Sort Ratio: 99.96

           Soft Parse Ratio: 100.00

            Latch Hit Ratio: 99.44  

Top 5 Wait Events

~~~~~~~~~~~~~~~~                                             Wait     %
Total
Event                                               Waits  Time (cs)   Wt
Time
-------------------------------------------- ------------ ------------
-------
log file sync                                      38,815       56,837
81.58
log file parallel write                            18,998        6,492
9.32
latch free                                          9,206        3,192
4.58
buffer busy waits                                   3,184        2,381
3.42
db file parallel write                                 70          262
.38
         -------------------------------------------------------------

 

->cs - centisecond - 100th of a second

->ms - millisecond - 1000th of a second (unit often used for disk IO timings)  

                                                                   Avg

                                                     Total Wait   wait
Waits 
Event                               Waits   Timeouts  Time (cs)   (ms)
/txn
---------------------------- ------------ ---------- ----------- -----
------ 
log file sync                      38,815          0      56,837    15
1.1 
log file parallel write            18,998          0       6,492     3
0.5 
latch free                          9,206      8,321       3,192     3
0.3 
buffer busy waits                   3,184          4       2,381     7
0.1 
db file parallel write                 70          0         262    37
0.0 
SQL*Net more data to client        10,140          0         167     0
0.3 
control file parallel write           176          0         110     6
0.0 
enqueue                               337          0          91     3
0.0 
direct path write                     227          0          33     1
0.0 
direct path read                      309          0          32     1
0.0 
refresh controlfile command           112          0          22     2
0.0 
control file sequential read          741          0          19     0
0.0 
LGWR wait for redo copy               162          0          14     1
0.0 
log file switch completion              3          0          14    47
0.0 
file open                             317          0           4     0
0.0 
log file single write                   4          0           4    10
0.0 
file identify                           4          0           0     0
0.0 
log file sequential read                2          0           0     0
0.0 
db file sequential read                 1          0           0     0
0.0 
db file single write                    1          0           0     0
0.0 
SQL*Net message from client       747,297          0   4,894,938    66
20.9 
SQL*Net message to client         747,297          0         502     0
20.9 
         -------------------------------------------------------------

 

Background Wait Events  

                                                                        Avg

                                                     Total Wait   wait
Waits 
Event                               Waits   Timeouts  Time (cs)   (ms)
/txn
---------------------------- ------------ ---------- ----------- -----
------ 
log file parallel write            18,998          0       6,492     3
0.5 
db file parallel write                 70          0         262    37
0.0 
control file parallel write           176          0         110     6
0.0 
direct path read                      120          0          26     2
0.0 
LGWR wait for redo copy               162          0          14     1
0.0 
file open                             317          0           4     0
0.0 
log file single write                   4          0           4    10
0.0 
control file sequential read          374          0           3     0
0.0 
direct path write                      60          0           1     0
0.0 
file identify                           4          0           0     0
0.0 
log file sequential read                2          0           0     0
0.0 
db file sequential read                 1          0           0     0
0.0 
db file single write                    1          0           0     0
0.0 
rdbms ipc message                  35,448        429     205,163    58
1.0 
smon timer                              2          2      61,443 #####
0.0 
pmon timer                            159        159      48,965  3080
0.0 
      

Instance Activity Stats  

Statistic                                    Total   per Second    per Trans

--------------------------------- ---------------- ------------ ------------

parse count (total)                        139,890        285.5          3.9

parse time cpu                              11,138         22.7          0.3

parse time elapsed                          20,408         41.7          0.6



 

Buffer wait Statistics  

                                 Tot Wait    Avg

Class                    Waits  Time (cs) Time (cs)

------------------ ----------- ---------- ---------

data block               2,990      2,326         1

undo block                  67         41         1

undo header                127         22         0       


Latch Activity
->"Pct Misses" should be very close to 0.0  

                                                Pct    Avg
Pct 
                                    Get         Get Sleeps      Nowait
Nowait 
Latch Name                        Requests     Miss  /Miss    Requests
Miss
------------------------------- ------------ ------ ------ -----------
------ 
Active checkpoint queue latch            232    0.0                  0

Checkpoint queue latch                15,196    0.0                  0

Token Manager                            122    0.0                  0

archive control                          120    0.0                  0

archive process latch                      1    0.0                  0

cache buffers chains               1,386,788    0.4    0.1       3,221
0.7 
cache buffers lru chain              300,673    0.0    0.0       7,770
0.0 
dml lock allocation                   83,994    0.0    0.0           0

dropped object history latch               5    0.0                  0

enqueue hash chains                  156,451    0.2    0.1           0

enqueues                             272,986    0.1    0.1           0

job_queue_processes parameter l            7    0.0                  0

ktm global data                            1    0.0                  0

latch wait list                        7,363    1.0    0.1       7,327
0.1 
library cache                      1,292,955    1.9    0.3           0

list of block allocation              70,950    0.0    0.0           0

loader state object freelist              20    0.0                  0

messages                             131,361    0.1    0.0           0

ncodef allocation latch                    7    0.0                  0

redo allocation                      110,048    0.1    0.0           0

redo writing                         148,423    0.5    0.0           0

row cache objects                      1,546    0.0                  0

sequence cache                             3    0.0                  0

session allocation                   279,492    0.3    0.0           0

session idle bit                   1,634,184    0.1    0.3           0

session switching                          7    0.0                  0

shared pool                              483    0.0                  0

sort extent pool                          11    0.0                  0

transaction allocation               111,840    0.1    0.0           0

transaction branch allocation              7    0.0                  0

undo global data                     226,340    0.2    0.0           0

         -------------------------------------------------------------

 

Latch Sleep breakdown  

                                 Get                              Spin &

Latch Name                     Requests        Misses      Sleeps Sleeps 1-4

---------------------------- ------------ ----------- -----------
------------
library cache                   1,292,955      25,000       7,891
17723/6695/5
                                                                  56/26/0

 

cache buffers chains            1,386,788       5,005         540
4469/532/4/0
                                                                  /0

 

session idle bit                1,634,184       1,950         633
1329/609/12/
                                                                  0/0

 

session allocation                279,492         838          22
816/22/0/0/0
redo writing                      148,423         719          19
700/19/0/0/0
undo global data                  226,340         430          13
417/13/0/0/0
enqueues                          272,986         380          44
336/44/0/0/0
enqueue hash chains               156,451         289          32
261/24/4/0/0
transaction allocation            111,840          99           2 98/0/1/0/0

redo allocation                   110,048          96           3 93/3/0/0/0

latch wait list                     7,363          76           5 72/3/1/0/0

dml lock allocation                83,994          39           1 38/1/0/0/0

list of block allocation           70,950          28           1 27/1/0/0/0

         -------------------------------------------------------------

 

Latch Miss Sources  

                                                          No Wait

Latch Name                     Where                       Misses
Sleeps
------------------------------ -------------------------- -------
----------- 
cache buffers chains           kcbgtcr: kslbegin                0
378 
cache buffers chains           kcbchg: kslbegin: bufs not       0
83 
cache buffers chains           kcbgcur: kslbegin                0
40 
cache buffers chains           kcbchg: kslbegin: call CR        0
21 
cache buffers chains           kcbnlc                           0
7 
cache buffers chains           kcbrls: kslbegin                 0
5 
cache buffers chains           kcbzwb                           0
5 
cache buffers chains           kcbzgb: scan from tail. no       0
1 
dml lock allocation            ktaidm                           0
1 
enqueue hash chains            ksqgtl3                          0
24 
enqueue hash chains            ksqrcl                           0
8 
enqueues                       ksqies                           0
30 
enqueues                       ksqgtl2                          0
12 
enqueues                       ksqgel: create enqueue           0
2 
latch wait list                kslges                           8
4 
latch wait list                kslfre                           6
1 
library cache                  kglhdgn: child:                  0
7,333 
library cache                  kglpnal: child: alloc spac       0
198 
library cache                  kglpnal: child: before pro       0
161 
library cache                  kglpnc: child                    0
85 
library cache                  kgldte: child 0                  0
25 
library cache                  kgllkdl: child: cleanup          0
24 
library cache                  kglhdgc: child:                  0
23 
library cache                  kglupc: child                    0
22 
library cache                  kgllkdl: child: free pin         0
17 
list of block allocation       ktlbbl                           0
1 
redo allocation                kcrfwr: redo allocation          0
3 
redo writing                   kcrfsr                           0
12 
redo writing                   kcrfss                           0
6 
redo writing                   kcrfwi: after write              0
1 
session allocation             ksucri                           0
12 
session allocation             ksuxds: not user session         0
10 
session idle bit               ksupuc: set busy                 0
334 
session idle bit               ksupuc: clear busy               0
266 
session idle bit               ksuxds                           0
33 
transaction allocation         ktcxba                           0
2 
undo global data               kturdc: KSLBEGIN                 0
5 
undo global data               ktubnd                           0
4 
undo global data               ktudba: KSLBEGIN                 0
3 
undo global data               ktudnx: KSLBEGIN                 0          
-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.com
-- 
Author: Valiveru, Siva
  INET: SValiveru_at_ea.com

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from).  You may
also send the HELP command for other information (like subscribing).
Received on Tue May 22 2001 - 12:07:35 CDT

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US