event 10046 raw output

From: <danisment_at_yahoo.com>
Date: Sat, 06 Jan 2001 11:47:17 GMT
Message-ID: <9370k5$erq$1@nnrp1.deja.com>

Hello,

I'm comparing event 10046 raw file with it's tkprof output.

I did not encounter a mismatch in user SQLs, but encountered in recursive SQLs. For example:

PARSING IN CURSOR #4 len=350 dep=2 uid=0 oct=3 lid=0 tim=1073270 hv=2216582187 ad='323ecf4'
select
ts#,file#,block#,nvl(bobj#,0),nvl(tab#,0),intcols,nvl(clucols,0),audit$
,flags,pctfree$,pctused$,initrans,maxtrans,rowcnt,blkcnt,empcnt,avgspc,
chncnt,avgrln,analyzetime,
samplesize,cols,property,nvl(degree,1),nvl(instances,1),avgspc_flb,flbc nt,kernelcols,nvl(trigflag,
0),nvl(spare1,0),nvl(spare2,0),spare4,nvl(spare3,0) from tab$ where obj#=:1
END OF STMT
PARSE #4:c=0,e=5,p=0,cr=0,cu=0,mis=1,r=0,dep=2,og=0,tim=1073275 BINDS #4:
bind 0: dty=2 mxl=22(22) mal=00 scl=00 pre=00 oacflg=08 oacfl2=1 size=24 offset=0
bfp=02bf8660 bln=22 avl=02 flg=05
value=59
EXEC #4:c=0,e=12,p=0,cr=0,cu=0,mis=0,r=0,dep=2,og=4,tim=1073287 FETCH #4:c=0,e=2,p=0,cr=3,cu=0,mis=0,r=1,dep=2,og=4,tim=1073289

tkprof output of raw:

select
ts#,file#,block#,nvl(bobj#,0),nvl(tab#,0),intcols,nvl(clucols,0), audit$,flags,pctfree$,pctused$,initrans,maxtrans,rowcnt,blkcnt,empcnt, avgspc,chncnt,avgrln,analyzetime,
samplesize,cols,property,nvl(degree,1),
nvl(instances,1),avgspc_flb,flbcnt,kernelcols,nvl(trigflag, 0),nvl(spare1,0)
,nvl(spare2,0),spare4,nvl(spare3,0)

from
tab$ where obj#=:1

call count cpu elapsed disk query current rows ------- ------ -------- ---------- ---------- ---------- ----------

---

Parse 1 0.00 0.05 0 0 0 0
Execute 2 0.00 0.12 0 0 0 0
Fetch 2 0.00 0.02 0 6 0 2
------- ------ -------- ---------- ---------- ---------- ----------

---

total 5 0.00 0.19 0 6 0 2

In raw file, there is 1 PARSE, 1 EXECUTE, 1 FETCH. But, tkprof output shows 1 PARSEs, 2 EXECUTEs, 2 FETCHes. and then some values of cpu/elepased/current etc. do not match with raw file(I guess absence of some phases). for sample above, there 6 consistent reads in FETCH phase. but raw output
shows 3 consistent reads (cr=3).

this mismatch is not for just this sample, but also another samples.

not: SQL given above is unique in raw file.

is recursive SQL calculated differently ?

Sent via Deja.com
http://www.deja.com/ Received on Sat Jan 06 2001 - 05:47:17 CST