Re: What is v$mts telling me?

From: Daniel B. Bikle <dbikle_at_cco.caltech.edu>
Date: 24 Jul 1993 19:40:31 GMT
Message-ID: <22s37fINNdgc_at_gap.caltech.edu>


Since posting my original question I've learned a few points which I will share with you now.

Querying V$MTS



A simple query and its results are displayed below:

SELECT /* v$mts displays #'s for instance since last startup */

MAXIMUM_CONNECTIONS     /* max # connections for each dispatcher */
,SERVERS_STARTED        /* # servers automatically spawned */
,SERVERS_TERMINATED     /* # servers killed after they were not needed*/
,SERVERS_HIGHWATER      /* max # of simultaneous servers running at any */
FROM V$MTS              /* moment since last startup */
/

Example results obtained after pushing the dbs through a flurry of activity:

MAXIMUM_CONNECTIONS SERVERS_STARTED SERVERS_TERMINATED SERVERS_HIGHWATER

------------------- --------------- ------------------ -----------------
                 29              12                  9                 5

A process listing was obtained from the machine during the flurry of activity to verify the above query:

% ps -aux | grep x7

o7         352  0.0  4.0  212  872 ?  S    09:20   0:00 ora_pmon_x7
o7         353  0.0  4.0  208  856 ?  S    09:20   0:01 ora_dbwr_x7
o7         354  0.0  3.7  208  792 ?  S    09:20   0:00 ora_lgwr_x7
o7         355  0.0  0.0  332    0 ?  IW   09:20   0:02 ora_smon_x7
o7         356  0.0  0.0  228    0 ?  IW   09:20   0:00 ora_reco_x7
o7         357  0.0  0.6  280  132 ?  S    09:20   1:03 ora_s000_x7 <-
o7         358  0.0  0.6  276  132 ?  S    09:20   0:52 ora_s001_x7 <-
o7         455 10.3 24.1  220 5200 ?  R    09:34   0:05 ora_s002_x7 <-
o7         456  9.5 24.6  220 5292 ?  R    09:34   0:04 ora_s003_x7 <-
o7         457  9.5 23.7  220 5108 ?  R    09:34   0:03 ora_s004_x7 <-
o7         359  0.0  1.1  276  236 ?  S    09:20   0:22 ora_d000_x7
o7         360  0.0  1.1  276  236 ?  S    09:20   0:19 ora_d001_x7
o7         361  0.0  1.1  244  228 ?  S    09:21   0:00 ora_d002_x7

The scripts used to generate the flurry of activity are displayed below:

#! /bin/csh -f

# mtstst.sh

# tests connections to mts

setenv ORACLE_HOME /q/o7
setenv ORACLE_SID x7
setenv TWO_TASK ava2_x7

set path = ($ORACLE_HOME/bin $path)

# run a set of scripts to make the dbs busy
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.1.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.2.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.3.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.4.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.5.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.6.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.7.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.8.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.9.out &
$ORACLE_HOME/bin/sqlplus / _at_mtstst.sql > mtstst.10.out &

# wait a minute for the dbs to get busy
sleep 66

# obtain a process listing
ps -aux |fgrep x7 > mtstst.ps.out

# query the v$mts table; quote the $ sign so the shell doesn't # interpret it
$ORACLE_HOME/bin/sqlplus / > qrymts.lst << endofsql
select * from v\$mts
/
endofsql

# end of mtstst.sh

####################################################


rem
rem mtstst.sql
rem
SET ECHO ON

SELECT * FROM DICT              ;
SELECT * FROM V$MTS             ;
SELECT * FROM V$CIRCUIT         ;
SELECT * FROM V$DISPATCHER      ;
SELECT * FROM V$SESSION         ;
SELECT * FROM V$SHARED_SERVER   ;
SELECT * FROM DBA_TABLES        ;
SELECT * FROM DBA_INDEXES       ;
SELECT * FROM DBA_SYNONYMS      ;
SELECT * FROM CAT               ;

EXIT
rem end of mtstst.sql

The discussion above points out the obvious observation: V$MTS contains some of the information a DBA needs to pick a suitable value for for MTS_SERVERS. If SERVERS_STARTED and SERVERS_TERMINATED are 0 and SERVERS_HIGHWATER = MTS_SERVERS, then the instance has not been very busy since it was last started; the DBA might consider lowering the value of MTS_SERVERS if the instance has been up a long time.



Daniel B. Bikle
dbikle_at_alumni.caltech.edu
415/854-9542
Received on Sat Jul 24 1993 - 21:40:31 CEST

Original text of this message