Using current CVS (yesterday) I've rerun the benchmarks to see the
effects of various NAMEDATALEN settings.

3 times per setting.


First time is pgbench inserts (-s 5)
Second time is pgbench run (-t 3000 -s 5)

Third time is the postmaster during both of the above.

I'll run it again tonight on a computer with faster disk I/O to see if
that helps descrease the sys times.

-----
NAMEDATALEN: 32

       89.34 real         1.85 user         0.13 sys
      146.87 real         1.51 user         3.91 sys

      246.63 real        66.11 user        19.21 sys


NAMEDATALEN: 64

       93.10 real         1.82 user         0.16 sys
      147.30 real         1.45 user         3.90 sys

      249.28 real        66.01 user        18.82 sys


NAMEDATALEN: 128

       99.13 real         1.80 user         0.51 sys
      169.47 real         1.87 user         4.54 sys

      279.16 real        67.93 user        29.72 sys


NAMEDATALEN: 256

      106.60 real         1.81 user         0.43 sys
      166.61 real         1.69 user         4.25 sys

      283.76 real        66.88 user        26.59 sys


NAMEDATALEN: 32

       89.34 real         1.85 user         0.13 sys
      146.87 real         1.51 user         3.91 sys

      246.63 real        66.11 user        19.21 sys


NAMEDATALEN: 64

       93.10 real         1.82 user         0.16 sys
      147.30 real         1.45 user         3.90 sys

      249.28 real        66.01 user        18.82 sys


NAMEDATALEN: 128

       99.13 real         1.80 user         0.51 sys
      169.47 real         1.87 user         4.54 sys

      279.16 real        67.93 user        29.72 sys


NAMEDATALEN: 256

      106.60 real         1.81 user         0.43 sys
      166.61 real         1.69 user         4.25 sys

      283.76 real        66.88 user        26.59 sys


NAMEDATALEN: 512

       88.13 real         1.83 user         0.22 sys
      160.77 real         1.64 user         4.48 sys

      259.56 real        67.54 user        21.89 sys


#!/bin/sh

PGSRC=/home/rbt/postgresql/pgsqlnamedatalen
PGBASEPORT=6400
PGBASEBIN=/home/rbt/postgresql/dbname

LOG=/home/rbt/bench_namedatalen.log


for newDATALEN in 32 64 128 256 512 ; do

  PGBIN=${PGBASEBIN}_${newDATALEN}
  PGPORT=`echo "${PGBASEPORT}+${newDATALEN}" | bc`

  sed -E 's/NAMEDATALEN\s[0-9]+/NAMEDATALEN ${newDATALEN}/g' ${PGSRC}/src/include/postgres_ext.h > ${PGSRC}/src/include/postgres_ext.h.tmp
  mv ${PGSRC}/src/include/postgres_ext.h.tmp ${PGSRC}/src/include/postgres_ext.h

  cd ${PGSRC}
  ./configure --prefix=${PGBIN} --with-pgport=${PGPORT} || (echo "UNABLE TO CONFIGURE"; exit)

  make clean
  make clean install

  cd ${PGSRC}/contrib/pgbench

  gmake install


  ${PGBIN}/bin/initdb -D ${PGBIN}/data  || (echo "UNABLE TO INITIALIZE"; exit 1)

  time -a -o ${LOG} ${PGBIN}/bin/postmaster -D ${PGBIN}/data -F &

  sleep 5

  echo "NAMEDATALEN: ${newDATALEN}" >> ${LOG}
  echo  >> ${LOG}
  time -a -o ${LOG} ${PGBIN}/bin/pgbench -i -s 5 -d template1 -p ${PGPORT}

  time -a -o ${LOG} ${PGBIN}/bin/pgbench -t 3000 -s 5 -d template1 -p ${PGPORT}
  echo >> ${LOG}
  ${PGBIN}/bin/pg_ctl -D ${PGBIN}/data stop
  echo >> ${LOG}
  echo >> ${LOG}
  rm -rf ${PGBIN}
done

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to [EMAIL PROTECTED] so that your
message can get through to the mailing list cleanly

Reply via email to