Smith. Along with Japanese doc updation by Tasuo Ishii.
> This patch changes the way pgbench outputs its latency log files so that
> every transaction gets a timestamp and notes which transaction type was
> executed. It's a one-line change that just dumps some additional
> information that was already sitting in that area of code. I also made a
> couple of documentation corrections and clarifications on some of the more
> confusing features of pgbench.
>
> It's straightforward to parse log files in this format to analyze what
> happened during the test at a higher level than was possible with the
> original format. You can find some rough sample code to convert this
> latency format into CVS files and then into graphs at
> http://www.westnet.com/~gsmith/content/postgresql/pgbench.htm which I'll
> be expanding on once I get all my little patches sent in here.
-$PostgreSQL: pgsql/contrib/pgbench/README.pgbench,v 1.16 2007/04/06 08:49:44 ishii Exp $
+$PostgreSQL: pgsql/contrib/pgbench/README.pgbench,v 1.17 2007/04/06 09:16:15 ishii Exp $
pgbench README
default is 1. NOTE: scaling factor should be at least
as large as the largest number of clients you intend
to test; else you'll mostly be measuring update contention.
+ Regular (not initializing) runs using one of the
+ built-in tests will detect scale based on the number of
+ branches in the database. For custom (-f) runs it can
+ be manually specified with this parameter.
-D varname=value
- Define a variable. It can be refereed to by a script
- provided by using -f option. Multile -D options are allowed.
+ Define a variable. It can be refered to by a script
+ provided by using -f option. Multiple -D options are allowed.
-U login
Specify db user's login name if it is different from
with the name "pgbench_log.xxx", where xxx is the PID
of the pgbench process. The format of the log is:
- client_id transaction_no time
+ client_id transaction_no time file_no time-epoch time-us
- where time is measured in microseconds.
+ where time is measured in microseconds, , the file_no is
+ which test file was used (useful when multiple were
+ specified with -f), and time-epoch/time-us are a
+ UNIX epoch format timestamp followed by an offset
+ in microseconds (suitable for creating a ISO 8601
+ timestamp with a fraction of a second) of when
+ the transaction completed.
+
+ Here are example outputs:
+
+ 0 199 2241 0 1175850568 995598
+ 0 200 2465 0 1175850568 998079
+ 0 201 2513 0 1175850569 608
+ 0 202 2038 0 1175850569 2663
-d
debug option.
(7) end;
+If you specify -N, (4) and (5) aren't included in the transaction.
+
o -f option
This supports for reading transaction script from a specified
-$PostgreSQL: pgsql/contrib/pgbench/README.pgbench_jis,v 1.17 2007/04/06 08:49:44 ishii Exp $
+$PostgreSQL: pgsql/contrib/pgbench/README.pgbench_jis,v 1.18 2007/04/06 09:16:16 ishii Exp $
pgbench README
\e$B%G%U%)%k%H$N%9%1!<%j%s%0%U%!%/%?!<$O\e(B 1 \e$B$G$9!%\e(B
-f \e$B%*%W%7%g%s$G;XDj$7$?%U%!%$%k$+$i%9%1!<%j%s%0%U%!%/\e(B
\e$B%?!<$r;2>H$9$k$K$O\e(B scale \e$B$H$$$&JQ?tL>$r;HMQ$7$^$9!%\e(B
+ \e$BCm0U\e(B: \e$B%9%1!<%j%s%0%U%!%/%?!<$O>/$J$/$H$b\e(B -c \e$B$G;XDj$5$l$kF1;~@\\e(B
+ \e$BB3%f!<%6?t$HF1$8$+$=$l$h$j$bBg$-$/$7$F$/$@$5$$!%$G$J$$$H!$\e(B
+ pgbench\e$B$N7k2L$O\e(BUPDATE\e$B$N6%9g$KBg$-$/1F6A$5$l$F$7$^$$$^$9!%\e(B
+ \e$BAH9~$_$N%Y%s%A%^!<%/%Q%?!<%s$G$O!$\e(Bbraches\e$B$N?t\e(B(=braches\e$B%F!<%V%k\e(B
+ \e$B$N9T?t\e(B)\e$B$,<+F0E*$K%9%1!<%j%s%0%U%!%/%?!<$K@_Dj$5$l$^$9!%\e(B
+ \e$B%+%9%?%`%/%(%j\e(B(-f\e$B%*%W%7%g%s;HMQ\e(B)\e$B$G$O!$%9%1!<%j%s%0%U%!%/%?!<$O\e(B
+ \e$B<+F0@_Dj$5$l$^$;$s!%\e(B
-D varname=value
\e$B%+%l%s%H%G%#%l%/%H%j0J2<$N\e(Bpgbench_log.xxx\e$B$H$$$&%U%!%$\e(B
\e$B%k$G$9!%%U%!%$%k$N%U%)!<%^%C%H$O!$\e(B
- \e$B%/%i%$%"%s%H\e(BID \e$B%H%i%s%6%/%7%g%sHV9f\e(B \e$B;~4V\e(B
+ \e$B%/%i%$%"%s%H\e(BID \e$B%H%i%s%6%/%7%g%sHV9f\e(B \e$B;~4V\e(B \e$B%U%!%$%kHV9f\e(B Unix\e$B%?%$%`\e(B(\e$BIC\e(B) \e$B%*%U%;%C%H\e(B
\e$B$H$J$C$F$$$^$9!%;~4V$O%^%$%/%mICC10L$G$9!%\e(B
+ \e$B%U%!%$%kHV9f$O%F%9%H$K;H$o$l$?%U%!%$%k$NHV9f$G$9\e(B(-f\e$B%*%W%7%g%s\e(B
+ \e$B$rJ#?t;XDj$7$?$H$-$KM-MQ$G$9\e(B)\e$B!%\e(B
+
+ \e$B!V\e(BUnix\e$B%?%$%`\e(B(\e$BIC\e(B) \e$B%*%U%;%C%H!W$O!$%H%i%s%6%/%7%g%s$,40N;$7$?$H\e(B
+ \e$B$-$N\e(BUNIX\e$B%(%]%C%/%?%$%`$H%^%$%/%mICC10L$N%*%U%;%C%H$G$9\e(B(ISO
+ 8601\e$B%?%$%`%9%?%s%W$HIC$NCf$NC<?t\e(B)\e$B!%\e(B
+
+ \e$B%m%0$N=PNONc$r<($7$^$9!%\e(B
+
+ 0 199 2241 0 1175850568 995598
+ 0 200 2465 0 1175850568 998079
+ 0 201 2513 0 1175850569 608
+ 0 202 2038 0 1175850569 2663
+
-d \e$B%G%P%C%0%*%W%7%g%s!%MM!9$J>pJs$,I=<($5$l$^$9!%\e(B
\e$B"#%G!<%?%Y!<%9$N=i4|2=\e(B
(7) end;
+-N\e$B%*%W%7%g%s$r;HMQ$9$k$H!$\e(B(4)(5)\e$B$O<B9T$5$l$^$;$s!%\e(B
+
\e$B"#F~NO%U%!%$%k$N%U%)!<%^%C%H\e(B
-f \e$B%*%W%7%g%s$r;XDj$7$F%H%i%s%6%/%7%g%s$K4^$^$l$k\e(B SQL \e$B%3%^%s%I$NFbMF$r\e(B
/*
- * $PostgreSQL: pgsql/contrib/pgbench/pgbench.c,v 1.63 2007/04/06 08:49:44 ishii Exp $
+ * $PostgreSQL: pgsql/contrib/pgbench/pgbench.c,v 1.64 2007/04/06 09:16:16 ishii Exp $
*
* pgbench: a simple benchmark program for PostgreSQL
* written by Tatsuo Ishii
diff = (int) (now.tv_sec - st->txn_begin.tv_sec) * 1000000.0 +
(int) (now.tv_usec - st->txn_begin.tv_usec);
- fprintf(LOGFILE, "%d %d %.0f\n", st->id, st->cnt, diff);
+ fprintf(LOGFILE, "%d %d %.0f %d %ld %ld\n",
+ st->id, st->cnt, diff, st->use_file, now.tv_sec,now.tv_usec);
}
if (commands[st->state]->type == SQL_COMMAND)
projects / postgresql / commitdiff