<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 12pt;
font-family:Calibri
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>Hi Bernd,<br><br>It is difficult to say how many strings are kept in the table when <br>your have 4 threads in parallel, but it seems the value 100003 fits quite well.<br>Larger values did not improve anything in my tests (I ignore changes < 2 seconds).<br><br>BTW: It's very difficult to get exact values. <br>If you execute mkgmap two times after not using it for a while<br>the 2nd run will always be faster because disk caches are already <br>filled with good values.<br>So, you first have to execute mkgmap a few times with exactly the<br>same parms and input files to find out how many seconds <br>it executes on average and how large the measurement error is.<br>You may already see 5 % difference here<br>when using multiple threads (--max-jobs), because sometimes<br>a worse case happens, sometimes not. This also seems to<br>depend on actions from JRE like just in time (JIT) compiles etc, <br>not talking about other background programs like system updates, <br>downloads, virus scanners, media players etc.<br><br>The worst case that can happen is that you find a good value for <br>one set of input files which is bad for many other sets of input files.<br><br>Gerd<br><br><div>> From: weigelt.bernd@web.de<br>> To: mkgmap-dev@lists.mkgmap.org.uk<br>> Date: Thu, 8 May 2014 18:28:22 +0200<br>> Subject: Re: [mkgmap-dev] Java tuning hint<br>> <br>> Am Donnerstag, 8. Mai 2014, 17:33:22 schrieb Gerd Petermann:<br>> <br>> Hello Gerd<br>> <br>> I've made three tests with my bonn extract<br>> <br>> without -XX:StringTableSize<br>> Time started: Thu May 08 17:57:43 CEST 2014<br>> Time finished: Thu May 08 18:04:54 CEST 2014<br>> Total time taken: 430804ms<br>> <br>> with -XX:StringTableSize=100003<br>> Time started: Thu May 08 18:07:33 CEST 2014<br>> Time finished: Thu May 08 18:14:20 CEST 2014<br>> Total time taken: 406425ms<br>> <br>> with -XX:StringTableSize=1000003<br>> time started: Thu May 08 18:16:20 CEST 2014<br>> Time finished: Thu May 08 18:23:06 CEST 2014<br>> Total time taken: 405556ms<br>> <br>> This are tests on my small laptop with 3.6 GB RAM with 64bit Linux<br>> <br>> Bernd<br>> <br>> > I tried both, and the smaller values seemed to work better for mkgmap.<br>> > I guess it depends on the input files what value is best, but at least <br>> > the default 1009 doesn't work well, so any much higher prime value should<br>> > help.<br>> <br>> _______________________________________________<br>> mkgmap-dev mailing list<br>> mkgmap-dev@lists.mkgmap.org.uk<br>> http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev<br></div> </div></body>
</html>