<div dir="ltr">Hi,<div><br></div><div>You might want to read this to get some insights.</div><div><br></div><div><a href="http://www.researchgate.net/profile/Konstantinos_Sagonas/publication/262172496_On_the_scalability_of_the_Erlang_term_storage/links/53f3a22f0cf2155be351bc3b.pdf?origin=publication_detail"></a><a href="http://www.researchgate.net/profile/Konstantinos_Sagonas/publication/262172496_On_the_scalability_of_the_Erlang_term_storage/links/53f3a22f0cf2155be351bc3b.pdf?origin=publication_detail"></a><a href="http://www.researchgate.net/profile/Konstantinos_Sagonas/publication/262172496_On_the_scalability_of_the_Erlang_term_storage/links/53f3a22f0cf2155be351bc3b.pdf?origin=publication_detail">http://www.researchgate.net/profile/Konstantinos_Sagonas/publication/262172496_On_the_scalability_of_the_Erlang_term_storage/links/53f3a22f0cf2155be351bc3b.pdf?origin=publication_detail</a><br></div><div><br></div><div>cheers,</div><div>Chandru</div><div><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On 15 May 2015 at 12:00, Viacheslav V. Kovalev <span dir="ltr"><<a href="mailto:kovyl2404@gmail.com" target="_blank">kovyl2404@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">I've added writer process to my test. It perfoms random write per each<br>
100 ms. Implementation is here:<br>
<a href="https://gist.github.com/kovyl2404/622a9908c4e8a2abc214" target="_blank">https://gist.github.com/kovyl2404/622a9908c4e8a2abc214</a><br>
<br>
There is example of run on 8-core machine.<br>
<br>
29> ets_read_concurrency:analyze(WithoutWriter).<br>
[{{procs,1},{percent,-8.998057178933019}},<br>
 {{procs,2},{percent,1.474232611754453}},<br>
 {{procs,4},{percent,-0.4318193657099243}},<br>
 {{procs,8},{percent,4.796912026714365}},<br>
 {{procs,16},{percent,4.926194326111598}},<br>
 {{procs,32},{percent,7.4091491628647805}},<br>
 {{procs,64},{percent,6.7226404897426315}},<br>
 {{procs,128},{percent,7.129140726319386}},<br>
 {{procs,256},{percent,-28.200373148451757}},<br>
 {{procs,512},{percent,10.229583687247757}},<br>
 {{procs,1024},{percent,25.824989572270635}}]<br>
30> ets_read_concurrency:analyze(WithWriter).<br>
[{{procs,1},{percent,-9.233383915316411}},<br>
 {{procs,2},{percent,1.3554058972355476}},<br>
 {{procs,4},{percent,-1.3437122387165232}},<br>
 {{procs,8},{percent,0.3944371727018411}},<br>
 {{procs,16},{percent,21.719493229803017}},<br>
 {{procs,32},{percent,-26.32711009412866}},<br>
 {{procs,64},{percent,11.577461884371825}},<br>
 {{procs,128},{percent,19.608517106505893}},<br>
 {{procs,256},{percent,11.362311552960543}},<br>
 {{procs,512},{percent,20.935963863004808}},<br>
 {{procs,1024},{percent,22.512472513575506}}]<br>
<br>
Where percent calculated as `(NonTweakedTime - TweakedTime)/NonTweaked*100`<br>
<br>
Tweaked table shows stable performance advantage either with or<br>
without writer process. I think I was wrong when interpreted former<br>
results, but frankly I expected to see more drastic changes. Again, I<br>
don't know how to interpret this artifacts:<br>
<br>
{{procs,256},{percent,-28.200373148451757}}. %% WithoutWriter<br>
{{procs,32},{percent,-26.32711009412866}}. %% WithWriter<br>
<br>
Is this just statistical error or something else?<br>
<div class="HOEnZb"><div class="h5">_______________________________________________<br>
erlang-questions mailing list<br>
<a href="mailto:erlang-questions@erlang.org">erlang-questions@erlang.org</a><br>
<a href="http://erlang.org/mailman/listinfo/erlang-questions" target="_blank">http://erlang.org/mailman/listinfo/erlang-questions</a><br>
</div></div></blockquote></div><br></div>