<html>
<head>
<style><!--
.hmmessage P
{
margin:0px;
padding:0px
}
body.hmmessage
{
font-size: 10pt;
font-family:Tahoma
}
--></style></head>
<body class='hmmessage'><div dir='ltr'>
I thought I remember the question being asked but didn't see anything in my searches. Though I will admit I was pressed for time at work and probably didn't do as good a job at searching as I really should have. Thank you everyone for responding and especially for the report on your experiences with supervisors.<div><br></div><div>Chris</div><div><br><div><div id="SkyDrivePlaceholder"></div>> From: gleber.p@gmail.com<br>> Date: Thu, 25 Oct 2012 12:38:34 +0200<br>> Subject: Re: [erlang-questions] How much load can supervisors handle?<br>> To: jesper.louis.andersen@erlang-solutions.com<br>> CC: silent_vendetta@hotmail.com; erlang-questions@erlang.org<br>> <br>> > On Oct 24, 2012, at 11:18 PM, Chris Hicks <silent_vendetta@hotmail.com> wrote:<br>> >> Does anyone have any experience running one supervisor with a thousand,<br>> > ten-thousand, or more, workers under it with a high rate of churn? Would<br>> > a dynamically expanding tree of supervisors, which would obviously need to be<br>> > balanced, be a good solution for this?<br>> <br>> Lately my experience with simple_one_for_one was more or less like this:<br>> - thousands work without issues<br>> - tens of thousands are OK, but the limit is near and it can add some latency<br>> - hundred of thousands is stalling supervisor for minutes and system<br>> stops handling any requests which need spawning processes under the<br>> supervisor<br>> <br>> We solved this problem by "sharding" supervisors (spawn as many<br>> supervisors as there are schedulers) to ensure that all CPU cores are<br>> used.<br>> <br>> Another question: why does supervisors not use ETS tables? I believe<br>> it should be faster than "sets" for handling many thousands of entries<br>> (I haven't done any benchmarks here).<br></div></div> </div></body>
</html>