<div dir="ltr"><div class="gmail_quote"><div dir="ltr">On Sun, Jan 8, 2017 at 6:47 PM Gary E. Miller <<a href="mailto:gem@rellim.com">gem@rellim.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br class="gmail_msg">
I think not. Most of the threads will be servering clients. All a<br class="gmail_msg">
thread needs to serve a client is the incoming packet and the current<br class="gmail_msg">
local time. No locks needed for that.<br class="gmail_msg">
<br class="gmail_msg">
The client threads will prolly need some locking, but I doubt anyone<br class="gmail_msg">
has more than 40 refclocks or peers.<br class="gmail_msg">
<br class="gmail_msg">
The big win os for NTp servers with hundreds of clients.<br class="gmail_msg"></blockquote><div><br></div><div>That is the concurrency pattern that Erlang was designed for and wins at.</div><div><br></div><div>I am not recommending that we port NTPsec to Erlang, because the pool of available Erlang programmers with time to contribute is too low. </div><div><br></div><div>However, if I was a telco, spending a man-year or two writing an open source NTP server in Erlang as a lab project, and then upstreaming it into the various telco grade ecosystems would be a good idea.</div><div><br></div><div>..m</div></div></div>