The SO_REUSEPORT socket option [LWN.net]

The SO_REUSEPORT socket option

Posted Mar 15, 2013 2:02 UTC (Fri) by pr1268 (guest, #24648)
In reply to: The SO_REUSEPORT socket option by jengelh
Parent article: The SO_REUSEPORT socket option

> In other news, what Linux comes up is setting standards, because there have not been any standards before :)

Hey, Linux is THE trendsetter here. Linux hackers and organizations behind its development are certainly not going to wait for some standards body like POSIX, IETF, or whoever to get off their keisters to address this issue! Especially when we're...

...dealing with applications that accept 40,000 connections per second. :)

to post comments

The SO_REUSEPORT socket option

Posted Aug 24, 2013 21:12 UTC (Sat) by lyda (subscriber, #7429) [Link] (1 responses)

Well, again, it was in BSD first. Credit where it's due.

But generally, Unix "standards" have always trailed implementations. It's just that now Linux and BSD are in reality the primary Unix implementations.

(cue screams about which is the real unix)

The SO_REUSEPORT socket option

Posted Oct 1, 2014 16:09 UTC (Wed) by vsrinivas (subscriber, #56913) [Link]

DragonFly BSD has implemented SO_REUSEPORT since July 2013 too; http://lists.dragonflybsd.org/pipermail/users/2013-July/0... has some very interesting performance numbers.

SO_REUSEPORT naturally aligns with networks stacks parallelized like DFly's/Solaris's (hash connection state early, map & fanout to a fixed CPU per connection, no locking till you hit socket buffer layer).