From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <e@80x24.org>
X-Spam-Checker-Version: SpamAssassin 3.4.6 (2021-04-09) on dcvr.yhbt.net
X-Spam-Level: 
X-Spam-ASN:  
X-Spam-Status: No, score=-4.2 required=3.0 tests=ALL_TRUSTED,BAYES_00,
	DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF shortcircuit=no
	autolearn=ham autolearn_force=no version=3.4.6
Received: from localhost (dcvr.yhbt.net [127.0.0.1])
	by dcvr.yhbt.net (Postfix) with ESMTP id 9B8E41F47D;
	Sat, 25 Feb 2023 00:36:10 +0000 (UTC)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=80x24.org;
	s=selector1; t=1677285370;
	bh=ejP2rM0XIHL+GtNRYWa6EgxbCkY4m+8J9hMIrZ2gX+Y=;
	h=Date:From:To:Cc:Subject:References:In-Reply-To:From;
	b=gflkQLF/XmKag2OKabn3Ckt73Ueb7f/Hmkg6yBbu0MWBbTb4UEJ3qkBs6/ZkphpvQ
	 aJFkhYc/qLp+DsfHhA4ftuuyWUohXjpp1rhzqcWaU3toTIVZJ+l5gxK74d9YWKqsu3
	 LzK/aeveUZSjxXBwUliyeLtlSRCzABpI4mqf9iIs=
Date: Sat, 25 Feb 2023 00:36:10 +0000
From: Eric Wong <e@80x24.org>
To: Dale Hamel <dale.hamel@shopify.com>
Cc: raindrops-public@yhbt.net
Subject: Re: [PATCH] Fix queue stats for sockets with SO_REUSEPORT
Message-ID: <20230225003610.M538677@dcvr>
References: <20230224183544.51837-1-dale.hamel@shopify.com>
 <20230224190922.M629699@dcvr>
 <CA+G7BaHaftciDqK4gzE2ANVffB-DMdPaBAwCwCn1cB1i38jGCQ@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <CA+G7BaHaftciDqK4gzE2ANVffB-DMdPaBAwCwCn1cB1i38jGCQ@mail.gmail.com>
List-Id: <raindrops-public.yhbt.net>

Dale Hamel <dale.hamel@shopify.com> wrote:
> > I'm honestly shocked that these features get used;
> > I never convinced the org I worked for to use them :x
> 
> So far we haven't actually found a production use case
> for SO_REUSEPORT with Unicorn. We thought we had some
> promising early results - but it turned out to be this
> instrumentation error in raindrops.

Oh, I meant the socket instrumentation part of raindrops.

I figured atomic counters were the only thing that ever got used.

Anyways I think SO_REUSEPORT was too risky due to the
possibility of connection drops, but then again I haven't used
more than 4 real cores.

> At least by fixing this, efforts in the future will have the correct
> metrics to base tests off of.

Cool.

> > Anything else?  Could tag a release in a bit...
> 
> No, that's it! My apologies for sending as two patches,
> I didn't notice the bug in my test until I had already
> submitted, naturally.

No worries.  In the future, you can also ask me to do a squash
in the diffstat area or signature area, or use [SQUASH] instead of
[PATCH] in the subject prefix.

> Thanks very much for the quick review!

np.  Easy for me as long as I never leave a terminal :>