Received: by 10.224.27.10 with SMTP id g10mr4962216qac.5.1320371934512;
        Thu, 03 Nov 2011 18:58:54 -0700 (PDT)
X-BeenThere: mochiweb@googlegroups.com
Received: by 10.224.73.197 with SMTP id r5ls1599554qaj.4.gmail; Thu, 03 Nov
 2011 18:58:53 -0700 (PDT)
Received: by 10.224.186.72 with SMTP id cr8mr4969149qab.2.1320371933439;
        Thu, 03 Nov 2011 18:58:53 -0700 (PDT)
Received: by 10.224.186.72 with SMTP id cr8mr4969148qab.2.1320371933431;
        Thu, 03 Nov 2011 18:58:53 -0700 (PDT)
Return-Path: <b...@redivi.com>
Received: from mail-qw0-f44.google.com (mail-qw0-f44.google.com [209.85.216.44])
        by gmr-mx.google.com with ESMTPS id j29si3425596qcs.2.2011.11.03.18.58.53
        (version=TLSv1/SSLv3 cipher=OTHER);
        Thu, 03 Nov 2011 18:58:53 -0700 (PDT)
Received-SPF: neutral (google.com: 209.85.216.44 is neither permitted nor denied by best guess record for domain of b...@redivi.com) client-ip=209.85.216.44;
Authentication-Results: gmr-mx.google.com; spf=neutral (google.com: 209.85.216.44 is neither permitted nor denied by best guess record for domain of b...@redivi.com) smtp.mail=...@redivi.com
Received: by mail-qw0-f44.google.com with SMTP id c10so2193017qad.31
        for <mochiweb@googlegroups.com>; Thu, 03 Nov 2011 18:58:53 -0700 (PDT)
MIME-Version: 1.0
Received: by 10.182.36.100 with SMTP id p4mr2443533obj.65.1320371933131; Thu,
 03 Nov 2011 18:58:53 -0700 (PDT)
Received: by 10.182.48.97 with HTTP; Thu, 3 Nov 2011 18:58:53 -0700 (PDT)
In-Reply-To: <CACwMPm8QJqqkHLd0kgNG0Mauafb9Bx=82oVTFq=jfgYMLH5myg@mail.gmail.com>
References: <802FBA65-E4A5-4802-92E9-95F884C93F52@gmail.com>
	<CACwMPm8QJqqkHLd0kgNG0Mauafb9Bx=82oVTFq=jfgYMLH5myg@mail.gmail.com>
Date: Thu, 3 Nov 2011 18:58:53 -0700
Message-ID: <CACwMPm_k4_XkqHE0svV7O=a8myDR+DLUEXQPSYZV=8OQU+i-uw@mail.gmail.com>
Subject: Re: [mochiweb] Newest Mochiweb performance issues?
From: Bob Ippolito <b...@redivi.com>
To: mochiweb@googlegroups.com
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

And of course, which platform?

On Thu, Nov 3, 2011 at 6:58 PM, Bob Ippolito <b...@redivi.com> wrote:
> What was the old version of mochiweb? What is your loop actually
> doing? Have you tried using the default configuration options? What
> version of Erlang are you using?
>
> On Thu, Nov 3, 2011 at 6:41 PM, Valentino Volonghi <dialt...@gmail.com> w=
rote:
>> I'm using mochiweb at revision 62e460dbd42 from github.
>>
>> The configuration parameters for the loop are:
>>
>> =C2=A0[{ip, "0.0.0.0"}, {port, 8000}, {backlog, 256}, {max, 16384}, {nod=
elay, true}, {acceptor_pool_size, 256}]
>>
>> with an older mochiweb this was the configuration:
>>
>> =C2=A0 [{ip, "0.0.0.0"}, {port, 8000}, {backlog, 200}, {max, 16384}]
>>
>>
>> The issue I'm seeing is that with the old mochiweb I was able to handle =
well over 3000 req/sec on a single server.
>>
>> With the new one, although during a test it was able to handle even more=
, after I reach 1000/1200 req/sec I start getting timeouts.
>>
>> I could think that it's in my code but even the simple ping page fails t=
o answer and there's nothing going on in it, it immediately returns pong.
>>
>> I don't see anything particular in the erlang logs that could make me th=
ink about an internal timeout between erlang processes due to my code, I do=
 see a lot of cases where the client disconnected from the server and when =
I try to send a response it fails in gen_tcp:send().
>>
>> with a log like this in sasl:
>> =3DCRASH REPORT=3D=3D=3D=3D 4-Nov-2011::00:00:03 =3D=3D=3D
>> =C2=A0crasher:
>> =C2=A0 =C2=A0initial call: mochiweb_acceptor:init/3
>> =C2=A0 =C2=A0pid: <0.12016.168>
>> =C2=A0 =C2=A0registered_name: []
>> =C2=A0 =C2=A0exception exit: enotconn
>> =C2=A0 =C2=A0 =C2=A0in function =C2=A0mochiweb_request:send/2
>> =C2=A0 =C2=A0 =C2=A0in call from mochiweb_request:start_raw_response/2
>> =C2=A0 =C2=A0 =C2=A0in call from mochiweb_request:respond/2
>> =C2=A0 =C2=A0 =C2=A0in call from dyno_web:loop/1
>> =C2=A0 =C2=A0 =C2=A0in call from mochiweb_http:headers/5
>> =C2=A0 =C2=A0ancestors: [dyno_web,dyno_sup,<0.66.0>]
>> =C2=A0 =C2=A0messages: []
>> =C2=A0 =C2=A0links: [<0.89.0>,#Port<0.4613065>]
>> =C2=A0 =C2=A0dictionary: [{random_seed,{1320,1117,9871}},
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0{mochiweb_=
request_path,
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0"/view/5108159f8eebac4d22a074c1ca82f8b1"},
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0{mochiweb_=
request_cookie,
>> =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =
=C2=A0[{"foobar","42a98512d10520496684b53c0b29a938"}]}]
>> =C2=A0 =C2=A0trap_exit: false
>> =C2=A0 =C2=A0status: running
>> =C2=A0 =C2=A0heap_size: 1597
>> =C2=A0 =C2=A0stack_size: 24
>> =C2=A0 =C2=A0reductions: 5122
>> =C2=A0neighbours:
>>
>> This problem unfortunately seem to happen only in production so for now =
I've switched the load back to the older servers. Anyone can help me shed a=
ny light on this?
>>
>> --
>> Valentino Volonghi
>> http://www.adroll.com
>>
>> --
>> You received this message because you are subscribed to the Google Group=
s "MochiWeb" group.
>> To post to this group, send email to mochiweb@googlegroups.com.
>> To unsubscribe from this group, send email to mochiweb+unsubscribe@googl=
egroups.com.
>> For more options, visit this group at http://groups.google.com/group/moc=
hiweb?hl=3Den.
>>
>>
>