SocksJS performance with a lot of users - sockjs

The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Message from discussion SocksJS performance with a lot of users
Received-SPF: pass (google.com: domain of glebe...@gmail.com designates 209.85.212.49 as permitted sender) client-ip=209.85.212.49;
MIME-Version: 1.0
In-Reply-To: <CANZWjw7YHdkA1O9OA4zaFn+v1mZqMzTughgVg_44oc_SZEH...@mail.gmail.com>
References: <CANZWjw4yPegUGg4f7M3Odo0G_QEJPyRYcT5urkTd8r5uqnu...@mail.gmail.com>
	<CABzX+qwB+Wbd+qZcFM-RGXZE+jdvDve+4nt1gsK02Bd_QNu...@mail.gmail.com>
	<CADOS1qJvZARLQ0yoHvmyBPAy4qhAFcB3WpgY9s56ZOo75Du...@mail.gmail.com>
	<CANZWjw5u0YRGWrgeAWP6rveaNNS83ywh3Lim=DW8ua8K5Wq...@mail.gmail.com>
	<CABzX+qydQuATvyFsCiOk3THLBNSSVQAh1mSKnZ9pgtOpRg4...@mail.gmail.com>
	<CANZWjw50m9-fHtRiyRxv1fjBGiWqyM1h_3hZzjWjEMrfuVn...@mail.gmail.com>
	<CADOS1qL429BTqsfac52=GWGBo1CmuhtoXUroswe2uzEig7f...@mail.gmail.com>
	<CANZWjw5gE+DbMDdq77NmKRP+=ZmjzTGJYb_ZzEuK-eARRB7...@mail.gmail.com>
	<CADOS1qLYF3gCtihFGCAtgKedEtPMY2SrSmJZ059uHVW+sdo...@mail.gmail.com>
	<CANZWjw5OMgRwMpQKYf23fB+qeZATBnFCDePFAe4601yoRgu...@mail.gmail.com>
	<CADOS1qKdkpntH=3UrvTUTKiLmgBtbCGzKCc_n1jJVLY3VuP...@mail.gmail.com>
	<CANZWjw7YHdkA1O9OA4zaFn+v1mZqMzTughgVg_44oc_SZEH...@mail.gmail.com>
Date: Sat, 27 Oct 2012 09:47:25 +0200
Message-ID: <CADOS1qLAO=_8rqO1oEzAB0Zq9r4AgoaY82xQE_zs5wC+3kF...@mail.gmail.com>
Subject: Re: [sockjs] SocksJS performance with a lot of users
From: Gleb Peregud <glebe...@gmail.com>
To: AD <straightfl...@gmail.com>
Cc: Marek Majkowski <maje...@gmail.com>, sockjs <sockjs@googlegroups.com>
Content-Type: multipart/alternative; boundary=047d7b678494b6258904cd05a58f

--047d7b678494b6258904cd05a58f
Content-Type: text/plain; charset=UTF-8

Sounds good in general. But there's no need to do iolist_to_binary in the
end if you do properly prepare the frame to be sent. Also please note that
non-raw-websockets protocol needs different encodings, hence PreparedFame
will have multiple versions of encoded data in the end (I've used record
with 3 binary fields)

On Thursday, October 25, 2012, AD wrote:

> Here is my thought, let me know if this would work. This is just for raw
> websockets for now since this seems to be handled in different places
>
> 1) add a new method in sockjs_session called  send_multi/2 which would
> accept 2 params (Data, ListOfConn).
> 2) first step of send_multi would call sockjs_util:encode_frame on Data
> just once
> 3) loop through ListofConn and pass Frame output to send method in
> sockjs_session
> 4) modify sockjs_ws_handler:reply to NOT encode frame and just return
> iolist_to_binary(Frame)
>
> Thoughts ?
>
> On Thu, Oct 18, 2012 at 4:07 PM, Gleb Peregud <glebe...@gmail.com<javascript:_e({}, 'cvml', 'glebe...@gmail.com');>
> > wrote:
>
>> It's my own word, sorry for not explaining. Essentially I am doing it
>> this way:
>> 1) socket handler process (let's call it A) receives a request
>> 2) if a request is going to long-poll, then:
>> 3) A transfers ownership of socket to session process (let's call it B)
>> 4) A issues a infinity-timeout gen_server call with socket and Req to B
>> 5) A is sleeping
>> 6) B handles the request till the end of long-poll including sending the
>> data
>> 7) after it A resumes and finalizes all usual cowboy handling stuff
>> and, usually, dies
>>
>> On Thu, Oct 18, 2012 at 10:03 PM, AD <straightfl...@gmail.com<javascript:_e({}, 'cvml', 'straightfl...@gmail.com');>>
>> wrote:
>> > Thanks, not familiar with "attaching" a sockjs_session process really
>> means.
>> >
>> >
>> > On Wed, Oct 17, 2012 at 2:13 PM, Gleb Peregud <glebe...@gmail.com<javascript:_e({}, 'cvml', 'glebe...@gmail.com');>>
>> wrote:
>> >>
>> >> On Wed, Oct 17, 2012 at 8:09 PM, AD <straightfl...@gmail.com<javascript:_e({}, 'cvml', 'straightfl...@gmail.com');>>
>> wrote:
>> >> > Meaning you dont spawn a new sockjs_session for every connection ?
>> >>
>> >> I do. But that process "attaches" itself to sockjs_session process and
>> >> passes socket's ownership to session. This greatly reduce latency.
>> >>
>> >> For websocket connections, essentially, sockjs_ws_handler duplicates
>> >> all of sockjs_session functions, but in much simplified manner due to
>> >> simplicity of WS connection handling.
>> >
>> >
>>
>
>

--047d7b678494b6258904cd05a58f
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Sounds good in general. But there&#39;s no need to do iolist_to_binary in t=
he end if you do properly prepare the frame to be sent. Also please note th=
at non-raw-websockets protocol needs different encodings, hence PreparedFam=
e will have multiple versions of encoded data in the end (I&#39;ve used rec=
ord with 3 binary fields)<span></span><div>
<br>On Thursday, October 25, 2012, AD  wrote:<br><blockquote class=3D"gmail=
_quote" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:=
1ex">Here is my thought, let me know if this would work. This is just for r=
aw websockets for now since this seems to be handled in different places<di=
v>
<br></div><div>1) add a new method in sockjs_session called =C2=A0send_mult=
i/2 which would accept 2 params (Data, ListOfConn).</div>
<div>2) first step of send_multi would call sockjs_util:encode_frame on Dat=
a just once</div><div>3) loop through ListofConn and pass Frame output to s=
end method in sockjs_session</div><div>4) modify sockjs_ws_handler:reply to=
 NOT encode frame and just return iolist_to_binary(Frame)<br>

<br>Thoughts ?<br><br><div class=3D"gmail_quote">On Thu, Oct 18, 2012 at 4:=
07 PM, Gleb Peregud <span dir=3D"ltr">&lt;<a href=3D"javascript:_e({}, &#39=
;cvml&#39;, &#39;glebe...@gmail.com&#39;);" target=3D"_blank">gleber.p@gmai=
l.com</a>&gt;</span> wrote:<br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">
It&#39;s my own word, sorry for not explaining. Essentially I am doing it t=
his way:<br>
1) socket handler process (let&#39;s call it A) receives a request<br>
2) if a request is going to long-poll, then:<br>
3) A transfers ownership of socket to session process (let&#39;s call it B)=
<br>
4) A issues a infinity-timeout gen_server call with socket and Req to B<br>
5) A is sleeping<br>
6) B handles the request till the end of long-poll including sending the da=
ta<br>
7) after it A resumes and finalizes all usual cowboy handling stuff<br>
and, usually, dies<br>
<div><div><br>
On Thu, Oct 18, 2012 at 10:03 PM, AD &lt;<a href=3D"javascript:_e({}, &#39;=
cvml&#39;, &#39;straightfl...@gmail.com&#39;);" target=3D"_blank">straightf=
l...@gmail.com</a>&gt; wrote:<br>
&gt; Thanks, not familiar with &quot;attaching&quot; a sockjs_session proce=
ss really means.<br>
&gt;<br>
&gt;<br>
&gt; On Wed, Oct 17, 2012 at 2:13 PM, Gleb Peregud &lt;<a href=3D"javascrip=
t:_e({}, &#39;cvml&#39;, &#39;glebe...@gmail.com&#39;);" target=3D"_blank">=
glebe...@gmail.com</a>&gt; wrote:<br>
&gt;&gt;<br>
&gt;&gt; On Wed, Oct 17, 2012 at 8:09 PM, AD &lt;<a href=3D"javascript:_e({=
}, &#39;cvml&#39;, &#39;straightfl...@gmail.com&#39;);" target=3D"_blank">s=
traightfl...@gmail.com</a>&gt; wrote:<br>
&gt;&gt; &gt; Meaning you dont spawn a new sockjs_session for every connect=
ion ?<br>
&gt;&gt;<br>
&gt;&gt; I do. But that process &quot;attaches&quot; itself to sockjs_sessi=
on process and<br>
&gt;&gt; passes socket&#39;s ownership to session. This greatly reduce late=
ncy.<br>
&gt;&gt;<br>
&gt;&gt; For websocket connections, essentially, sockjs_ws_handler duplicat=
es<br>
&gt;&gt; all of sockjs_session functions, but in much simplified manner due=
 to<br>
&gt;&gt; simplicity of WS connection handling.<br>
&gt;<br>
&gt;<br>
</div></div></blockquote></div><br></div>
</blockquote></div>

--047d7b678494b6258904cd05a58f--
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
Account Options
