From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-vb0-f43.google.com (mail-vb0-f43.google.com [209.85.212.43]) (using TLSv1 with cipher RC4-SHA (128/128 bits)) (Client CN "smtp.gmail.com", Issuer "Google Internet Authority" (verified OK)) by huchra.bufferbloat.net (Postfix) with ESMTPS id 153BD201A98 for ; Sun, 16 Dec 2012 11:42:32 -0800 (PST) Received: by mail-vb0-f43.google.com with SMTP id fs19so6217824vbb.16 for ; Sun, 16 Dec 2012 11:42:31 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:from:date:message-id:subject:to:content-type; bh=FL+7jhfW73katn3IfSwVZxTwCh8h3V5fL9t+301I9uc=; b=kG5aCw4Qz8TPUgOnH0t6FIxDr0gu8c7oNiP/uyc5MV+jnGJ0gjAFcdT9UR+/j/Lu+z UaA11raNyV9g/PJUyEhlSwa4E4P8QZE86QVYpLGvijg5k4ub32LNY5/4wup2v0yUIwet ynI5p6QSXrC4Ss/3DqO/O0z+1g23a8hdFexLKK5ghAPZVNALnZkoQdM3GPxSAvkwIupG 8nv1UvHoXRQoPaZ6cbu8zkNatGLPbm/j566cqvUwdno/j/KQWYiOrV8EnhbNt5C/CpiK zzt40jI9MmVu+MNaEvW69UVGunh+LBzU/MF2rNNi2qM/OUsAOy67XRp1iOC578iWN5mU F0pA== Received: by 10.221.0.79 with SMTP id nl15mr19899907vcb.41.1355686951729; Sun, 16 Dec 2012 11:42:31 -0800 (PST) MIME-Version: 1.0 Received: by 10.59.8.130 with HTTP; Sun, 16 Dec 2012 11:42:16 -0800 (PST) From: Jesper Louis Andersen Date: Sun, 16 Dec 2012 20:42:16 +0100 Message-ID: To: bloat Content-Type: multipart/alternative; boundary=bcaec54ee3702f3a4c04d0fd7742 Subject: [Bloat] What is the best current pseudo-code of the CODEL algorithm. X-BeenThere: bloat@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: General list for discussing Bufferbloat List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 16 Dec 2012 19:42:33 -0000 --bcaec54ee3702f3a4c04d0fd7742 Content-Type: text/plain; charset=UTF-8 I am building a load-constraint system, for Erlang, called safetyvalve[1]. Because this system has a concept of a queue I was wondering on applying some AQM to this queue and CODEL looks like an interesting one to ponder about. Especially because the algorithm is wonderfully simple. But is the appendix in the ACM queue article[2] the newest one or does it need some fixes to work correctly? Do note that I am aware that CODEL might not fit perfectly without modifications. But I am still interested in how it works because the algorithm looks like a very simple way of controlling a given target latency in a queue. The point with safetyvalve is to be able to define that a given web server can take, say, 80K connections simultaneously and then queue excess requests up to a point where a user waits more than 3-4 seconds and then send back an HTTP 503 (Overloaded) if this is the case. Thanks in advance, [1] https://github.com/jlouis/safetyvalve [2] http://queue.acm.org/appendices/codel.html -- J. --bcaec54ee3702f3a4c04d0fd7742 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable
I am building a load-constraint system, for Erlang, called safetyvalve= [1]. Because this system has a concept of a queue I was wondering on applyi= ng some AQM to this queue and CODEL looks like an interesting one to ponder= about. Especially because the algorithm is wonderfully simple. But is the = appendix in the ACM queue article[2] the newest one or does it need some fi= xes to work correctly?

Do note that I am aware that CODEL might not fit perfec= tly without modifications. But I am still interested in how it works becaus= e the algorithm looks like a very simple way of controlling a given target = latency in a queue. The point with safetyvalve is to be able to define that= a given web server can take, say, 80K connections simultaneously and then = queue excess requests up to a point where a user waits more than 3-4 second= s and then send back an HTTP 503 (Overloaded) if this is the case.

Thanks in advance,

[1]=C2=A0https://github.com/jlouis/safety= valve

--
J.
--bcaec54ee3702f3a4c04d0fd7742--