No subject


Thu Feb 16 03:44:05 GMT 2012


- the growth of queues is perfectly acceptable, whether consumers exist at
the time or not.

ISTM Rabbit makes a guess about whether the application is "steady state"
or "bursty", choosing "steady state" all the time. Rabbit team, is this the
case? Is there a way of configuring the flow control to permit unbounded
queue growth to support the "bursty" case? Or do I have the wrong end of
the stick entirely?


> 1. A client is publishing 1.5K/sec to each of four exchanges, each of
> which has a queue.
> 2. There are no consumers.  Therefore the queue is growing.
> 3. RMQ does not stop this.  "Flow control" does not trigger.
> 4. The consumers appear to begin to tear down the backlog.
>  5. RMQ per-connection flow control suddenly decides that now there are
> some consumers, now it has a reason to throttle the sender.
> 6. If the consumers had not shown up, the producer would not have been
> blocked.
>

7. The consumers drain some of the messages from the queues.
8. The producers are unblocked.
9. Profit!!!

Regards,
  Tony




-- 
Tony Garnock-Jones
tonygarnockjones at gmail.com
http://homepages.kcbbs.gen.nz/tonyg/

--14dae93403d7eb63fb04c0ec0197
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Aha, after this bounced, I'm now forwarding to the proper list address =
rather than the possibly-bogus google-groups alias.<br><br><div class=3D"gm=
ail_quote">---------- Forwarded message ----------<br>From: <b class=3D"gma=
il_sendername">Tony Garnock-Jones</b> <span dir=3D"ltr">&lt;<a href=3D"mail=
to:tonygarnockjones%2Brabbitmq at gmail.com">tonygarnockjones+rabbitmq at gmail.c=
om</a>&gt;</span><br>
Date: 26 May 2012 09:08<br>Subject: Re: [rabbitmq-discuss] Per-Connection F=
low Control -- The Case Against (Try #2)<br>To: Chip Salzenberg &lt;<a href=
=3D"mailto:rev.chip at gmail.com">rev.chip at gmail.com</a>&gt;<br>Cc: <a href=3D=
"mailto:rabbitmq-discuss at googlegroups.com">rabbitmq-discuss at googlegroups.co=
m</a><br>
<br><br>On 25 May 2012 18:57, Chip Salzenberg <span dir=3D"ltr">&lt;<a href=
=3D"mailto:rev.chip at gmail.com" target=3D"_blank">rev.chip at gmail.com</a>&gt;=
</span> wrote:<br><div class=3D"gmail_quote"><blockquote class=3D"gmail_quo=
te" style=3D"margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"=
>


<div>This is not acceptable.</div></blockquote><div><br></div><div>Why not?=
 I&#39;d be interested to learn what problems this is causing for you.</div=
><div><br></div><div>From one point of view, any time consumers exist but a=
re not keeping up with offered traffic is a good time to throttle the produ=
cers. Perhaps call this the &quot;steady state&quot; point of view - I&#39;=
m unsure if it has a better name already.</div>

<div><br></div><div>From another point of view - call this the &quot;bursty=
&quot; point of view, perhaps - the growth of queues is perfectly acceptabl=
e, whether consumers exist at the time or not.</div><div><br></div><div>

ISTM Rabbit makes a guess about whether the application is &quot;steady sta=
te&quot; or &quot;bursty&quot;, choosing &quot;steady state&quot; all the t=
ime. Rabbit team, is this the case? Is there a way of configuring the flow =
control to permit unbounded queue growth to support the &quot;bursty&quot; =
case? Or do I have the wrong end of the stick entirely?</div>
<div class=3D"im">
<div>=C2=A0</div><blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8=
ex;border-left:1px #ccc solid;padding-left:1ex"><div>1. A client is publish=
ing 1.5K/sec to each of four exchanges, each of which has a queue.</div>
<div>2. There are no consumers. =C2=A0Therefore the queue is growing.</div>=
<div>3. RMQ does not stop this. =C2=A0&quot;Flow control&quot; does not tri=
gger.</div><div>4. The consumers appear to begin to tear down the backlog.<=
/div>

<div>
5. RMQ per-connection flow control suddenly decides that now there are some=
 consumers, now it has a reason to throttle the sender.</div><div>6. If the=
 consumers had not shown up, the producer would not have been blocked.</div=
>

</blockquote><div><br></div></div><div>7. The consumers drain some of the m=
essages from the queues.</div><div>8. The producers are unblocked.</div><di=
v>9. Profit!!!</div><div><br></div><div>Regards,</div><div>=C2=A0 Tony</div=
>
<div><br>
</div></div>

</div><br><br clear=3D"all"><div><br></div>-- <br>Tony Garnock-Jones<br><a =
href=3D"mailto:tonygarnockjones at gmail.com" target=3D"_blank">tonygarnockjon=
es at gmail.com</a><br><a href=3D"http://homepages.kcbbs.gen.nz/tonyg/" target=
=3D"_blank">http://homepages.kcbbs.gen.nz/tonyg/</a><br>


--14dae93403d7eb63fb04c0ec0197--


More information about the rabbitmq-discuss mailing list