Output Saturation Error - OpenVPN Users

The old Google Groups will be going away soon, but your browser is incompatible with the new version.
Message from discussion Output Saturation Error
Received-SPF: pass (google.com: domain of openvpn-users-boun...@lists.sourceforge.net designates 216.34.181.88 as permitted sender) client-ip=216.34.181.88;
Received-SPF: pass (3b2kzd1.ch3.sourceforge.com: domain of gmail.com
	designates 209.85.198.251 as permitted sender)
	client-ip=209.85.198.251; envelope-from=nya...@gmail.com;
	helo=rv-out-0708.google.com; 
MIME-Version: 1.0
In-Reply-To: <34aee8f10905292232v41cf60aej1f9bb3f0ff83eab6@mail.gmail.com>
References: <7DD0C0AC30DE4BE699E8E6DFFE29907A@carteslap>
	<9618647f0905292040v71694282n877248229fda0...@mail.gmail.com>
	<34aee8f10905292232v41cf60aej1f9bb3f0ff83e...@mail.gmail.com>
Date: Sun, 31 May 2009 04:37:54 +0600
Message-ID: <9618647f0905301537y66e052ecm3764c5def1eab...@mail.gmail.com>
From: Nyamul Hassan <mnhas...@usa.net>
To: OpenVPN List <openvpn-us...@lists.sourceforge.net>
Subject: Re: [Openvpn-users] Output Saturation Error
Precedence: list
Content-Type: multipart/mixed; boundary="===============7016985033437779996=="
Errors-To: openvpn-users-boun...@lists.sourceforge.net

--===============7016985033437779996==
Content-Type: multipart/alternative; boundary=000e0cd1460613faea046b28d8ab

--000e0cd1460613faea046b28d8ab
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit

Thank you Michael for your email.  Yes, you're absolutely right.  I've been
doing some reading during this time too, and have stumbled on the elaborate
explanation from the CIPE site.  We have been running GRE during the
endpoints, and the latency is excellent.  Only 5 ms "mdev" on an average of
320 ms.
I've tried doing OpenVPN over UDP, but it seems to fail, giving errors on
the server side about some "read UDPv4 [ECONNREFUSED]: Connection refused
(code=111)".  Someone on the mailing list in the past had pointed out this
means that the UDP to the client failed.  The client is on a statically
assigned private ip which is natted over the internet.  Could this be the
reason?

TCP seems to be a sure bet, as it bypasses almost any firewall rules, and
you can pretend to be a genuine service like HTTP or POP3, and also do nice
stuff like "http-proxy".  I was just wondering, is there a way to make the
TCP stack ignore the inherent "reliability" measures for either of the TCP
packet groups (wrapper or payload of the tunnel)?

Regards
HASSAN



On Sat, May 30, 2009 at 11:32 AM, Michael Chletsos <mpchl...@gmail.com>wrote:

> It looks like you are running this over TCP, correct me if I am wrong,
> but I think that your voip application will also be using tcp, and
> basically tcp over tcp is a bad idea, see
> http://sites.inka.de/~W1011/devel/tcp-tcp.html   I think that you want
> to run openvpn over udp.
>
> good luck.
>
>
> On Fri, May 29, 2009 at 9:40 PM, Nyamul Hassan <mnhas...@usa.net> wrote:
> > Hi,
> > To add my config a bit more:
> > Server is a AMD Opteron Dual Core with 4G RAM hosted in a data center
> with
> > 50Mbps limit
> > Host is a Intel Core 2 Duo with 4G RAM hosted locally sharing a 50 Mbps
> line
> > We are trying to run a VoIP application over this link, and we're seeing
> > these errors, couple with high latency variations.
> > Can someone please advise?
> > Regards
> > HASSAN
> >
> >
> >
> > On Thu, May 28, 2009 at 4:00 PM, Nyamul Hassan <mnhas...@usa.net> wrote:
> >>
> >> Hi,
> >>
> >> We have started running a small VPN link between two remote destination,
> >> and
> >> we are getting the following error:
> >>
> >> Thu May 28 15:57:27 2009 us=721222 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=721806 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=723656 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=724482 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=724504 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=724804 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=726005 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=726027 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=726448 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >> Thu May 28 15:57:27 2009 us=727143 xxx/a.b.c.d:z MULTI: packet dropped
> due
> >> to output saturation (multi_process_incoming_tun)
> >>
> >> There is no shortage of BW in either end.  Interestingly, even when the
> >> tunnel size is only ~250kbps, it suddenly gives a few bursts of these
> >> errors.  Then, it stops giving these errors.  Then, the tunnel size
> keeps
> >> increasing, and intermittently, these errors show up for a second or so.
> >>
> >> Could someone point to me what I could be doing wrong?
> >>
> >> Regards
> >> HASSAN
> >>
> >>
> >>
> >>
> ------------------------------------------------------------------------------
> >> Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
> >> is a gathering of tech-side developers & brand creativity professionals.
> >> Meet
> >> the minds behind Google Creative Lab, Visual Complexity, Processing, &
> >> iPhoneDevCamp as they present alongside digital heavyweights like
> >> Barbarian
> >> Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com
> >> _______________________________________________
> >> Openvpn-users mailing list
> >> Openvpn-us...@lists.sourceforge.net
> >> https://lists.sourceforge.net/lists/listinfo/openvpn-users
> >>
> >
> >
> >
> ------------------------------------------------------------------------------
> > Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT
> > is a gathering of tech-side developers & brand creativity professionals.
> > Meet
> > the minds behind Google Creative Lab, Visual Complexity, Processing, &
> > iPhoneDevCamp as they present alongside digital heavyweights like
> Barbarian
> > Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com
> > _______________________________________________
> > Openvpn-users mailing list
> > Openvpn-us...@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/openvpn-users
> >
> >
>
>

--000e0cd1460613faea046b28d8ab
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

Thank you Michael for your email. =A0Yes, you&#39;re absolutely right. =A0I=
&#39;ve been doing some reading during this time too, and have stumbled on =
the elaborate explanation from the CIPE site. =A0We have been running GRE d=
uring the endpoints, and the latency is excellent. =A0Only 5 ms &quot;mdev&=
quot; on an average of 320 ms.<div>
<br></div><div>I&#39;ve tried doing OpenVPN over UDP, but it seems to fail,=
 giving errors on the server side about some &quot;read UDPv4 [ECONNREFUSED=
]: Connection refused (code=3D111)&quot;. =A0Someone on the mailing list in=
 the past had pointed out this means that the UDP to the client failed. =A0=
The client is on a statically assigned private ip which is natted over the =
internet. =A0Could this be the reason?</div>
<div><br></div><div>TCP seems to be a sure bet, as it bypasses almost any f=
irewall rules, and you can pretend to be a genuine service like HTTP or POP=
3, and also do nice stuff like &quot;http-proxy&quot;. =A0I was just wonder=
ing, is there a way to make the TCP stack ignore the inherent &quot;reliabi=
lity&quot; measures for either of the TCP packet groups (wrapper or payload=
 of the tunnel)?</div>
<div><br></div><div>Regards</div><div>HASSAN</div><div><br></div><div><br><=
br><div class=3D"gmail_quote">On Sat, May 30, 2009 at 11:32 AM, Michael Chl=
etsos <span dir=3D"ltr">&lt;<a href=3D"mailto:mpchl...@gmail.com">mpchlets@=
gmail.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 looks like you are running this over TCP=
, correct me if I am wrong,<br>
but I think that your voip application will also be using tcp, and<br>
basically tcp over tcp is a bad idea, see<br>
<a href=3D"http://sites.inka.de/~W1011/devel/tcp-tcp.html" target=3D"_blank=
">http://sites.inka.de/~W1011/devel/tcp-tcp.html</a> =A0 I think that you w=
ant<br>
to run openvpn over udp.<br>
<br>
good luck.<br>
<div><div></div><div class=3D"h5"><br>
<br>
On Fri, May 29, 2009 at 9:40 PM, Nyamul Hassan &lt;<a href=3D"mailto:mnhass=
a...@usa.net">mnhas...@usa.net</a>&gt; wrote:<br>
&gt; Hi,<br>
&gt; To add my config a bit more:<br>
&gt; Server is a AMD Opteron Dual Core with 4G RAM hosted in a data center =
with<br>
&gt; 50Mbps limit<br>
&gt; Host is a Intel Core 2 Duo with 4G RAM hosted locally sharing a 50 Mbp=
s line<br>
&gt; We are trying to run a VoIP application over this link, and we&#39;re =
seeing<br>
&gt; these errors, couple with high latency variations.<br>
&gt; Can someone please advise?<br>
&gt; Regards<br>
&gt; HASSAN<br>
&gt;<br>
&gt;<br>
&gt;<br>
&gt; On Thu, May 28, 2009 at 4:00 PM, Nyamul Hassan &lt;<a href=3D"mailto:m=
nhas...@usa.net">mnhas...@usa.net</a>&gt; wrote:<br>
&gt;&gt;<br>
&gt;&gt; Hi,<br>
&gt;&gt;<br>
&gt;&gt; We have started running a small VPN link between two remote destin=
ation,<br>
&gt;&gt; and<br>
&gt;&gt; we are getting the following error:<br>
&gt;&gt;<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D721222 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D721806 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D723656 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D724482 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D724504 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D724804 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D726005 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D726027 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D726448 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt; Thu May 28 15:57:27 2009 us=3D727143 xxx/a.b.c.d:z MULTI: packet d=
ropped due<br>
&gt;&gt; to output saturation (multi_process_incoming_tun)<br>
&gt;&gt;<br>
&gt;&gt; There is no shortage of BW in either end. =A0Interestingly, even w=
hen the<br>
&gt;&gt; tunnel size is only ~250kbps, it suddenly gives a few bursts of th=
ese<br>
&gt;&gt; errors. =A0Then, it stops giving these errors. =A0Then, the tunnel=
 size keeps<br>
&gt;&gt; increasing, and intermittently, these errors show up for a second =
or so.<br>
&gt;&gt;<br>
&gt;&gt; Could someone point to me what I could be doing wrong?<br>
&gt;&gt;<br>
&gt;&gt; Regards<br>
&gt;&gt; HASSAN<br>
&gt;&gt;<br>
&gt;&gt;<br>
&gt;&gt;<br>
&gt;&gt; ------------------------------------------------------------------=
------------<br>
&gt;&gt; Register Now for Creativity and Technology (CaT), June 3rd, NYC. C=
aT<br>
&gt;&gt; is a gathering of tech-side developers &amp; brand creativity prof=
essionals.<br>
&gt;&gt; Meet<br>
&gt;&gt; the minds behind Google Creative Lab, Visual Complexity, Processin=
g, &amp;<br>
&gt;&gt; iPhoneDevCamp as they present alongside digital heavyweights like<=
br>
&gt;&gt; Barbarian<br>
&gt;&gt; Group, R/GA, &amp; Big Spaceship. <a href=3D"http://p.sf.net/sfu/c=
reativitycat-com" target=3D"_blank">http://p.sf.net/sfu/creativitycat-com</=
a><br>
&gt;&gt; _______________________________________________<br>
&gt;&gt; Openvpn-users mailing list<br>
&gt;&gt; <a href=3D"mailto:Openvpn-us...@lists.sourceforge.net">Openvpn-use=
r...@lists.sourceforge.net</a><br>
&gt;&gt; <a href=3D"https://lists.sourceforge.net/lists/listinfo/openvpn-us=
ers" target=3D"_blank">https://lists.sourceforge.net/lists/listinfo/openvpn=
-users</a><br>
&gt;&gt;<br>
&gt;<br>
&gt;<br>
&gt; ----------------------------------------------------------------------=
--------<br>
&gt; Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT<b=
r>
&gt; is a gathering of tech-side developers &amp; brand creativity professi=
onals.<br>
&gt; Meet<br>
&gt; the minds behind Google Creative Lab, Visual Complexity, Processing, &=
amp;<br>
&gt; iPhoneDevCamp as they present alongside digital heavyweights like Barb=
arian<br>
&gt; Group, R/GA, &amp; Big Spaceship. <a href=3D"http://p.sf.net/sfu/creat=
ivitycat-com" target=3D"_blank">http://p.sf.net/sfu/creativitycat-com</a><b=
r>
&gt; _______________________________________________<br>
&gt; Openvpn-users mailing list<br>
&gt; <a href=3D"mailto:Openvpn-us...@lists.sourceforge.net">Openvpn-users@l=
ists.sourceforge.net</a><br>
&gt; <a href=3D"https://lists.sourceforge.net/lists/listinfo/openvpn-users"=
 target=3D"_blank">https://lists.sourceforge.net/lists/listinfo/openvpn-use=
rs</a><br>
&gt;<br>
&gt;<br>
<br>
</div></div></blockquote></div><br></div>

--000e0cd1460613faea046b28d8ab--


--===============7016985033437779996==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

------------------------------------------------------------------------------
Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT 
is a gathering of tech-side developers & brand creativity professionals. Meet
the minds behind Google Creative Lab, Visual Complexity, Processing, & 
iPhoneDevCamp as they present alongside digital heavyweights like Barbarian 
Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com 
--===============7016985033437779996==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

_______________________________________________
Openvpn-users mailing list
Openvpn-us...@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/openvpn-users

--===============7016985033437779996==--
Create a group - Google Groups - Google Home - Terms of Service - Privacy Policy
Account Options
