From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-ie0-x22d.google.com (mail-ie0-x22d.google.com [IPv6:2607:f8b0:4001:c03::22d]) (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 14CE321F1B1 for ; Mon, 1 Jul 2013 10:50:32 -0700 (PDT) Received: by mail-ie0-f173.google.com with SMTP id k13so9739138iea.4 for ; Mon, 01 Jul 2013 10:50:30 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; bh=tGKLnUh6lfMYznLDcPfAVAaLwUT9YvQppdM4KD2bvI8=; b=HERMps9ujAwUwEzKMT4qOA7uhqit2lpQMDv3fCVo7nejpFckWIkZbnTTDJ9XXTqbeS g6eYhTYVi9xzZtZR+qTInQK9LeQuUmKmIYiqVx62qmcuZUpwe5Wr+jJM7/fmUwdSd6Dm 418jyKEP/oyklozwjfpal1idW35tRS7l+Cyj17pemPSGuziSbjgGzKjE7is8VwNvYHYk RC4Kgp58q8Q9Y2XEesB95Fh86XOy2F81uoOuQ1Nd87LzcLBrdHRb3ULDE11OX6OcY4ob mRFQ/KP0aWalrHTGYSgoe/0us66iOylw11R7SAVlEb6vVzTipLdDzitqsIzadv8ZxayF tjYw== MIME-Version: 1.0 X-Received: by 10.50.65.42 with SMTP id u10mr16198885igs.47.1372701030691; Mon, 01 Jul 2013 10:50:30 -0700 (PDT) Received: by 10.64.249.37 with HTTP; Mon, 1 Jul 2013 10:50:30 -0700 (PDT) In-Reply-To: References: Date: Mon, 1 Jul 2013 12:50:30 -0500 Message-ID: From: jason arends To: Cerowrt-devel@lists.bufferbloat.net Content-Type: multipart/alternative; boundary=047d7b414330513b8e04e076dd25 Subject: Re: [Cerowrt-devel] DLNA between wired and wireless again X-BeenThere: cerowrt-devel@lists.bufferbloat.net X-Mailman-Version: 2.1.13 Precedence: list List-Id: Development issues regarding the cerowrt test router project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 01 Jul 2013 17:50:32 -0000 --047d7b414330513b8e04e076dd25 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Would this ipmr part in the 902-debloat_proc.patch script be messing with us? --- a/net/ipv4/ipmr.c +++ b/net/ipv4/ipmr.c @@ -70,6 +70,10 @@ #define CONFIG_IP_PIMSM 1 #endif +#ifdef CONFIG_PROC_STRIPPED +#undef CONFIG_PROC_FS +#endif + struct mr_table { struct list_head list; #ifdef CONFIG_NET_NS Unfortunately I am not sure how to test this or recompile without that part, I'm just browsing on github. Oh, and I just found this while trying to understand what it does also: https://dev.openwrt.org/changeset/36529/trunk where it looks like they took that part out. I notice in my /proc/net I have ip6_mr* but no ip_mr stuff ~jason On Fri, Jun 28, 2013 at 3:07 PM, jason arends wrote= : > After realizing you can't put numeric ip's in replies... here's my > reformatted reply. I removed the middle octets in some places, but I hop= e > it's still obvious. Let me know if there's a better place to discuss thi= s. > > I've done a few things based on some googling... I'm no expert, just usin= g > this as my home router. Here's what i can remember changing so far: > > To make sure the packets with TTL of 1 don't die going through the router= : > > iptables -t mangle -A PREROUTING -i se00 -d 224 dot 0 dot 0 dot 0/4 -j TT= L --ttl-inc 1 > iptables -t mangle -A PREROUTING -i sw00 -d (224..0/4) -j TTL --ttl-inc 1 > iptables -t mangle -A PREROUTING -i sw10 -d (224..0/4) -j TTL --ttl-inc 1 > > some websites said you need a route for multicast so I did: > > ip route add to 224 dot 0 dot 0 dot 0/4 dev se00 > > I've tried adding the same for sw10 and sw00 but didnt' seem to help, but > I'm not sure about putting just the se00 in there. The thing about this = is > when i do: > > ip mroute show > > it's blank, but I'm not sure if that should show something. I also get > this, which I'm not sure is good or bad, but someone may be able to > interpret: > > #ip route get 224 dot 0 dot 0 dot 1 from 172 dot 30 dot 42 dot 70 iif sw0= 0 > multicast 224..1 from 172..70 dev lo > cache iif * > > From http://www.policyrouting.org/iproute2.doc.html it has an example > where the output looks like this though with the Oifs showing the other > device and pimreg, so I'm not sure pim is working right: > cache iif eth0 Oifs eth1 pimreg > > I've setup my pimd.conf like this (comments removed and middle octets > removed, hopefully obviously): > > default_source_preference 101 > default_source_metric 1024 > > phyint ge00 disable > phyint gw00 disable > phyint gw10 disable > > cand_rp 172..1 time 60 priority 20 > cand_bootstrap_router 172..1 priority 5 > rp_address 172..1 224..0 masklen 4 priority 5 > group_prefix 224..0 masklen 4 > switch_data_threshold rate 50000 interval 20 # 50kbps (approx.= ) > switch_register_threshold rate 50000 interval 20 # 50kbps (approx.= ) > > I did also install miniDLNA and setup the router to stream media from a u= sb drive, both wired and wireless can see and stream from that just fine, b= ut not sure if that has any impact on the issue. > > Right now if I open a dlna application on wireless, it can see my wired c= omputer. They appear to exchange SSDP search and notify packets, but when = they try to view contents of the media on the wired computer or the wired c= omputer tries to discover other devices on wireless, it fails. If I turn o= n the wifi card in the wired computer as well, all the devices instantly sh= ow up and everything works, so that makes me think the computer itself is f= ine... unless there's something that tells it to ignore or refuse devices o= n other subnets. > > > ~jason > > > On Fri, Jun 28, 2013 at 1:56 PM, Maciej Soltysiak w= rote: > >> On Fri, Jun 28, 2013 at 5:16 PM, jason arends wr= ote: >> >>> Saw this post >>> https://lists.bufferbloat.net/pipermail/cerowrt-devel/2013-January/0009= 24.html which >>> said it was working in 3.7.2-4, but I haven't been able to get this wor= king >>> right for me on 3.7.5-2. Did something change that broke this or have = I >>> misconfigured something? After some fiddling with pimd, route, miniupn= pd, >>> minissdpd, etc, I have ended up where the client on wireless (Xbox360 >>> console) can see my wired computer (Win 8) but when it tries to open it= , it >>> can't browse the contents. This works when both are on the same wirele= ss, >>> but I get some buffering/lag in video playback >>> >> I had the same problems. I was then able to hint Dave where the issue >> might be and it really was working out of the box for me (Wired Samsung = TV, >> win7 laptop on 802.11g and n). >> Unfortunately since then I was never able to figure out what's going on. >> Are you able to share what you did to improve your situation? >> >> In wireshark, I see the SSDP search from the Xbox and then when the Xbox >>> tries to access it, the computer replying to the console with a Server >>> Error 500 containing "Access Denied" and I think it's because the compu= ter >>> can't see the console. I can ping it though. When I put the computer = on >>> the same wireless as the console, then open the Network folder and hit >>> refresh, the Xbox pops up along with other things on wireless (Roku, et= c) >>> but when the computer is wired, it doesn't see any of those, so I think >>> something about the SSDP packets isn't working quite right between subn= ets >>> or with pimd. My guess is the SSDP search from the computer isn't gett= ing >>> to the wifi, only vice versa. >>> >> You're in a better situation than I am in. To me, Error 500 suggests the >> issue is located outside minissdpd or the router. >> >> >> >>> >>> Not sure where to go next, any ideas? (is this the right place to >>> ask/troubleshoot this?) >>> >> I don't know a better one. >> >> Maciej >> > > --047d7b414330513b8e04e076dd25 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable
Would this ipmr part in the 902-debloat_proc.patch script = be messing with us?

--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -70,6 +70,10 @@
=A0#define CO= NFIG_IP_PIMSM 1
=A0#endif
=A0
+#ifdef CONFIG_PROC_STRIPPED
+#undef CONFIG_PROC_FS
+#endif
+
=A0struct= mr_table {
=A0 struct list_head list;
=A0#ifdef CONFIG_NET_NS

Unfortunately I am not sure how to test this or reco= mpile without that part, I'm just browsing on github. =A0Oh, and I just= found this while trying to understand what it does also:=A0https://dev.openwrt.org/changese= t/36529/trunk=A0where it looks like they took that part out. =A0I notic= e in my /proc/net I have ip6_mr* but no ip_mr stuff

~jason


On Fri, Jun 28, 2013 at 3:07 PM, jason a= rends <jason.arends@gmail.com> wrote:
After realizing you can't put numeric ip's in replies... her= e's my reformatted reply. =A0I removed the middle octets in some places= , but I hope it's still obvious. =A0Let me know if there's a better= place to discuss this.

I've= done a few things based on some googling... I'm no expert, just using = this as my home router. =A0Here's what i can remember changing so far:<= /span>

To mak= e sure the packets with TTL of 1 don't die going through the router:
iptables -t mangle -A PREROUTING -i se00 -d 224 dot 0 dot 0 dot 0/4 -j TTL =
--ttl-inc 1
iptables -t mangle -A PREROUTING -i sw00 -d (224..0/4) -j TTL --ttl-inc 1
iptables -t mangle -A PREROUTING -i sw10 -d (224..0/4) -j TTL --ttl-inc 1
some websites said you need a route for multicas=
t so I did:
ip route ad=
d to 224 dot 0 dot 0 dot 0/4 dev se00



I've tried adding the same for sw10 and sw=
00 but didnt' seem to help, but I'm not sure about putting just the=
 se00 in there. =A0The thing about this is when i do:
ip mroute show
it'= s blank, but I'm not sure if that should show something. =A0I also get = this, which I'm not sure is good or bad, but someone may be able to int= erpret: =A0
#ip route get 224 dot 0 dot=
 0 dot 1 from 172 dot 30 dot 42 dot 70 iif sw00
multicast 224..1 from 172..70 dev lo=20
    cache <local,mc>  iif *
From= =A0http://www.policyrouting.org/iproute2.doc.html=A0it has an example= where the output looks like this though with the Oifs showing the other de= vice and pimreg, so I'm not sure pim is working right:=A0
cache <mc> iif eth0 Oifs eth1 pi= mreg

I've setup my pimd.conf like this (comments remov= ed and middle octets removed, hopefully obviously):
default_source_preference       101    =
              =20
default_source_metric           1024 
phyint ge00 disable                                                    =
 =20
phyint gw00 disable                                                     =20
phyint gw10 disable                                                     =20
                                                                           =
                         =20
cand_rp 172..1 time 60 priority 20                                 =20
cand_bootstrap_router 172..1 priority 5                               =20
rp_address 172..1 224..0 masklen 4 priority 5                  =20
group_prefix 224..0 masklen 4                                              =
                      =20
switch_data_threshold           rate 50000 interval 20  # 50kbps (approx.)
switch_register_threshold       rate 50000 interval 20  # 50kbps (approx.)<=
/pre>
I did also install miniDLNA and setup the router to=
 stream media from a usb drive, both wired and wireless can see and stream =
from that just fine, but not sure if that has any impact on the issue.




Right now if I open a dlna application on wireless, it can=
 see my wired computer. =A0They appear to exchange SSDP search and notify p=
ackets, but when they try to view contents of the media on the wired comput=
er or the wired computer tries to discover other devices on wireless, it fa=
ils. =A0If I turn on the wifi card in the wired computer as well, all the d=
evices instantly show up and everything works, so that makes me think the c=
omputer itself is fine... unless there's something that tells it to ign=
ore or refuse devices on other subnets.

~jason


On Fri, = Jun 28, 2013 at 1:56 PM, Maciej Soltysiak <maciej@soltysiak.com>= wrote:
On Fri, Jun 28, 2013 at 5:16 PM, jason aren= ds <jason.arends@gmail.com> wrote:
Saw this post=A0https://= lists.bufferbloat.net/pipermail/cerowrt-devel/2013-January/000924.html= =A0which said it was working in 3.7.2-4, but I haven't been able to get= this working right for me on 3.7.5-2. =A0Did something change that broke t= his or have I misconfigured something? =A0After some fiddling with pimd, ro= ute, miniupnpd, minissdpd, etc, I have ended up where the client on wireles= s (Xbox360 console) can see my wired computer (Win 8) but when it tries to = open it, it can't browse the contents. =A0This works when both are on t= he same wireless, but I get some buffering/lag in video playback
I had the same problems. I was then able to hint Da= ve where the issue might be and it really was working out of the box for me= (Wired Samsung TV, win7 laptop on 802.11g and n).
Unfortunately since t= hen I was never able to figure out what's going on. Are you able to sha= re what you did to improve your situation?

In wireshark, I see t= he SSDP search from the Xbox and then when the Xbox tries to access it, the= computer replying to the console with a Server Error 500 containing "= Access Denied" and I think it's because the computer can't see= the console. =A0I can ping it though. =A0When I put the computer on the sa= me wireless as the console, then open the Network folder and hit refresh, t= he Xbox pops up along with other things on wireless (Roku, etc) but when th= e computer is wired, it doesn't see any of those, so I think something = about the SSDP packets isn't working quite right between subnets or wit= h pimd. =A0My guess is the SSDP search from the computer isn't getting = to the wifi, only vice versa.
You're in a better situation t= han I am in. To me, Error 500 suggests the issue is located outside minissd= pd or the router.

=A0

Not sure where to go next, any ideas? =A0(is this the right place to = ask/troubleshoot this?)
I don't know= a better one.

Maciej


--047d7b414330513b8e04e076dd25--