[NANOG] Fiber Cut at 60 Hudson

Does anyone know of any NY fiber cuts going on near/around 60 Hudson Street? I have a Level3 DIA Gig-E that's been out for almost 36 hours and each time I call them I get a different answer on what the problem is and exactly how much longer this is going to take to be resolved. We noticed this go down around 7pm EDT on Sunday and the following morning the dark fiber we have going through 60 Hudson took a hit for about an hour on one side of our DWDM ring...

Anyone know whats up?

Hello,
several months ago we have had a discussion about IPv6 content. There has been a proposal that having some adult content IPv6 only should be a good idea.

I'm not p0rn hoster, but I'm very close to IP content delivery network for Czech public TV. They have news channel (unfortunatelly for most of you in czech language) running round the clock.

So we made available their content over IPv6 and made available TV resolution for IPv6 only. So if you have IPv6, you will get video content at http://master.nacevi.cz/ct24v6.asp in 720x576 (bitrate ~1.5 Mb/s). If you have old IP only, you will see this content only in 320x240 (bitrate ~400 Kb/s).

This service is experimental, and if you have any ideas, complains or questions, please contact me off the list.

                Regards
                    Michal

Hello Michael,

I'm getting the permanent error message:

maxtul@greentheatre ~/temp $ mplayer http://master.nacevi.cz/ct24v6.asp
MPlayer dev-SVN-rUNKNOWN-4.1.2 (C) 2000-2008 MPlayer Team
CPU: Intel(R) Core(TM)2 Duo CPU E4500 @ 2.20GHz (Family: 6, Model: 15, Stepping: 13)
CPUflags: MMX: 1 MMX2: 1 3DNow: 0 3DNow2: 0 SSE: 1 SSE2: 1
Compiled for x86 CPU with extensions: MMX MMX2 SSE SSE2

Playing http://master.nacevi.cz/ct24v6.asp.
Resolving master.nacevi.cz for AF_INET6...
Connecting to server master.nacevi.cz[2a02:1d0:1:0:217:a4ff:feaa:e6f3]: 80...
STREAM_ASF, URL: http://master.nacevi.cz/ct24v6.asp
Resolving master.nacevi.cz for AF_INET6...
Connecting to server master.nacevi.cz[2a02:1d0:1:0:217:a4ff:feaa:e6f3]: 80...
size_confirm mismatch!: 30835 28271
Error while parsing chunk header
Failed, exiting.
Resolving master.nacevi.cz for AF_INET6...
Connecting to server master.nacevi.cz[2a02:1d0:1:0:217:a4ff:feaa:e6f3]: 80...
Cache size set to 320 KBytes
Cache fill: 0.10% (323 bytes)

Exiting... (End of file)

maxtul@greentheatre ~/temp $ links -dump http://master.nacevi.cz/ct24v6.asp
    Ceska televize (c) 2007 Ceska televize (c) 2007

I'm trying from 2a01:d0:9:0:21c:c0ff:fe23:793f.

Michal Krsek wrote:

"Michal Krsek" <michal@krsek.cz> writes:

Hello,
several months ago we have had a discussion about IPv6 content. There has
been a proposal that having some adult content IPv6 only should be a good
idea.

I'm not p0rn hoster, but I'm very close to IP content delivery network for
Czech public TV. They have news channel (unfortunatelly for most of you in
czech language) running round the clock.

So we made available their content over IPv6 and made available TV
resolution for IPv6 only. So if you have IPv6, you will get video content at
http://master.nacevi.cz/ct24v6.asp in 720x576 (bitrate ~1.5 Mb/s). If you
have old IP only, you will see this content only in 320x240 (bitrate ~400
Kb/s).

This service is experimental, and if you have any ideas, complains or
questions, please contact me off the list.

So, we've been native v6 here for a number of years. Figured I'd give
this a shot, only to be greeted with:

   The application VLC quit unexpectedly

   Mac OS X and other applications are not affected.

As long as we're doing forward-looking, brave-new-world stuff, would
it be asking too much to say "use h.264 or something else 'standard'
for the video"?

Best,
                                        ---Rob

Works fine on VLC/OS X for me - but not with flip4mac - flip4mac does IPv4 only it seems.

Michal,

I can see the video here from Brazil.

The quality is very good, but I am using the Go6 Tunnel to watch it.

Sometimes we have some interruption, but in general it is ok.

Thanks,

Giuliano

hello,

works on videolan osx leopard , just a few seconds then it stops , because my connections is not good enough

15:12:25.779523 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: S 613680145:613680145(0) win 65535 <mss 1440,nop,wscale 2,nop,nop,timestamp 603390695 0,sackOK,[|tcp]>
15:12:25.852420 IP6 fe80::20f:66ff:fea7:2d48 > ff02::1:ff79:f1e: ICMP6, neighbor solicitation, who has mini.stattfernsehen.com, length 32
15:12:25.852549 IP6 mini.local > fe80::20f:66ff:fea7:2d48: ICMP6, neighbor advertisement, tgt is mini.stattfernsehen.com, length 32
15:12:25.853012 IP6 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming > mini.stattfernsehen.com.55362: S 963848731:963848731(0) ack 613680146 win 16384 <mss 1440>
15:12:25.853115 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: . ack 1 win 65535
15:12:25.855914 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: P 1:217(216) ack 1 win 65535
15:12:25.932982 IP6 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming > mini.stattfernsehen.com.55362: P 1:145(144) ack 217 win 16864
15:12:25.933132 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: . ack 145 win 65535
15:12:25.933552 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: P 217:329(112) ack 145 win 65535
15:12:26.009816 IP6 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming > mini.stattfernsehen.com.55362: P 145:241(96) ack 329 win 16752
15:12:26.009943 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: . ack 241 win 65535
15:12:26.010238 IP6 mini.stattfernsehen.com.55362 > 2a02:1d0:2::217:a4ff:feaa:e6e7.ms-streaming: P 329:433(104) ack 241 win 65535

first time that i ever saw a ipv6 stream by the way !!!!

thank you very much !!!

Marc

Dear Marc,
if you (or other users) have not enough capacity for watching 1.5 Mb/s stream, you can use lower (comodity) bitrate. You can use comodity URLs:

http://master.nacevi.cz/asx/ct24livewh.asx (400 Kb/s)

http://master.nacevi.cz/asx/ct24livewl.asx (225 Kb/s)

       Regards
             Michal

P.S: Replacing "master" with "master6" will drive you to IPv6 only streams.

P.P.S: Last three RIPE meetings have been broadcasted via IPv6 as well.

Hi,

Robert Blayzor wrote:

Does anyone know of any NY fiber cuts going on near/around 60 Hudson
Street? I have a Level3 DIA Gig-E that's been out for almost 36 hours
and each time I call them I get a different answer on what the problem
is and exactly how much longer this is going to take to be resolved.
We noticed this go down around 7pm EDT on Sunday and the following
morning the dark fiber we have going through 60 Hudson took a hit for
about an hour on one side of our DWDM ring...

Anyone know whats up?

- ---------------------------
Possible bad module / fiber. Techs are in route back to Albany site. ETA
1 hr.

On a different note, outages mailing list (outages@isotf.org) should be
up / running by this week at the latest. We are finishing up on the last
pieces (hw/software).

regards,
/virendra

Dear Marc,
if you (or other users) have not enough capacity for watching 1.5 Mb/s stream, you can use lower (comodity) bitrate. You can use comodity URLs:

http://master.nacevi.cz/asx/ct24livewh.asx (400 Kb/s)

http://master.nacevi.cz/asx/ct24livewl.asx (225 Kb/s)

     Regards
           Michal

exellent Michal

is this multicasted ? what server software you use for ipv6 streaming ?
is there a way to stream via rtp/ rtsp over ipv6 aswell :wink: ?

P.S: Replacing "master" with "master6" will drive you to IPv6 only streams.

P.P.S: Last three RIPE meetings have been broadcasted via IPv6 as well.

cheers

Marc

Hi Marc,

> if you (or other users) have not enough capacity for watching 1.5 Mb/
> s stream, you can use lower (comodity) bitrate. You can use comodity
> URLs:
>
> http://master.nacevi.cz/asx/ct24livewh.asx (400 Kb/s)
>
> http://master.nacevi.cz/asx/ct24livewl.asx (225 Kb/s)

exellent Michal

is this multicasted ?

No it is not. I have no reliable access to mbone and multicast penetration on public Internet here in central europe is "not very wide". So it makes no sense to deal with multicast. Rather I'm investing my time to support IPv6, this looks like it has more perspective :slight_smile:

what server software you use for ipv6 streaming ?

Windows Media Server on top of POS (Picture Operating System - WM 2003 server).

is there a way to stream via rtp/ rtsp over ipv6 aswell :wink: ?

WM is serving data over rtsp as well as over http. ASX file is only pointer to the stream. As I understand the technology, server will negotiate with your client and they try to use ports in following order 1775 (mms) -> 554 (rtsp) -> 80 (http).

                Regards
                    Michal

On Tue, 20 May 2008 14:52:24 +0300
Max Tulyev <president@ukraine.su> mentioned:

Hello Michael,

I'm getting the permanent error message:

Works fine here. You should try different URL. The page you're
requesting contains an actual URL to the video,
http://cdn4.nacevi.cz//CT24-PAL in IPv6 case.

Hello Michael,

I'm getting the permanent error message:
    
Works fine here. You should try different URL. The page you're
requesting contains an actual URL to the video,
http://cdn4.nacevi.cz//CT24-PAL in IPv6 case.

Server name is generated dynamically - depends on your IP/IPv6 address.

Not a rocket science behind :slight_smile:

             Regards
                Michal

I wonder when IPv6porn.com is coming online. We're all waiting on Kevin
Day @ Your.org. The latest mailing list updated was positive [This
morning 5 AM PST8PDT]. Seems DNS has dissapeared for it though. It
should give a decent boost to IPv6 traffic. We're all going to have a
fun time dealing with this once Your.org breaks out IPv6. :slight_smile:

Michal Krsek wrote:

I wonder when IPv6porn.com is coming online. We're all waiting on Kevin
Day @ Your.org.

It honestly is coming soon! :slight_smile: As I mentioned on the mailing list ( http://mail.your.org/pipermail/v6test/2008-May/000065.html ), there are some copyright clearance issues causing a bit of a delay. Once that's resolved, we'll do a bit of testing with those willing to help, then launch.

The latest mailing list updated was positive [This
morning 5 AM PST8PDT]. Seems DNS has dissapeared for it though.

In preparation for launch, all the "real" domains are being moved to their correct IPs. Info about the project itself is still available at http://www.ipv6experiment.com.

It should give a decent boost to IPv6 traffic.

If you want advance warning of a decent amount (many gbps) of v6 traffic being dumped on the world, sign up for the mailing list for the countdown. We've got some major mainstream media exposure going to happen after we're live that should bring a lot of public attention to v6.

We're all going to have a
fun time dealing with this once Your.org breaks out IPv6. :slight_smile:

-- Kevin

FWIW, Level 3 work at 60 Hudson took out Vermont and adjacent areas on the 20th, the same day as you wrote. Quoting their explanation:

---begin quote---
Level 3 experienced a protected outage that impacted customers in Vermont. While a
fiber outage on the working side was being repaired, we experienced a module failure on
the protect side. Traffic had been switched to the protect path.
Root cause: There were 2 separate situations.
1. A working route between Albany and New York failed due to bad fiber jumpers in
at 60 Hudson and this outage was extended due to fibers being moved during
troubleshooting which needed to be re-terminated to their original assignments.
Traffic was switched to the protect path.
2. While the traffic was switched to the protect path, there was also an Optical
Amplifier card failure on an OC-48 riding the protect path through Albany, which
resulted in the outage in South Burlington, VT. Once the Field Technician arrived
with the new OA card and replaced it, traffic restored.
--- end quote ---

I've also heard contradictory information from Level3 reps on some of the above, so I'm not asserting any accuracy for it; so just FYI.

Maybe they finally got to looking into your problem and unplugged the fiber labeled 'Vermont' by accident. :wink:

-Bill

That's about what we experienced. After about 45 hours they finally managed to bring our service back up. :-/ Our problem is that when it did come back up (assuming on the protect side of the ring) our first hop latency to Level3 was through the roof, so we decided to keep our BGP peer shutdown until things completely cleared up. Unfortunately for us I think there is a lot of "legacy" pre-Level3 network between us in Albany and 60 Hudson/111 8th in NYC... a lot of stuff that probably isn't well documented or maintained. (thus, the 40+ hour outage)