[c-nsp] huge amount of mcast traffic
james list
jameslist72 at gmail.com
Mon Oct 17 13:08:05 EDT 2016
Hi Pavel
obviously not, as you see from previous email market is connected to an
ASR. then there is a C6807 and the only customer currently requesting this
huge mcast feed is connected to C6500 with a WAN link...
I've attached my hw configuration, facing my cisco knowledge it shouldn't
be an issue, but I was asking for hints...
Matt has pretty confirmed it should work, but he is not sure about the
overall quality if I get it well, if you have another view please let me
know.
Cheers
James
2016-10-17 13:02 GMT+02:00 Pavel Skovajsa <pavel.skovajsa at gmail.com>:
> James,
> So all your customers are on 6708?
>
> Why thing you can try is check the internal architecture of the 6708 cards
> especially the egress replication asic.Probably also depends on which ports
> you have the customers connected...
>
> -pavel
>
> Dňa 13.10.2016 18:44 používateľ "Matthew Huff" <mhuff at ox.com> napísal:
>
>> A sustained 6Gps on a 10GB pipe is hard to do already, but with
>> multicast…. Typically that large of multicast is broken up into different
>> multicast addresses can be split on multiple lines. The burst nature of the
>> feed is going to be an issue. Will it work, yes. Will it work well, I doubt
>> it.
>>
>> ----
>> Matthew Huff | 1 Manhattanville Rd
>> Director of Operations | Purchase, NY 10577
>> OTA Management LLC | Phone: 914-460-4039
>> aim: matthewbhuff | Fax: 914-694-5669
>>
>> From: james list [mailto:jameslist72 at gmail.com]
>> Sent: Thursday, October 13, 2016 12:34 PM
>> To: Matthew Huff <mhuff at ox.com>
>> Cc: cisco-nsp at puck.nether.net
>> Subject: Re: [c-nsp] huge amount of mcast traffic
>>
>> well we'll connect to 10 Gbs interface a traffic up to 6 Gbs, not on 6748
>> 1 Gbs blades... no other issue you see ?
>>
>> 2016-10-13 18:31 GMT+02:00 Matthew Huff <mhuff at ox.com<mailto:mhuff at ox.com
>> >>:
>> The 6748 blades are going to be an issue with buffer overruns. Whether
>> this will be a minor or major issue depends on the application that uses
>> the multicast data.
>>
>> ----
>> Matthew Huff | 1 Manhattanville Rd
>> Director of Operations | Purchase, NY 10577
>> OTA Management LLC | Phone: 914-460-4039<tel:914-460-4039>
>> aim: matthewbhuff | Fax: 914-694-5669<tel:914-694-5669>
>>
>> From: james list [mailto:jameslist72 at gmail.com<mailto:
>> jameslist72 at gmail.com>]
>> Sent: Thursday, October 13, 2016 12:25 PM
>> To: Matthew Huff <mhuff at ox.com<mailto:mhuff at ox.com>>
>> Cc: cisco-nsp at puck.nether.net<mailto:cisco-nsp at puck.nether.net>
>> Subject: Re: [c-nsp] huge amount of mcast traffic
>>
>>
>> Hi
>>
>>
>>
>> I’m not able to find the multicast replication mode on ASR..
>>
>>
>>
>> On core routers:
>>
>>
>>
>> C6807 has Supervisor Engine 2T 10GE and IOS 15.1(2)SY4
>>
>>
>>
>> xxx>sh module
>> Mod Ports Card Type Model
>> Serial No.
>> --- ----- -------------------------------------- ------------------
>> -----------
>> 1 20 DCEF2T 4 port 40GE / 16 port 10GE WS-X6904-40G xx
>> 2 20 DCEF2T 4 port 40GE / 16 port 10GE WS-X6904-40G xx
>> 3 5 Supervisor Engine 2T 10GE w/ CTS (Acti VS-SUP2T-10G xx
>> 5 48 CEF720 48 port 10/100/1000mb Ethernet WS-X6848-GE-TX xx
>>
>>
>>
>> Mod Sub-Module Model Serial Hw
>> Status
>>
>>
>>
>> ---- --------------------------- ------------------ ----------- -------
>> -------
>>
>>
>>
>> 1 Distributed Forwarding Card WS-F6K-DFC4-E xxx 1.2 Ok
>>
>>
>>
>> 2 Distributed Forwarding Card WS-F6K-DFC4-E xxx 1.2 Ok
>>
>>
>>
>> 3 Policy Feature Card 4 VS-F6K-PFC4 xxx 3.0 Ok
>>
>>
>>
>> 3 CPU Daughterboard VS-F6K-MSFC5 xxx 3.0 Ok
>>
>>
>>
>> 5 Distributed Forwarding Card WS-F6K-DFC4-A xxx 1.4 Ok
>>
>>
>>
>> xxx#sh platform multicast routing replication
>>
>>
>>
>> Current mode of replication is Egress
>>
>>
>>
>> Configured mode of replication is Egress
>>
>>
>>
>>
>>
>> Switch Slot Multicast replication capability
>>
>>
>>
>> 1 1 Egress
>>
>>
>>
>> 1 2 Egress
>>
>>
>>
>> 1 3 Egress
>>
>>
>>
>> 1 5 Egress
>>
>>
>>
>> 2 1 Egress
>>
>>
>>
>> 2 2 Egress
>>
>>
>>
>> 2 3 Egress
>>
>>
>>
>> 2 5 Egress
>>
>>
>>
>> 4 1 Ingress
>>
>>
>>
>> 3 1 Ingress
>>
>>
>>
>> 5 1 Ingress
>>
>>
>>
>>
>>
>> C6500 has Supervisor Engine 720 10GE and IOS 12.2(33)SXI5
>>
>>
>>
>>
>>
>> xxx>sh module
>>
>>
>>
>> Mod Ports Card Type Model
>> Serial No.
>>
>>
>>
>> --- ----- -------------------------------------- ------------------
>> -----------
>>
>>
>>
>> 1 48 CEF720 48 port 10/100/1000mb Ethernet WS-X6748-GE-TX xxx
>>
>>
>>
>> 2 8 CEF720 8 port 10GE with DFC WS-X6708-10GE xxx
>>
>>
>>
>> 3 48 CEF720 48 port 10/100/1000mb Ethernet WS-X6748-GE-TX xxx
>>
>>
>>
>> 4 48 CEF720 48 port 1000mb SFP WS-X6748-SFP xxx
>>
>>
>>
>> 5 5 Supervisor Engine 720 10GE (Active) VS-S720-10G xxx
>>
>>
>>
>>
>>
>> Mod Sub-Module Model Serial Hw
>> Status
>>
>>
>>
>> ---- --------------------------- ------------------ ----------- -------
>> -------
>>
>>
>>
>> 1 Distributed Forwarding Card WS-F6700-DFC3C xxx 1.6 Ok
>>
>>
>>
>> 2 Distributed Forwarding Card WS-F6700-DFC3C xxx 1.8 Ok
>>
>>
>>
>> 3 Distributed Forwarding Card WS-F6700-DFC3C xxx 1.6 Ok
>>
>>
>>
>> 4 Centralized Forwarding Card WS-F6700-CFC xxx 4.2 Ok
>>
>>
>>
>> 5 Policy Feature Card 3 VS-F6K-PFC3C xxx 1.1 Ok
>>
>>
>>
>> 5 MSFC3 Daughterboard VS-F6K-MSFC3 xxx 1.0 Ok
>>
>>
>>
>>
>>
>> xxx>show mls ip multicast capability
>>
>>
>>
>> Current mode of replication is Egress
>>
>>
>>
>> Configured replication mode is Auto
>>
>>
>>
>>
>>
>> Slot Multicast replication capability
>>
>>
>>
>> 1 Egress
>>
>>
>>
>> 2 Egress
>>
>>
>>
>> 3 Egress
>>
>>
>>
>> 4 Egress
>>
>>
>>
>> 5 Egress
>>
>>
>>
>> Cheers
>>
>> 2016-10-13 17:59 GMT+02:00 Matthew Huff <mhuff at ox.com<mailto:mhuff at ox.com
>> >>:
>> Even with fabric enable blades in the c6500, you are going to get massive
>> output buffer overflows. Market data has very uneven traffic patterns
>> causing microburst effects. What sup-engines/blades are on the boxes? What
>> type of multicast replication is being used (ingress/egress). QoS policies
>> typically make matters worse. What type of interfaces are on the 6500?
>>
>>
>> ----
>> Matthew Huff | 1 Manhattanville Rd
>> Director of Operations | Purchase, NY 10577
>> OTA Management LLC | Phone: 914-460-4039<tel:914-460-4039>
>> aim: matthewbhuff | Fax: 914-694-5669<tel:914-694-5669>
>>
>> > -----Original Message-----
>> > From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net<mailto:
>> cisco-nsp-bounces at puck.nether.net>] On Behalf Of
>> > james list
>> > Sent: Thursday, October 13, 2016 10:45 AM
>> > To: cisco-nsp at puck.nether.net<mailto:cisco-nsp at puck.nether.net>
>> > Subject: [c-nsp] huge amount of mcast traffic
>> >
>> > Dear experts,
>> >
>> > I’ve a multicast financial market connected to my infrastructure, I’ve
>> > been
>> > informed that a new data multicast flow could reach up to 6 Gbs, so an
>> > huge
>> > amount of traffic needs to be replicated.
>> >
>> > Market is connected to an ASR 1001, than to a C6807-XL and customers
>> > are
>> > connected to C6500.
>> >
>> > ASR1001 is running 15.3(3)S1 and currently has a license for 2.5Gbs (to
>> > be
>> > upgrade)
>> >
>> > C6807 has Supervisor Engine 2T 10GE and IOS 15.1(2)SY4
>> >
>> > C6500 has Supervisor Engine 720 10GE and IOS 12.2(33)SXI5
>> >
>> > I’d like to understand in your experience if the mentioned
>> > infrastructure
>> > could suffer in performance or throughput or other, having to replicate
>> > the
>> > mentioned amount of traffic.
>> >
>> > Thanks in advance for any feedback
>> >
>> > Cheers
>> >
>> > James
>> > _______________________________________________
>> > cisco-nsp mailing list cisco-nsp at puck.nether.net<mailto:
>> cisco-nsp at puck.nether.net>
>> > https://puck.nether.net/mailman/listinfo/cisco-nsp
>> > archive at http://puck.nether.net/pipermail/cisco-nsp/
>>
>>
>> _______________________________________________
>> cisco-nsp mailing list cisco-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>
>
More information about the cisco-nsp
mailing list