PDA

View Full Version : DNS/TRACERT Naming explanation?


Rik
29-12-2008, 07:22
Is there any logic to the VM DNS Tracert naming, can someone give me a dummies guide to understanding what im looking at?

Is it or all top secret and thats strictly agaisnt the rules? ;)

Thanks

For instance.

Tracing route to nthellworld.co.uk [87.106.245.143]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms unknown [192.168.1.1]
2 6 ms 6 ms 8 ms 10.xxx.220.1
3 6 ms 8 ms 7 ms lutn-t2cam1-b-v108.network.virginmedia.net [80.4.118.221]
4 8 ms 7 ms 7 ms lutn-t3core-1b-ae1-0.network.virginmedia.net [195.182.174.189]
5 10 ms 11 ms 8 ms nth-bb-b-so-120-0.network.virginmedia.net [213.105.175.153]
6 13 ms 12 ms 14 ms tele-ic-1-as0-0.network.virginmedia.net [62.253.184.2]

I know the locations for instance, lutn and nth are Luton and Northampton (i assume) and tele, is that Telehouse in London?
Im more interested in the cam, core and numbers afterwards and their relevance.

Yes im a geek

:)

Axegrinder
29-12-2008, 09:21
t2cam is the "Cable Access Module I think. I'm not sure about the rest.

Ignitionnet
29-12-2008, 09:33
3 6 ms 8 ms 7 ms lutn-t2cam1-b-v108.network.virginmedia.net [80.4.118.221]
4 8 ms 7 ms 7 ms lutn-t3core-1b-ae1-0.network.virginmedia.net [195.182.174.189]
5 10 ms 11 ms 8 ms nth-bb-b-so-120-0.network.virginmedia.net [213.105.175.153]
6 13 ms 12 ms 14 ms tele-ic-1-as0-0.network.virginmedia.net [62.253.184.2]

luton - type 2 cable access module cluster 1 - b (second in cluster, 1st is a) - vlan 108
luton - type 3 core cluster 1 - b - aggregated ethernet 1-0
northampton - backbone - b - serial interface 1-2-0-0
telehouse east - interconnect - 1 - aggregated serial interface 0-0

Have a nice day!

cybernetic_tiger
29-12-2008, 09:44
The network is layered with Tier2 (T2) being closer to the customer and Tier3 (T3) being the interconnect between the local network and the backbone core (BB).

The 1's, 2's, a's and b's all denote the number of the device as there are usually between two and four of each for load balancing and resilience.

The last portion before the "network.virginmedia.net" usually denotes the interface type Vlan (V), Routing AS (as), OSPR ABR (ae), Interconnect (ic).

The new naming convention is not finished as yet so things are still being finalised across the core.

Ignitionnet
29-12-2008, 09:51
The network is layered with Tier2 (T2) being closer to the customer and Tier3 (T3) being the interconnect between the local network and the backbone core (BB).

The 1's, 2's, a's and b's all denote the number of the device as there are usually between two and four of each for load balancing and resilience.

The last portion before the "network.virginmedia.net" usually denotes the interface type Vlan (V), Routing AS (as), OSPR ABR (ae), Interconnect (ic).

The new naming convention is not finished as yet so things are still being finalised across the core.

No sir!

T2 = type 2, the 2nd version of the cam/core, T3 = type 3, 3rd evolution. T3 core = type 3 = 3rd version of the hardware. The t2cores are Cisco hardware the t3core is a more powerful Juniper chassis. The t2cam is an upgraded version of the t1cam which dates back to when the cable modem service was backhauled over ATM.

Units closer to the customer are suffixed *am meaning 'access module' - these are in the case of the cable network where the CMTS connects, in the case of leased lines they connect directly to these. Core is the next tier of the network, and backbone is the supercore which aggregates and transports traffic from all core nodes and the interconnects, which are used to connect to LINX and/or other ISPs.

AE is not OSPF, OSPF does not run on the VM network past core level, AE is aggregated ethernet as per http://www.juniper.net/techpubs/software/junos/junos56/swconfig56-interfaces/html/interfaces-ethernet-config34.html The VM core is an IBGP / MPLS overlay, with IS-IS used as the IGP to redistribute loopbacks from which the IBGP messages source.

As above AS is aggregated serial interface, similar to the above but using SDH framing rather than gigabit / 10-GigE framing on the router port.

Nothing in the DNS is relevant to routing, they are strictly the names of the physical ports.

Also the 2's are not resilient with the 1's - the resilience is arranged 1a and b, 2a and b, the resilience is only between a and b nodes of the same number. Where there are multiply resilient nodes in a cluster they would be a, b, c, etc.

Rik
29-12-2008, 11:21
I love a bit of Networking Speak in the morning!!

Thanks guys!! :)

Joking aside thats exactly the kind of indepth answer I actually wanted :)

Thankyou

ccarmock
03-01-2009, 13:49
Are the 'osr' entries the 'headend then? Also I've seen 'tenge' in a lot of traces.

1 1 ms 1 ms 1 ms [192.168.10.1]
2 10 ms 8 ms 16 ms [10.89.0.1]
3 9 ms 12 ms 9 ms nmal-geam-1-v11.network.virginmedia.net [62.30.1
12.131]
4 10 ms 12 ms 11 ms osr01croy-tenge72.network.virginmedia.net [62.30
.242.37]
5 12 ms 12 ms 18 ms bre-bb-a-ge-300-0.network.virginmedia.net [195.1
82.178.82]

cybernetic_tiger
03-01-2009, 20:24
OSR is Optical Switching Router and is at the Head End. Tenge is just showing that the interface is a TenGigabitEthernet port (10Gbps).

Ignitionnet
04-01-2009, 09:24
New Malden is a bit of a weird one in that it has a Telewest office there as well, and uses GEAMs / Gigabit Ethernet Access Modules.

Most Telewest hubs have OSRs locally:

3 11 ms 9 ms 12 ms osr01mort-v15.network.virginmedia.net [62.30.242.129]
4 11 ms 22 ms 13 ms osr02croy-tenge73.network.virginmedia.net [62.30.242.57]
5 14 ms 10 ms 11 ms pop-bb-b-ge-210-0.network.virginmedia.net [195.182.178.86]

With Cisco GSRs dealing with supercore duties. These have become less and less relevant as they have been replaced with the ex-ntl Juniper supercore.

Bonglet
05-01-2009, 02:08
Can anyone explain how if i do a traceroute then i do a reverse traceroute there are 2 extra hops on the ubr on a reverse traceroute as there are in an outgoing traceroute?
Always seemed funny to me why its taking a longer path to come in than go out :S.

Ignitionnet
05-01-2009, 10:04
Routing is not symettrical basically. It's also possible that some hops are hidden via MPLS in one direction while they are present in another.

tfewster
22-01-2009, 23:59
Hi,

try playing with some of the traceroute and DNS tools on

www.network-tools.in (http://www.network-tools.in)

they have remote and local tracert options so i is good to be able to compare both to see where any hold ups might be.

Tim