cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
3955
Views
0
Helpful
11
Replies

Question about VC4-4c in Cisco ONS15454

Yuyang Jin
Level 1
Level 1

Hi everyone,

I have a question about VC4-4c failure. There is an ethernet circuit created on ONS1545 sized VC4-4c. In the middle of the network, the VC4-4c is separated transmitted as 4x individual VC4. What will happen if one individual VC4 fail? Will the whole circuit be torn down or just suffer loss of part of the bandwidth? Can the behavior be manipulated by configuration?

Thanks,

Henry

3 Accepted Solutions

Accepted Solutions

In that case, the entire VC4-4c circuit will fail. There isn't a recovery mechanism for circuits built using standard concatenation. 

If your Ethernet service cards in the 15454 support Virtual Concatenation (VCAT) and Link Capacity Adjustment Scheme (LCAS), then your Ethernet circuit can remain up even if one or more of the individual VC4 circuits go down.  The Ethernet service cards act to bond together individual VC4s to create a VC4-4v circuit. The VC4s traverse the SDH network as 4 independent circuits.  If one VC4 goes down, the Ethernet service card then adjusts for the loss, and operates as having only a 3 x VC4 trunk.

Many of the latest 15454 Ethernet cards support VCAT & LCAS.

View solution in original post

Michael Dooley
Cisco Employee
Cisco Employee

Tom is correct, but if you are using VCAT and LCAS at the end-points the circuit between will not be a VC4-4c.  It will be 4 individual VC4 circuits that are stiched together by the VCAT/LCAS code.   VCAT will make the 4 individual VC4 circuits look like a pipe the size of a VC4-4c.  And if one VC4 fails, LCAS will take it out the failed VC4 leaving 3 VC4s still carrying traffic between the end-points.

Mike

View solution in original post

Viyuan700 brings up a good point.  Rereading the original question, if the circuit is concatonated, then how could an individual VC4 fail or how could the concatonated VC4 be carried as individual VC4s in the middle of the network (some type of intermediate VC4-4c to VC4-4v mux)?.  Yuyang Jin, do you have more details regarding the network? Maybe you are already deploying VC4-4v circuit between 15454 cards?

Good catch Viyuan700. 

View solution in original post

11 Replies 11

viyuan700
Level 5
Level 5

What will happen if one individual VC4 fail? Will the whole circuit be  torn down or just suffer loss of part of the bandwidth? Can the behavior  be manipulated by configuration?

In SDH, when there is problem with one of the fiber it switches to other. Suppose if there is STM 16 ring then the whole STM 16 switches (all 16 VC4) . 1-2 VC 4 dont fail as SDH/SONET protects the whole circuit not part.

So your circuit will neither be torn down nor suffer any loss.

When you carry and Ethernet circuit over separate VC4 then only the end point need to understand VC4-4c (means concatenation) in between it can be separate VC4.

Hi viyuan700,

How about the occasion when there is no fibre protection? Will the loss of 1x VC4 cause the whole VC4-4c becomes unusable or Is there machnasim in balancing the bandwidth?

Thanks

Henry

In that case, the entire VC4-4c circuit will fail. There isn't a recovery mechanism for circuits built using standard concatenation. 

If your Ethernet service cards in the 15454 support Virtual Concatenation (VCAT) and Link Capacity Adjustment Scheme (LCAS), then your Ethernet circuit can remain up even if one or more of the individual VC4 circuits go down.  The Ethernet service cards act to bond together individual VC4s to create a VC4-4v circuit. The VC4s traverse the SDH network as 4 independent circuits.  If one VC4 goes down, the Ethernet service card then adjusts for the loss, and operates as having only a 3 x VC4 trunk.

Many of the latest 15454 Ethernet cards support VCAT & LCAS.

Thanks, Tom.

Can somebody explain how a standard concatenation can be carried over individual VC4 as asked in first question?

VCAT/LCAS are not bleeding edge technology which are available in latest cards, pizza type STM1 box I used in 2004 have the VCAT/LCAS we are in 2012 now. 

How many times in an carrier network single VC4 fails rarely? LCAS was  more helpful in situation where there was requirement to increase or decrease the bandwidth of the circuit. Customer ask for a 20M circuit today and later he want 30M.

if there in no LCAS then carrier have to create the a different circuit with 30M LCAS can bump the same circuit from 20 to 30M.

Viyuan700 brings up a good point.  Rereading the original question, if the circuit is concatonated, then how could an individual VC4 fail or how could the concatonated VC4 be carried as individual VC4s in the middle of the network (some type of intermediate VC4-4c to VC4-4v mux)?.  Yuyang Jin, do you have more details regarding the network? Maybe you are already deploying VC4-4v circuit between 15454 cards?

Good catch Viyuan700. 

The topology is we lease STM-16 from 3rd party. VC4#1 - VC4#12 is individual VC4; the last four VC4 is VC4-4c. The ethernet circuit is created as VC4-4c on the ONS15454 ADM10G card.

From the discussion, I realized it must be VC4-4c all along the path inside 3rd party capacity as a whole, it cannot be spearated into individual VC4s for transport in 3rd party network (correct me if I am wrong).

Perhaps another option is configure VC4-4v circuit on ADM10G card using VC4#1 - #4, which was transported as VC4s inside 3rd party network.

VC4-4c is made of of 4 VC4 like Vc4-4v. Only difference in case of VC4-4c all VC4 have to be contiguous (like you have from 13-16). If it is Vc4-4v it can be any 4 out of 16VC4 . Don,t have to be contiguous as in earlier case. Vc4-4v is useful for case where you dont have contiguous VC4.

You can create Vc4-4v and they can go as single VC4 but most probably they all be following the same path. If they are going through same path and chances of one VC4 failure is rare. As error in one VC4 will chnage the error in STM 16 frame and it will switch to redundant path. So LCAS will not be used in this case.

LCAS will be used in such cases but VC4 in such cases are travelling through different physical route. SO if one route fails still you have other 3. LCAS is useful if you want to add or remove Vc4 from exisitng bandwidth.

Most probabaly router will not know the decrease in bandwidth but if the traffic is TCP based then will reduce the transfer rate to handle dropping of packet.

However with the VCAT and LCAS, when there is one individual VC4 fails, it is still a problem to have a feedback mechanism to let the end-router realise the bandwidth change. I can make the router be bandwidth aware by configuring MPLS TE using RSVP. But once the bandwidth adjustsed on ONS15454 (SDH Ethernet circuit), is there any way to let the router aware of the bandwidth decrease? Otherwise congestion will be incurred and end-user will suffer the packet loss.

Michael Dooley
Cisco Employee
Cisco Employee

Tom is correct, but if you are using VCAT and LCAS at the end-points the circuit between will not be a VC4-4c.  It will be 4 individual VC4 circuits that are stiched together by the VCAT/LCAS code.   VCAT will make the 4 individual VC4 circuits look like a pipe the size of a VC4-4c.  And if one VC4 fails, LCAS will take it out the failed VC4 leaving 3 VC4s still carrying traffic between the end-points.

Mike

Thanks, Mike

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Innovations in Cisco Full Stack Observability - A new webinar from Cisco