Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Mike Lewis on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Passport 15K - New to ATM want to remap PVC so that it is redundant.

Status
Not open for further replies.

rainman

ISP
Mar 22, 2001
186
US
Hello,

About two years ago I was introduced to the Nortel Passport series of ATM switches (7440/7480/15000). Currently on these ATM switches I have many static PVC's configured as NRP's (nailed up relay points). I am trying to dig up some information on how to setup a PVC so that if it fails then to take an alternate path. I have done some reading and found that possibly a SVC (switched VC's) or something else referred to as "soft PVC's".

Again, I am rather new to ATM, and any help/thoughts/URL's would be appreciated.


Tx - Rainman
 
A soft PVC (SPVC) will accomplish what I think you are looking for; with a soft PVC you nail up each endpoint, but the connection itself is set up dynamically so in the case of a link failure the circuit can automatically reroute (but otherwise it stays up all the time unlike an SVC which is torn down after an idle period).
 
Does anybody here know how to get a Passport ATM switch to do SPVC's? What about a message board that specializes in Nortel ATM configurations?


-Rainman
 
cprx7 is right - 15K's are PNNI capable. Rainman, if you're not already using PNNI in your network, it is fairly easy to configure a simple circuit. Like other dynamic routing technologies such as OSPF, there is a lot of fancy functions that you can do with PNNI - my advice is to keep it as simple as you can. PNNI addressing is similar to IP addressing in that if you try to build in *too much* room for scalability and all the functionality that PNNI provides, it will become horrendous to manage. Of course if you don't build in enough, you're just as screwed, so... tread carefully.
Your second post to this thread, Rainman, made me wonder if you have access to the NTPs? Although they are painful to read, the ATM guides are fairly comprehensive for at least the basic ATM configurations. Start with 241-2701-710 "Passport 7400/15000/20000 ATM Configuration Guide"
It would be funny to see a message board that specializes in Passport configs - have you ever saved a config (view) as ascii and looked at it? One that I just looked at from my network is 140 pages long! Obviously the size varies depending on the cards/services you use, but Tek-Tips may have to get another server just for us if we do start posting Passport configs! ;-)
 
Hi,

Thanks for the reply. I do have access to the NMS servers however no documentation. Somehow the documentation that Nortel provided us for our 7440/7480/15k/s was lost, then I came into the network and had to pickup ATM and the Passport from scratch. I think a message board dedicated to Passport configurations would be pretty helpful, as it seems to be a topic that is not often talked about.

I will look into PNNI's - I have been studying ATM for awhile now, and I know that PNNI's are private network to network interfaces, but I just need to relate this to Cisco networking which I am very familiar with, and then how to implement it. Currently our ATM switches use just NRP PVC's with no SPVC's. I can post a sample config of how I'd setup our current PVC's / nexthops to get input from some ppl?


Tx - Rainman!
 
Hello again,

Here's how I currently have say a PVC setup that I'd like to setup some redundancy for:

a atmif/10+11 vcc/6.708+6.708 nrp
set atmif/10 vcc/6.708 nrp nexthop atmif/11 vcc/6.708 nrp


wkb1> d -p atmif/10 vcc/6.708 nrp
==|================================================
#| WKB1
| AtmIf/10 Vcc/6.708 Nrp
==|========================(A)=====================
1|nextHop AtmIf/11 Vcc/6.708 Nrp
2|oamSegmentBoundary sameAsInterface
3|bandwidthElastic no
4|overrideHoldingPriority noOverride
==|================================================
1 ok Sep 03/04 05:31:03

wkb1>


If atmif/11 goes down, I want to tell vc 6/708 to take an alternate path. How could I use PNNI in this situation to setup something dynamic to solve this?


Tx again - Rainman
 
Rainman,
The quick answer to your question "How could I use PNNI in this situation to setup something dynamic to solve this?" is that you may not be able to in your case.

It depends mostly on where your passport ATM interfaces sit in the whole data path. If, for instance, your ATMIF/10 is a single (non-redundant) connection point from a customer, or a carrier, then there may be no point using a dynamic routing protocol like PNNI, as ATMIF/10 is always a single point of failure (SPOF). If, however, your ATMIF/11 etc gives you multiple egress points (and/or you don't care about ATMIF/10 being a SPOF) for your next hop, then you need to understand whether your passport will be initiating an SPVC to other passports etc in your network, or if you are just supporting a 'passed through' switched connection from somebody else (and in which case you need to support the PNNI signalling/addressing from the customer/carrier). The latter description is less likely, because supporting someone else's addressing scheme is like supporting their IP routing - would you want a fault in your routing to be able to affect your carrier/customer, or vice versa? You don't have to SHARE the same addressing, but you will need to establish a gateway functionality - a bit like BGP, I suppose. Anyway, it's a lot of planning work for possibly not much return. What is the alternative to ATMIF/11? Is ATMIF/11 your primary carrier, and the alternative interface is another carrier? If so, what do you expect to happen in the event of a failure - how do you want the second carrier to handle ATM routing? How long would it take you to identify a fault and simply repoint the existing NRP?

If you're set on PNNI, you must get hold of Nortel's manuals - the NTPs. Don't take no for an answer from your local support folks - make them earn their money. Tell them if their web page and documentation wasn't so crap (compare it with Cisco.com!) you wouldn't have to bug them. Look into PNNI in NTP 241-5701-702 "ATM Routing" and NTP 241-5701-710 "Configuration Guide".

If you want to provide some more info on what you want to have happen over the whole data path (not just switching cells out a different interface), I/others may be able to supply some more definite answers...
 
Hi Rainman

To set up the spvc, you first would need to delete the sub- component nrp off the atmif of the point of origin.Passport nearest the DSLAM).
delete atmif/xx vcc/0.32 nrp

(on the neighbouring atmif/, delete the entire corresponding vcc)
delete atmif/xx vcc/x.xx

then you would add an src
add atmif/xx vcc/0.32 src
which would include the following fields

calledvpivci:
callingvpivci:
calledaddress:

The called vpi vci can be left blank. The calling vpivci would be the VCC assigned to your subscriber. The called address would be the NSAP address of the atmif connected to your destination, for example B-RAS.

With this configuration the Passport would pick up the vcc(0.32) look at the calledaddress and define the shortest route to the destination end point. with this it would then route throughout the network with whatever vcc was available, until it reached the destination where it would return to 0.32. The route through the network is totally dynamic and will re-route if any outages (and congestion with ebr)are encountered

The other beauty of Passport, is you can add a secondary destination address, in case of failure on the first.

This is a quick overview, so may not be verbatim, but if you want to go down this road, please let me know if I can help
 
Mankyway -

I am going to give this a try - How do I determine the NSAP address of the Atmif? I tried doing a d atmif/x or a d -p atmif/x however the NSAP addy isn't listed.


-Rainman
 
go to the atmif you need to sort, do NOT be in prov mode, as d -o wil not work but command:

d atmif/x uni addr/*
 
Mankyway,

I just tried "d atmif/x uni addr/*" and it is not a valid command. I did a list to see what is valid and I'm getting:

switch> list atmif/1
AtmIf/1 CA
AtmIf/1 Pm
AtmIf/1 ConnMap
AtmIf/1 NAcct
AtmIf/1 Vcc/*
AtmIf/1 Pnni
ok


Is it possible that the Atmif is already setup for PNNI? When I do a d atmif/1 pnni/*,* I am getting:

wkb1> d atmif/80 pnni addr/*,*
==|=================================================================================
#| WKB1
| AtmIf/1 Pnni Addr/45000005702709500F010110100020480D005000,default
==|=============(A)=================================================================
1|scope -1
2|reachability internal
==|=================================================================================
1 ok



Also - if anyone wants to IM me thru AIM to chat about this, my AIM name is "DeadRX7".


Thanks again!

Rainman
 
Yep

Thats the one, sorry, but uni is usual for the "termination points" but that pnni addr/ is what you want to put in the calledaddress
 
I'm going to try it in the next day or so. Let me make sure I have everything correct. Here's my actual setup:

[dslam] --- [passport 15k] --|-- [passport 15k] --- [BSN B]
|
|
[BSN A]

Right now my DSL PPPoE connection goes through the DSLAM, thru the first passport 15k, and terminates on a Shasta (BSN A). My goal is to setup redundancy, so if BSN A is down, then I can terminate my PPPoE session on BSN B. I want to setup my passports so that the PVC re-routes.

Here's the steps I think that I need to take:

1 - Delete the NRP's off of the first 15K
2 - Delete the VCC nexthop on the first 15k
3 - Add atmif/X (nearest to the DSLAM) and setup src (calling vpi/vci, and the callingaddress need to be = to the Pnni addr).

Once these things are setup on my first atm switch, is there anything that has to be setup on the 2nd?

I just want to make sure I fully understand this before attmepting it, since it's on a production switch and I don't want to disrupt anything.


Thanks again,

Rainman
 
OK I must have misread the original request. The set up so far gives you redundancy through the network, but will always attempt to terminate on the NSAP Address given.

To include redundancy on the NSAP Addresses you will need to add a primary and a secondary NSAP Address. This is easy enough to do, but you must make sure that subscriber configuration is the same on both Shastas.

To give this redundancy you need to

add atmif/xx pnni addr/xxxxxxxxxxxxxxxxxxxxxxxxxxxx,primary
xx being the NSAP address of BSN A
add atmif/xx pnni addr/yyyyyyyyyyyyyyyyyyyyyyyyyy,alternate
yy being the NSAP address of BSN B
and because addr/yyy is a virtual interface, it will need a terminateSpvpAndSpvc component
add atmif/xx pnni addr/yyyyyyyyyyyyyyyyyyyyyyyyyy,alternate termsp

Hope this helps

 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top