x
Page 1 / 2   >   >>
bollocks187 12/5/2012 | 2:58:50 PM
re: Nortel Preps New PBT Switch This looks like the makings of a "new agressive" Nortel. Their current pioneering work in the PBT space is to be congratulated.

Nortel from the ashes.
fsubob 12/5/2012 | 2:58:49 PM
re: Nortel Preps New PBT Switch I agree, impressive direction after PBT launch...
go_ON 12/5/2012 | 2:58:48 PM
re: Nortel Preps New PBT Switch Oh please. Congradulations are far too premature here. Let them deliver a stable product and get real depoyments that are not press anouncements around lab trials. When was the last succesful packet platform they released? The Italian deployment is a disaster from what I hear. And the "simple" PBT/PBB gets more complex (read expensive) every quarter in order to plug the gaps that MPLS already fills

digits 12/5/2012 | 2:58:48 PM
re: Nortel Preps New PBT Switch You have to admit, though, that Nortel has, this time, managed to come up with a decent name for a product - MetroNext. That's actually pretty good.

Now all they need to do is stick with it -- surely that is not a name they would replace -- AND deliver something that works.... the tough end of the proposition.

Ray
chook0 12/5/2012 | 2:58:48 PM
re: Nortel Preps New PBT Switch I'll believe it when I see it. Nortel has been better at announcing products than delivering them recently. But I'm hoping they can bring it off this time.

--chook
Metropolitian 12/5/2012 | 2:58:44 PM
re: Nortel Preps New PBT Switch I think Nortel is getting ahead of themselves.
The internet world is (and always will be) Cisco-centric. MPLS is the defacto standard for good reason. It has a 10 year proven track record. PBT is essentially an unproven technology with numerous limitations.

1) It is a point-to-point technology
2) It does not support mutlicasting, therefore is useless for applications like IPTV.
3) By disabling Spanning Tree (which Nortel claims is a good thing), the network topology has to be manually entered. A NMS nightmare.

When "Nortel says many carriers haveshown interest in PBT", this actually translates to: Lets tell Cisco we are considering PBT, so that we can negiotate concessions on pricing for future MPLS equipment purchases.

Metro
macster 12/5/2012 | 2:58:44 PM
re: Nortel Preps New PBT Switch I think Nortel is getting ahead of themselves.
The internet world is (and always will be) Cisco-centric. MPLS is the defacto standard for good reason. It has a 10 year proven track record. PBT is essentially an unproven technology with numerous limitations.

1) It is a point-to-point technology
2) It does not support mutlicasting, therefore is useless for applications like IPTV.
3) By disabling Spanning Tree (which Nortel claims is a good thing), the network topology has to be manually entered. A NMS nightmare.

When "Nortel says many carriers haveshown interest in PBT", this actually translates to: Lets tell Cisco we are considering PBT, so that we can negiotate concessions on pricing for future MPLS equipment purchases.

Metro

======================

1 & 2). PLSB will go a long way to address these.
3). There'll be automated control.

There is a big change in going from enterprise class to carrier class (e.g. mindset, cost, etc), and my concern is the cost justifications of carrier class ethernet - this is the pitch used! Saying that, Eth transport for primarily (if not solely) ethernet services has its place in the market!!!
go_ON 12/5/2012 | 2:58:43 PM
re: Nortel Preps New PBT Switch 1 & 2). PLSB will go a long way to address these.

PLSB looks a lot like ISIS-TE, so hence my point about PBT getting more complex and expensive to mimic what MPLS already does, AND PLSB is slideware as of now

There'll be automated control.

When and how? and when will it be multi-vendor? there are two routes:- 1 Control Plane 2 NMS managing multiple vendors - Each option adds expense. In fact I think option 1 is a better long term solution and that too is already done and deployed wrt MPLS

PBT only has merit (to me) if it can be deployed and managed significantly cheaper than say a ethernet MPLS solution, and I have not seen reasons to suggest this is the case so far
Metropolitian 12/5/2012 | 2:58:43 PM
re: Nortel Preps New PBT Switch My other issue with PBT is the added overhead.

Granted, PLSB might address broadcast and multicast issues.

PLSB adds 8 bytes extra overhead to your existing payload. This translates to 12% added overhead for 64 byte traffic and 0.5% added overhead for 1522 byte traffic. This averages out to 6% added overhead for a random packet flow. This translates to 5 to 6% reduction in traffic capacity of your existing links.


macster 12/5/2012 | 2:58:41 PM
re: Nortel Preps New PBT Switch 1 & 2). PLSB will go a long way to address these.

PLSB looks a lot like ISIS-TE, so hence my point about PBT getting more complex and expensive to mimic what MPLS already does, AND PLSB is slideware as of now

There'll be automated control.

When and how? and when will it be multi-vendor? there are two routes:- 1 Control Plane 2 NMS managing multiple vendors - Each option adds expense. In fact I think option 1 is a better long term solution and that too is already done and deployed wrt MPLS

PBT only has merit (to me) if it can be deployed and managed significantly cheaper than say a ethernet MPLS solution, and I have not seen reasons to suggest this is the case so far

=================================================

go_ON,

If you want to quote me, please include the whole thing! If you go back to my previous post, what you are doing is merely repeating what I have said.

I, on the other hand, pointed out aspects of your previous post which was not entirely true!

Your last paragraph above...
"PBT only has merit (to me) if it can be deployed and managed significantly cheaper than say a ethernet MPLS solution, and I have not seen reasons to suggest this is the case so far"

Again, if you refer back to my previous post...
Page 1 / 2   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE