Routing

I heard from the news that EIGRP will be an open standard protocol.

I am just wondering if JunOS will have a support in its new software versions?

EIGRP is a cisco proprietay protocol? dont think they'll just release that...

---

@Mr_Reyes wrote:

EIGRP is a cisco proprietay protocol? dont think they'll just release that...

---

http://www.cisco.com/en/US/prod/collateral/iosswrel/ps6537/ps6554/ps6599/ps6630/qa_C67-726299.html

EIGRP is being re-released as an open standard.

I am also curious if Junos will take on support for it, if for no other reason than it will make it easier for Juniper to break into many markets and customers that are "pure Cisco" because, among other reasons, they were designed around EIGRP.

It'll certainly be interesting:

• Cisco as intimated is the dominant player in the Enterprise space, or at least those enterprises with money 
• Juniper et everyone else not named Cisco have spent years bashing EIGRP, some marketing, some technical reasons in my estimation

Networking has taken a back seat in my career (until recently), but back in the day EIGRP was phenomenal not so much for it's IP routing, but for making IPX and Appletalk not suck on WAN's.  Huge efficiencies in the old Frame Relay and X.25 networks.

These days with transport (comparitively) cheap, and the world has moved to IP (exception of some SNA traffic and even that was encapsulated everywhere we could in IP 15 years ago) I'm not so certain it has the compelling attraction it once did.  Previously EIGRP was everywhere in Cisco land: simple config, worked better than OSPF in some earlier versions of IOS, and it scaled well to the size of the networks I was routinely working with at the time (10-50 sites).

Now, I have to think that Cisco's doing this because it's no longer a competitive advantage to them; namely, I suspect more and more of their customers have transitioned to OSPF in the Enterprise space and so it doesn't make sense to keep it propietary any longer.  I certainly wouldn't recommend anyone adopt EIGRP these days, vendor lockin is problematic when we're talking about non-trivial amounts of money being spent on infrastrucure.

About the only compelling reason I could see for Juniper / Arista / Brocade / etc,. would be for migrations off CIsco; may reduce the complexity of migrations and the operational support requirements during the transition.

--Rev

I agree to that.

Most of the customers I know, they transitioned to OSPF. 

And I only encountered EIGRP on my Cisco Labs. hahahaha. 

But anyways. Lets just wait for JunOS updates if they will adopt the concept of EIGRP. 

I beleive JunOS will be even better when it will adopt the EIGRP because customers from other vendors will test and find out the good thing with JunOS. 

---

@networkSwagger wrote:

 

Most of the customers I know, they transitioned to OSPF. 

And I only encountered EIGRP on my Cisco Labs. hahahaha. 

---

Actually, I've seen a LOT of EIGRP out "in the wild."

Some very large customers... I can't name the places, but let's just say some really large enterprises and some really large service providers...  they have huge investments in Cisco and huge networks built on EIGRP.  The amount of time it would take and the risk of core network outages are huge blocks in the way of them moving off EIGRP.  The sentiment is usually "The network is working fine, we can't risk massive problems for the sake of changing something just to change it.  We have no reason to stop using EIGRP..."

That mindset pretty much knocks everyone but Cisco off the table when new products are evaluated, tested, etc.  I've seen some HUGE network messes (we're talking 40,000+ static routes) on networks where they layered Juniper SRX 5k series into huge data centers built on Cisco 6500s and CSM modules (on the order of 50 SRX devices and 100+ Cisco 6500s and Nexus 7ks)

Most places (understandably) don't want to get themselves into those kind of situations... so the answer is "well if we just keep using Cisco this is all really easy..."   which means in many cases they may not be buying the best product to suit their needs -- they're buying the path of least resistance.

By having EIGRP open and support start to show up in competing products (honestly, I would be very surprised if Juniper had working, usable, field-ready support in less than 2 years), at least the door would, in theory, open up to other vendors because the argument of "we don't want to mess with our EIGRP core" no longer automatically precludes other vendors from making an entry into the network.

I am not a pundit, nor a blogger, nor an expert of any kind whatsoever (  ) but I would say once the standard is ratified, look to some of the more agile vendors (Brocade comes to mind...) to have support for it earlier than Juniper.  Juniper may have something in a non-supported release, but typically those bleeding-edge releases take a really long time to filter into supported releases.  We're still a long ways off from having ubiquitous support for EIGRP, but at least strides are being made.

FYI:

http://www.ietf.org/staging/draft-savage-eigrp-00.txt