Solved

discontiguous network and EIGRP

Posted on 2016-11-09
12
38 Views
Last Modified: 2016-11-16
I have 5 facilities that communicate with each other via EIGRP.  I'd like to use a management subnet/VLAN for each of the facilities. I was thinking to keep the first 2 octects and assign a number to the third octect that will identify the facility. For example, 10.10.3.0/24 for facility 1, 10.10.13.0/24 for facility 2, and so on. But I am not sure if it is the best practice for EIGRP and route summarization. Also, is this considered as discountiguous network? Thanks
0
Comment
Question by:leblanc
  • 5
  • 4
  • 2
  • +1
12 Comments
 
LVL 24

Assisted Solution

by:Ken Boone
Ken Boone earned 125 total points
ID: 41881160
So that should not be a problem with EIGRP.  But what happens when you have multiple subnets at each facility.  I would recommend going with the 2nd octet to identify the location, and the 3rd octet to identify the function..
i.e:

faciility #1  10.10.1.0 /  24 mgmt
                     10.10.2.0 / 24 users
                      10.10.3.0 / 24 voice
                      10.10.4.0 / 24 wireless

facitliy #2  10.20.1.0 / 24 mgmt
                    10.20.2.0 / 24 users
                     etc..

Either way you go EIGRP will not have an issue with it.
0
 
LVL 26

Accepted Solution

by:
Predrag Jovic earned 250 total points
ID: 41881629
Recommendations
 - for voice vlan is have vlan number be as low possible
 - stop thinking in decimal and always create your network boundaries as power of 2       ☺

explanation:
Start your network with the first available possible power of 2 number
10.10.0.0/24, 10.10.2.0/24, 10.10.4.0/24, 10.10.8.0/24, 10.10.16.0/24 etc

You should choose network addresses to prepare network ranges for nice, clean summarization.
10.10.3.0/24 and 10.10.4.0/24 can't be summarized even as with 10.10.0.0/22 (.0.0  - .3.0).
The best (the first one that have both of those with the least number of networks that are not in use) summarization that you can have for those two network is 10.10.0.0/21 (.0.0 - 7.0).
So, starting your network with number that can be starting point for summarization is a good practice.
is this considered as discountiguous network?
It could be, but it does not have to be discountiguous network depending how it will be summarized. Basically when networks is divided separate by another network you have discontiguous network.

summarization:
192.168.0.0/20 --- 10.10.10.0/8  --- 192.168.1.0/24
192.168.0.0 - .31/24 --- 10.10.10.0/8 --- 192.168.0.32 - .255/24
discountiguous - some network or part of a network is located outside summarized part.

summarization:
192.168.0.0/20 --- 10.10.10.0/8  --- 192.168.16.20/20
Could be considered contiguous

Discontiguous Subnets
One of the problems frequently encountered with classful routing protocols is the need to support discontiguous subnets. A discontiguous subnet is two or more portions of a major network that are divided by another major network. Figure 3.3 illustrates the concept.
0
 
LVL 45

Expert Comment

by:Craig Beck
ID: 41882742
Recommendations
 - for voice vlan is have vlan number be as low possible

I'm intrigued as to why!? :-)

EIGRP will be fine.  You don't have to use /24 addresses for each subnet so EIGRP won't care.
0
 
LVL 26

Expert Comment

by:Predrag Jovic
ID: 41883036
I read that recovery time per VLAN is around 10 ms after RSTP reconvergence, not all VLANs at once. Lower number VLANs converge faster, if voice is among the first VLANs calls will not be dropped. Sure if you use STP there will be no difference at all.
0
 
LVL 45

Expert Comment

by:Craig Beck
ID: 41883134
The time to reconverge is barely noticeable, especially if you don't have many VLANs.

If I had 4 VLANS, 1-4 for example, it would take the same time as if I had VLANs 10, 100, 500 and 2008.
0
 
LVL 1

Author Comment

by:leblanc
ID: 41883152
So it looks like I can use 10.10.x.0/24 for each of the location for the management VLAN. It will work but I will nto be able to take advantage of EIGRP summarization. Correct? Thanks experts
0
VMware Disaster Recovery and Data Protection

In this expert guide, you’ll learn about the components of a Modern Data Center. You will use cases for the value-added capabilities of Veeam®, including combining backup and replication for VMware disaster recovery and using replication for data center migration.

 
LVL 26

Assisted Solution

by:Predrag Jovic
Predrag Jovic earned 250 total points
ID: 41883348
You still can summarize if network ranges in different areas do not overlap. You can place more than one summary route under interface or you can move management range outside of range for summarization (if it is possible), or include it in summary route at once.

The best solution would be good planning, so you need just one route for each location
Management VLAN is one of summarized networks
Location 1# IP address ranges 10.10.0.0/19          /management 10.10.0.0/24
Location 2 # IP address ranges 10.10.32.0/19      /management 10.10.32.0/24
Location 3 #  IP address ranges 10.10.64.0/20     /management 10.10.64.0/24

Good solution is also
Location 1# IP address ranges 10.10.0.0/19          /management 192.168.0.0/24
Location 2 # IP address ranges 10.10.32.0/19      /management 192.168.32.0/24
Location 3 #  IP address ranges 10.10.64.0/19     /management 192.168.64.0/24
Management IP address range is outside of IP address range for summarization.

Create more than one summary route under interface to EIGRP neighbor (one for whole area and one for management), this one need to be planned carefully.
Location 1 # IP address range 10.10.0.0/19 (.0.0 - .31.0)
Management VLAN  10.10.3.0/24
interface FastEthernet0/0
 ip summary-address eigrp 1 10.10.3.0 255.255.255.0
 ip summary-address eigrp 1 10.10.0.0 255.255.224.0

Route table of of neighboring routers will have both configured summary routes in their tables
D    10.10.3.0/24 [90/409600] via x.x.x.x, 00:00:03, FastEthernet0/0
D    10.10.0.0/19 [90/409600] via x.x.x.x, 00:00:03, FastEthernet0/0
And sure, more specific route will always be used.
Creating summary routes causes breaking and reestablishing network relationship for a few seconds (schedule downtime).

But, from previous questions on EE, you are currently in the process of creating new IP address ranges for your locations. Create good plan (create IP address ranges per location with "wider" IP address range than you need currently (for example you need 6 network per location and create IP address range for each location /19 - future use)) and make it happen (if it is possible of course). :)

@Craig Beck
The time to reconverge is barely noticeable, especially if you don't have many VLANs.
Yes, especially if you don't need voice, or have only one VLAN, you don't have to care about it at all.
:)
If I had 4 VLANS, 1-4 for example, it would take the same time as if I had VLANs 10, 100, 500 and 2008.
Definitively. Just don't put voice at the end, most of the VLANs should  be assigned with higher VLAN number than voice... Having one digit VLAN for voice makes it idiot proof that not too much VLANs can be placed in front of voice VLAN by someone. By assigning to that VLAN 2008 for voice you are simply ignoring the fact about delay when networks are "coming back", so it might and will be used against you. ☺Many network engineers from my experience are typically unaware of the fact that there is a delay and order for "bringing VLANs back".
0
 
LVL 45

Assisted Solution

by:Craig Beck
Craig Beck earned 125 total points
ID: 41883491
Definitively. Just don't put voice at the end, most of the VLANs should  be assigned with higher VLAN number than voice... Having one digit VLAN for voice makes it idiot proof that not too much VLANs can be placed in front of voice VLAN by someone. By assigning VLAN 2008 for voice you are simply ignoring the fact about delay when networks are "coming back" and sending "bad message" and it might and will be used against you. ☺Network engineers from my experience are typically unaware of the fact that there is a delay and order for "bringing VLANs back".

Delay depends mostly on how well your STP is configured, how your links are configured and which version you're using.  If you're using PVST I'd be concerned, but not with RPVST.

You're right - there is a delay.  As I said though, it's largely irrelevant nowadays if you use Rapid Spanning Tree or MST (which uses Rapid anyway).  If you don't have a lot of VLANs you'll NEVER notice the reconvergence.  As an example, I've deployed networks with 10-15 VLANs on trunks with Voice VLAN IDs configured all over the place and not seen a single drop in traffic when RSTP reconverges.

You shouldn't be using STP for Voice VLANs anyway.  All Voice VLANs should be L3 separate in a good design. ;-)
0
 
LVL 1

Author Comment

by:leblanc
ID: 41883858
Nice explanation Predrag. Appreciate that.

"You shouldn't be using STP for Voice VLANs anyway" This is interesting as I am not aware of this all this time. Is there a white paper or some sort of article on this analysis? I always like to validate my understanding. Thanks
0
 
LVL 45

Expert Comment

by:Craig Beck
ID: 41883864
You only quoted half of my statement, leblanc ;-)

In a good design you shouldn't be stretching VLANs across access switches.  That means you'd have one Voice VLAN per switch.  It's not just a Voice recommendation, but a recommendation in campus LAN design in general, particularly where you have a core/distribution/access hierarchy.  Even if you only have a collapsed-core architecture you should still separate VLANs by L3.
0
 
LVL 26

Expert Comment

by:Predrag Jovic
ID: 41884103
RPVSTP according to Cisco is estimated to 1 second (but it is topology dependent), googling timers strech from 70 milliseconds, few hundred milliseconds, but according to Cisco Design documentation is suggested time up to 4 seconds depending on place in STP where link failure happens, but from my experience it was faster (but still there was occasional lost of pings). Regarding routing between access and distribution:
Deploying the L2/L3 Boundary at the Access Layer

Advances in routing protocols and campus hardware have made it viable to deploy a routing protocol in the access layer switches and utilize an L3 point-to-point routed link between the access and distribution layer switches
Newest design is placing routing between access (rack) and distribution L3 device for building. So VLANs would be localized per rack and not cross rack border, there can be multiple switches inside of rack. CCDA official book for 200-310 Design.
Limit VLANs to a single closet when possible to provide the most deterministic and highly available topology.
However, on the other hand if you stack access switches in rack you will get one logical device, so it could be considered access switch (not switches).

Cisco HA network design
Enterprise Campus 3.0 Architecture: Overview and Framework

Routing in the Wiring Closet
The Routed Access solution uses routing protocols already deployed throughout the network. Routing protocols are used in the wiring closet or access layer as well as in the distribution and network core. Extending routing to the wiring closet allows the network to respond to outages with fast recovery times, often less than 200 milliseconds (ms), and to better utilize existing network links and ports.
0
 
LVL 45

Expert Comment

by:Craig Beck
ID: 41884323
I didn't say routing between access and distribution, I said L3 at distribution to separate VLANs from different access switches/stacks.

You can do routing at the access layer but it's usually overkill in a campus. It's probably best placed where your access uplinks are the same speed as your access switchports, but even then there are other techniques that can help rather than using L3 at the access layer.

The three-tiered routing at distribution isn't new by any means. Routing at access is though, but it does just the same where STP is concerned, only with extra complexity.
0

Featured Post

How to run any project with ease

Manage projects of all sizes how you want. Great for personal to-do lists, project milestones, team priorities and launch plans.
- Combine task lists, docs, spreadsheets, and chat in one
- View and edit from mobile/offline
- Cut down on emails

Join & Write a Comment

Suggested Solutions

Configuring network clients can be a chore, especially if there are a large number of them or a lot of itinerant users.  DHCP dynamically manages this process, much to the relief of users and administrators alike!
ADCs have gained traction within the last decade, largely due to increased demand for legacy load balancing appliances to handle more advanced application delivery requirements and improve application performance.
After creating this article (http://www.experts-exchange.com/articles/23699/Setup-Mikrotik-routers-with-OSPF.html), I decided to make a video (no audio) to show you how to configure the routers and run some trace routes and pings between the 7 sites…
In this tutorial you'll learn about bandwidth monitoring with flows and packet sniffing with our network monitoring solution PRTG Network Monitor (https://www.paessler.com/prtg). If you're interested in additional methods for monitoring bandwidt…

705 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

20 Experts available now in Live!

Get 1:1 Help Now