There has been some interesting discussions on configuration differences with IOS and JUNOS on Twitter the past few days.  One of the biggest things has been the inclusion of a Rollback command in JUNOS; and to be perfectly honest, that has to be one of the best commands that JUNOS has!
In case you are not familiar with it, in JUNOS if you submit a change and then enter commit confirm (default 10 minutes), the router will not “save” the configuration change until you commit the change again.  This way if you make a change, lock yourself out of the device, it will revert to the previous configuration and you should have your access back.
(some housekeeping)
Router/Switch Output
Commands
Notes

[edit]
user@host# commit confirmed
commit confirmed will be automatically rolled back in 10 minutes unless confirmed
commit complete
#commit confirmed will be rolled back in 10 minutes
By entering commit again, the change will become permenant
[edit]
user@host# commit

Now, what about IOS?  Well, typically we do the reload in x command before we make a change, but is there another way?  Well, yes there is!  We have the archive command and the config replace command string that will perform a similar function to to the JUNOS command and less drastic then the reload in command.
So, lets first cover the archive command.
The archive command allows you to make an archive of the config either over a time-period (say 1440 minutes) or when the write-memory command is initiated.  The archive can be stored in flash, tftp, ftp, or a few other places. To setup archiving:
Rack1R2#conf t
Enter configuration commands, one per line.  End with CNTL/Z.
To enter the archive config, enter the archive command
Rack1R2(config)#archive
From there, you need to tell it 1) where to store the file and 2) what to call it.
Here it will be flash:/archive with a name of Rollback
Rack1R2(config-archive)#path flash:/archive/Rollback
Here I tell it to archive the previous config whenever the write-memory command is executed
Rack1R2(config-archive)#write-memory
Also, I might as well make sure that we back it up once a day at a minimum.  Range is from 1 minute to 525600 minutes (1 year)
Rack1R2(config-archive)#time-period ?
<1-525600>  Number of minutes to wait between archive creation
Rack1R2(config-archive)#time-period 1440
Rack1R2(config-archive)#
Now, if we look at the flash:/archive/ folder, what will we see?  Lets look (I already have some saved in memory)
Rack1R2#dir flash:archive
Directory of flash:/archive/
17  -rw-        1806   Jun 3 2011 19:34:16 +00:00  Rollback-2
18  -rw-        1806   Jun 3 2011 19:35:02 +00:00  Rollback-3
16  -rw-        1744   Jun 3 2011 19:31:54 +00:00  Rollback-1
512057344 bytes total (328622080 bytes free)
Rack1R2#
As you can see, we have a few files there, each appended with a number (call it revision with highest being most recent). To check, use the show archive command
Rack1R2#show archive
There are currently 4 archive configurations saved.
The next archive file will be named flash:/archive/Rollback-4
Archive #  Name

1       flash:/archive/Rollback-1
2       flash:/archive/Rollback-2
3       flash:/archive/Rollback-3 <- Most Recent
4
5
6
7
8
9
10
11
12
13
14
Rack1R2#
So, what does all this mean and how does this relate to the commit confirm command in Junos?  Good question!  Here we go!
So, when you need to do a change, just execute the archive config command prior to the change.  That will create a snapshot of the existing config in the archive directory.
Rack1R2#archive config

Lets take a look and see what revision we have saved:
Rack1R2#show archive
There are currently 5 archive configurations saved.
The next archive file will be named flash:/archive/Rollback-5
Archive #  Name

1       flash:/archive/Rollback-1
2       flash:/archive/Rollback-2
3       flash:/archive/Rollback-3
4       flash:/archive/Rollback-4 <- Most Recent
5
6
7
8
9
10
11
12
13
14
Rack1R2#
Now, there is another command that we need to know about – its the configure replace command!  We can use that similar to the commit confirm command because of the archive command.
Rack1R2#configure replace flash:archive/Rollback-4 time 60
That is all good – lets test the command!
 
I have R1 and R2 configured like the diagram below, connected via Frame-Relay and all interfaces are in RIPv2

Here is the routing table and interface status on R2:
Rack1R2#sh ip route
Codes: C – connected, S – static, R – RIP, M – mobile, B – BGP
D – EIGRP, EX – EIGRP external, O – OSPF, IA – OSPF inter area
N1 – OSPF NSSA external type 1, N2 – OSPF NSSA external type 2
E1 – OSPF external type 1, E2 – OSPF external type 2
i – IS-IS, su – IS-IS summary, L1 – IS-IS level-1, L2 – IS-IS level-2
ia – IS-IS inter area, * – candidate default, U – per-user static route
o – ODR, P – periodic downloaded static route
Gateway of last resort is not set
150.1.0.0/16 is variably subnetted, 4 subnets, 2 masks
R       150.1.1.1/32 [120/1] via 150.1.12.1, 00:00:23, Serial4/0.1
C       150.1.12.0/24 is directly connected, Serial4/0.1
R       150.1.11.0/24 [120/1] via 150.1.12.1, 00:00:23, Serial4/0.1
R       150.1.111.1/32 [120/1] via 150.1.12.1, 00:00:23, Serial4/0.1
150.2.0.0/32 is subnetted, 3 subnets
C       150.2.222.2 is directly connected, Loopback2
C       150.2.22.2 is directly connected, Loopback1
C       150.2.2.2 is directly connected, Loopback0
Rack1R2#sh ip int br | ex un
Interface                  IP-Address      OK? Method Status                Protocol
Serial4/0.1                150.1.12.2      YES manual up                    up
Loopback0                  150.2.2.2       YES manual up                    up
Loopback1                  150.2.22.2      YES manual up                    up
Loopback2                  150.2.222.2     YES manual up                    up
Rack1R2#
Here is the same information for R1
Rack1R1#sh ip route
Codes: C – connected, S – static, R – RIP, M – mobile, B – BGP
D – EIGRP, EX – EIGRP external, O – OSPF, IA – OSPF inter area
N1 – OSPF NSSA external type 1, N2 – OSPF NSSA external type 2
E1 – OSPF external type 1, E2 – OSPF external type 2
i – IS-IS, su – IS-IS summary, L1 – IS-IS level-1, L2 – IS-IS level-2
ia – IS-IS inter area, * – candidate default, U – per-user static route
o – ODR, P – periodic downloaded static route
Gateway of last resort is not set
150.1.0.0/16 is variably subnetted, 4 subnets, 2 masks
C       150.1.1.1/32 is directly connected, Loopback0
C       150.1.12.0/24 is directly connected, Serial4/0.1
C       150.1.11.0/24 is directly connected, Loopback1
C       150.1.111.1/32 is directly connected, Loopback2
150.2.0.0/32 is subnetted, 3 subnets
R       150.2.222.2 [120/1] via 150.1.12.2, 00:00:12, Serial4/0.1
R       150.2.22.2 [120/1] via 150.1.12.2, 00:00:12, Serial4/0.1
R       150.2.2.2 [120/1] via 150.1.12.2, 00:00:13, Serial4/0.1
Rack1R1#sh ip int br | ex un
Interface                  IP-Address      OK? Method Status                Protocol
Serial4/0.1                150.1.12.1      YES manual up                    up
Loopback0                  150.1.1.1       YES manual up                    up
Loopback1                  150.1.11.1      YES manual up                    up
Loopback2                  150.1.111.1     YES manual up                    up
Rack1R1#
Lets test the connectivity by PINGing R2 Loopback 2 from R1 Loopback2
Rack1R1#ping 150.2.222.2 so lo2
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 150.2.222.2, timeout is 2 seconds:
Packet sent with a source address of 150.1.111.1
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 56/57/60 ms
Rack1R1#
Good, now lets test the config replace command (IOS Rollback)
First, we will archive the config
Rack1R2#archive config
Now lets see what archives we have:
Rack1R2#sh archive
There are currently 6 archive configurations saved.
The next archive file will be named flash:/archive/Rollback-6
Archive #  Name

1       flash:/archive/Rollback-1
2       flash:/archive/Rollback-2
3       flash:/archive/Rollback-3
4       flash:/archive/Rollback-4
5       flash:/archive/Rollback-5 <- Most Recent
[– SNIP –]

Ok, Rollback-5 is the most recent.
Now on R1 we will start a 1000 packet ping
Rack1R1#ping 150.2.222.2 so lo2 repeat 100
Type escape sequence to abort.
Sending 1000, 100-byte ICMP Echos to 150.2.222.2, timeout is 2 seconds:
Packet sent with a source address of 150.1.111.1
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Now that we have that started, lets move on
Now, on R2 we will execute the configure replace command, set the timer to 60 seconds, then shut down the Serial 4/0 (frame-relay) interface
Rack1R2#configure replace flash:/archive/Rollback-5 time 60
Timed Rollback: Backing up to flash:/archive/Rollback-7
It will now ask you to make sure you want to do this – checks and balances
This will apply all necessary additions and deletions
to replace the current running configuration with the
contents of the specified configuration file, which is
assumed to be a complete configuration, not a partial
configuration. Enter Y if you are sure you want to proceed. ? [no]: Y
Total number of passes: 0
Rollback Done
*Jun  3 21:43:22.026: Rollback:Acquired Configuration lock.
Rack1R2#conf t
Enter configuration commands, one per line.  End with CNTL/Z.
Rack1R2(config)#int ser 4/0
Rack1R2(config-if)#shut
Rack1R2(config-if)#
*Jun  3 21:43:28.274: %LINK-5-CHANGED: Interface Serial4/0, changed state to administratively down
*Jun  3 21:43:29.274: %LINEPROTO-5-UPDOWN: Line protocol on Interface Serial4/0, changed state to down
Rack1R2(config-if)
Hmm, if we look at the ping that we have going on R1 we see that it can no longer PING R2
Rack1R1#ping 150.2.222.2 so lo2 repeat 1000
Type escape sequence to abort.
Sending 1000, 100-byte ICMP Echos to 150.2.222.2, timeout is 2 seconds:
Packet sent with a source address of 150.1.111.1
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!………………………….
……..
Lets wait…. there, the config just rolled back on R2 and R1 can PING again!
Rack1R1#ping 150.2.222.2 so lo2 repeat 1000
Type escape sequence to abort.
Sending 1000, 100-byte ICMP Echos to 150.2.222.2, timeout is 2 seconds:
Packet sent with a source address of 150.1.111.1
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!………………………….
…………………!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
!!!!!!!!!!!!!!!!!!!!
Success rate is 94 percent (948/1000), round-trip min/avg/max = 56/57/240 ms
Rack1R1#
So what happened on R2?  Here is the output on the console:
Rack1R2(config-if)#Timed Rollback: rolling to:flash:/archive/Rollback-7
!Pass 1
Neat, it will show you the commands it is about to enter!

!List of Commands:
interface Serial4/0
no shutdown
end
Total number of passes: 1
Rollback Done
*Jun  3 21:44:22.218: Rollback:Acquired Configuration lock.
Rack1R2(config-if)#
*Jun  3 21:44:22.778: %PARSER-3-CONFIGNOTLOCKED: Unlock requested by process ’21’. Configuration not locked.
Rack1R2(config-if)#
*Jun  3 21:44:24.586: %LINK-3-UPDOWN: Interface Serial4/0, changed state to up
*Jun  3 21:44:25.586: %LINEPROTO-5-UPDOWN: Line protocol on Interface Serial4/0, changed state to up
Rack1R2(config-if)#
If all is good after your change and you need to not roll-back, just issue the command configure confirm and it will cancel the rollback._
_
Rack1R2#configure confirm
 
If you already issued that command, you will get the following output:
Rack1R2#configure confirm
%No timed rollback pending
Rack1R2#
Not as simple as the JUNOS commands, but is still valuable in the IOS world!
 
R1 Config Here
R2 Config Here
 

Enterprise Network Testing

 
Let me start off by saying that this is not the “normal” Cisco Press book where you learn technologies; no with this book you actually see how to use what you have learned and ways to approach the testing of those technologies.   This book covers the often neglected part of what we do, proper testing and planning.  It is refreshing to be able to read a book like this, it is heavy on the acronyms, but that is almost expected for a Cisco Press book anyways!
The book is divided into three main parts.  Part 1 – Introduction to Enterprise Network Testing, Part 2 – Case Studies, and Part 3 – Test plans.
Part 1 reviews the basics of business cases, network life-cycles, testing and lab strategies, test approach, and finally, executing a test plan.  The Network Life-cycle is something that we know about but never tend to look at.  Not that we intentionally ignore it or anything, but I think most shops run a bit slim in the Network Engineering group and tend to have to support the business as well as take care of current projects.   It is sometimes nice just to be able and sit back and follow the PPDIOO, but alas those times are few and far between.  Below is a diagram that can act as a friendly reminder that we should do that more often then we do today.  Perhaps the material in this book is something that can help each of us remind the senior management about this process and they might start to understand the benefits that this approach can bring.

 
 
When it comes to the case studies of this book, they are some amazing studies!  They are not the small networks that you would think, they are larger networks and show you the thought that is behind the design phase. It can help you to understand the bigger picture of a design as well as provide a sheltered approach to larger network designs.  It is not everyday that we need to design a large MPLS network for a company, but if we ever need to support one we might have a better understanding of what we are looking at. The case studies walk you though the process discussed above, so you can start to see the actual application of the concepts – and thus be able to gain a better understanding of them.
The last part, Part 3, is all test plans.  These are very nice reference materials that can be looked back upon when you are designing.  They can help you to formulate your own test plans, and by referencing back you can make sure you do not miss anything.  There are numerous plans for different technologies – from DMVPN/GetVPN to Nexus to Wide Area Application Services.  They cover the whole 10 yards of information – everything you need to think about is there!
 
There are a few other gems in this book that surprised me.  When you think of testing a network, typically you take a router and use it for a single function.  In the beginning of this book they show you how to make 1 router into 100 routers in a relatively simple fashion.  I have done small scales of this in the past, but the way that they covered this approach was easy to understand and comprehend.  Another gem was the mention of the little know service, chargen, and how to use it in your test scenarios.  We very nice to see that mentioned – need to remember to use that one from time to time as I forgot about it!
Overall this is a very good book – one that I am glad is now in my Cisco Press library!
 

There was some quick conversation on Twitter this morning about the CCIE Lab being available during Cisco Live 2011.  Well, after some digging around I was able to find a copy of the ad that talks about the availability of the lab.  Based on what the ad says, seats are available on July 9th and July 10th only.  This lab can only be scheduled via the Cisco Live registration site, below are the instructions on how to schedule.
I have to say that this is a long time coming, and lets hope they continue this with – say – the CCDE!?!?  It does not say what CCIE labs will be available, and my best guess would be R&S only at this time.
UPDATE – 5/31 – It is my understanding (from a beautiful source) that the following CCIE Labs will be available:

 
Welcome to the new site!  I have recently moved the site from the WordPress hosted site so a new location.  This site is now reachable via blog.fryguy.net or www.fryguy.net, www.fryguy.net being the preferred method going forward.  I still need to finish up a few things here and there, but all the posts have been moved with all their content – so we should be good to go!
Why did I move it you may ask, well I just felt it was time to be able to branch out a bit.  I can now host video, flash, or anything else that suit my readers.
So, if you have any suggestions or ideas – just let me know!  Also, if you find any problems – please let me know so that I can correct them quickly.
Thank you, and again – Welcome!

 
I will eventually be migrating the current site – blog.fryguy.net – over to here as time allows.  For now, please visit the site – blog.fryguy.net

Scott – @scottm32768 on twitter – posted a link about this nifty piece of software called DisplayLink.  It caught my attention very quick as he said it makes your iPad into a display for your Windows 7 PC.  This is something I have been looking for for quite some time.  I have seen some of the people at my work do this with their Apple computers using either the DisplayLink software or another vendor’s I believe.
So why am I doing a post on it, well to help get the word out, that is why!  The DisplayLink application definitely helps to free up the real-estate on your computer as you can toss your Twitter application to it, your e-mail, perhaps a terminal application, or anything else you want to keep an eye on.  This additional real estate on our computers is wonderful!
From their website, they say that the key features are:

This is just a quick post on a new venture one of my good friends, Brandon, has embarked on.  Brandon is a well recognized and respected Cisco trainer in the industry and has published a few books for Cisco Press over the years.  His books include CCNA Wireless Official Exam Certification Guide and _Cisco Access Control Security _as well as a few others.
Recently he has decided to embark on another endeavor, and this one could not have better timing.  He is launching a website for IPv6 tutoring called MyIPv6Tutor.com
If you want to learn more about IPv6 and such, might be a good time to head over there and see what a Cisco Certified Systems Instructor has to say about IPv6!

We have all been hearing about IPv6 for a few years now. There have been numerous white papers published, blog posts, audio podcasts, as well as the normal twitter chatter on the topic. Lately this topic has come to the forefront more and more since the RIR have handed out the last of the IPv4 /8 networks to the regional registrars. So what does all this mean to you and I? Well, if you have not fully embraced IPv6, the time has definitely come.
Luckily Cisco Press has recently released IPv6 for Enterprise Networks (Networking Technology) for our reading pleasure. Originally I was going to wait until Cisco Live 2011 to get this book, but I just could not wait and my friend, Jamie, was able to help me get it sooner. I am glad that I did not wait! This book is a great read and will help the reader to understand IPv6 as well as help to design and deploy a network.
The book consists of 12, well formatted chapters. They take you from the market drivers, to designing and services, deployment, and all the way to the data center and testing.  What is nice about this book is the progression on the topics.  You do not really need to know IPv6 to read this book, it will actually guide you through the whys and hows then onto the how to as well as testing.  It is nice to have a book that will cover most of the topics that one will experience in an Enterprise design.
The first two chapters in the book talk about the market drivers for IPv6, these are some of the whys that people are wondering.  This is a pretty quick chapter as the killer application is not really out there yet, but is sure to be coming (my best guess will be Asian/Pac e-commerce).  What this chapter does elude to is that this is a good time to be able to restructure your network in a more efficient design.  Consider this, you probably inherited your current network and the design that is currently in place.  With a transition to IPv6, you now have the opportunity to reconsider what was done and determine were you should take it.  The first chapter does good at giving you some points to consider as you read on and helps you to understand some of the underlying design scenarios.
The next two chapters, 3 and 4, start to go over how IPv4 and IPv6 will co-exist in the network and some of the ways to approach this challenge.  There are dual-stacking, 6-to-4 tunnels, MPLS, and the good old duct tape of the network – GRE.  It also continues to talk about services and some of the routing options that one has to consider.  Service such as multicast and QoS are discussed for a decent amount of pages.  These tend to be some of the more important items in an enterprise network.  These chapters also discuss the different routing protocol options – OSPFv3, EIGRPv6, and IS-IS with a mention of BGP.  What does surprise me, there is no discussion on LISP 🙂  Ok, that might have been asking for a bit much, but I do like LISP.
I think that Chapter 5 is a really important chapter in some ways.  This is the chapter where Planning an IPv6 Deployment is discussed.  This is a topic that should not be taken lightly but one that should be a primary focus.  After all the technology and numbers, all that you really have to stand upon is good design.  As i mentioned before, what is nice with IPv6 is that you do not have to base your new design on the existing design, you can actually rethink everything and start fresh.  For instance, if you are currently running EIGRP but want to move to OSPF, deploying IPv6 is the time to start that move.  The two protocols (v4 and v6) are separate from each other.
The next chapters, 6 through 10, are the heart of this book.  Discussed in some good detail is deploying IPv6 in Campus, Virtualized, WAN/Branch, Data Centers as well as Remote Access.  Each of these topics has a good chapter dedicated to it and there are many things that one needs to consider with each of these deployment scenarios.  Again, it all comes down to the choices made during the design and testing phase.
Chapter 11 is a nice chapter on the aspects of Managing IPv6 Networks.  It covers the different ways to monitor and secure various parts of the network.  It is nice to actually see a chapter dedicated to some of the stuff that is quickly overlooked during a design an deployment – how to manage and support all the work that you have done.  Most of this stuff is usually figured out near the end of a project when we hand off things to the NOC – and they say IPvWhat?
The final chapter in the book was a welcomed surprise to me.  This chapter actually talks about how to setup a lab to test IPv6.  It gives some good sample setups that should help the person be able to play with the protocol as well as test different scenarios.  Granted, some of the hardware suggested might be a bit out of reach for most of us (VSS, ISR, etc) – but it is a good starting point to see what you can do.  One can easily see ways to substitute certain hardware for other hardware so one can learn and develop.  If you are going to roll some of this into production, test hardware is invaluable for knowing the problems before they are experienced in a live network.
Overall it is a good read and one that should be handy to those who are working with IPv6 right now.  Not only can you use the information within the book for your own knowledge, but it gives you information on how you can explain it to others.  One of the most difficult things that I come across in an enterprise is explaining difficult topics.  This book helps me to find ways to explain things in ways that others might be able to understand.

Well, it is almost that time a year when all the “Networkers” get together at the annual Cisco Live event. It is still a few months away when I am writing this – but as we all know – time slows down for no one.  I wanted to take a few moments and share with you why I attend as well as why you should consider attending.   Please keep in mind that my experience is a bit different from the normal attendee as I am considered a NetVet as well as a CCIE, and being that we get a few additional perks during the event.