And we’re back for an exciting conclusion to the CCIE Troubleshooting series. Just what fate is in store for our heroes? Well, if we’re anything like the Friday the 13th movies, you may need to wait about 20+ years to find out!
You’re probably asking yourself, “With the amount of stuff we’ve gone through already, you’re really telling me there’s more to be concerned about?”
Yup, that’s exactly what I’m talking about. So what next then? How about little unexpected zingers that may just confuse, confound and otherwise astound you?
Have you ever rebooted your router/switch and had it reply to you “Would you like to enter basic management setup?” Of course you have. All the time in practice labs! But what if it happened in the middle of your actual lab exam? You DID save your configuration, right? You didn’t see any error messages pop up did you? Well, how about that?
How about a bit of preventive medicine? When you start the day, in SecureCRT that you’ll be using in the lab, I would click on the Session Options and then on the Emulation (under Terminal in the left-hand pane). See the part about Scrollback buffer. Set that to AT LEAST 5000 lines. Gives you the ability to review things you’ve typed or what the router has said, or recreate your steps in case of an emergency.
But back to this dilemma. What do you do? Run in circles, scream and shout? While that may be entertaining, it would scare everyone else in the room and not be overly helpful to your cause! Whatever you do, DO NOT enter the setup mode. Before completely panicking, just do a “show start” and see whether your config was really saved or not!
If it did not save, go check your scrollback and see what error you missed. Then hope you have enough scrollback to reconstruct your configuration. THEN run in circles, scream and shout!
You likely were victimized by the configuration-register. Well, ok, you were victimized by the Proctor’s warped sense of humor, but the weapon of choice was the configuration-register. Deal with it. You should have checked ahead of time! “show version | include Config”
Rack1R1(config)#do sh ver | in Config Configuration register is 0x2102 Rack1R1(config)#
That would be the sign of a good router.
0×2142 is the sign of a bad router – your startup configuration will be skipped
0×2101 is the sign of a really bad router – you will go into bootstrap or ROMMON mode
Those are the primary ones to think about. There are others that involve the changing of console speed. While particularly entertaining, being that you do not have access to the terminal server configuration or physical access to devices any longer, that’s not something you’ll be worrying about. Back in my day, that was one method of torture potentially bestowed upon us.
So what do you do if you end up in ROMMON? Partly that will depend on what routing platform you happen to be using. But let’s go with the standard ISR routers (28xx and 38xx which I believe are the currently listed devices on the blueprint for Lab Hardware).
rommon 1 > rommon 1 > ? alias set and display aliases command boot boot up an external process break set/show/clear the breakpoint confreg configuration register utility cont continue executing a downloaded image context display the context of a loaded image cookie display contents of motherboard cookie PROM in hex dev list the device table dir list files in file system dis disassemble instruction stream dnld serial download a program module frame print out a selected stack frame help monitor builtin command help history monitor command history iomemset set IO memory percent meminfo main memory information repeat repeat a monitor command reset system reset rommon-pref Select ROMMON set display the monitor variables showmon display currently selected ROM monitor stack produce a stack trace sync write monitor environment to NVRAM sysret print out info from last system return tftpdnld tftp image download unalias unset an alias unset unset a monitor variable xmodem x/ymodem image download rommon 2 >
You’ll have a very specific set of commands there having to do with the boot process and/or reloading IOS through the serial port (not much fun, and not possible over telnet as configured!).
You can use the “dir flash:” command to see what files are available (or “dev” if you need to know the names of current devices like DISK0: or DISK1: for PCMCIA cards) and then “boot flash:(filename)” if there’s any doubt.
Knowing how to get out of ROMMON is a great skill. Murphy’s Law says whatever can go wrong will go wrong, and generally at the most inopportune moment! Know how to reboot!
So what if you get into a router at the beginning of your lab and you see this:
rommon 1> rommon 1>? Exec commands: access-enable Create a temporary Access-List entry access-profile Apply user-profile to interface call Voice call clear Reset functions connect Open a terminal connection crypto Encryption related commands. disable Turn off privileged commands disconnect Disconnect an existing network connection enable Turn on privileged commands exit Exit from the EXEC help Description of the interactive help system lat Open a lat connection lock Lock the terminal login Log in as a particular user logout Exit from the EXEC modemui Start a modem-like user interface mrinfo Request neighbor and version information from a multicast router mstat Show statistics after multiple multicast traceroutes mtrace Trace reverse multicast path from destination to source name-connection Name an existing network connection pad Open a X.29 PAD connection ping Send echo messages ppp Start IETF Point-to-Point Protocol (PPP) release Release a resource renew Renew a resource resume Resume an active network connection rlogin Open an rlogin connection set Set system parameter (not config) show Show running system information slip Start Serial-line IP (SLIP) ssh Open a secure shell client connection systat Display information about terminal lines tclquit Quit Tool Command Language shell telnet Open a telnet connection terminal Set terminal line parameters tn3270 Open a tn3270 connection traceroute Trace route to destination tunnel Open a tunnel connection udptn Open an udptn connection where List active connections x28 Become an X.28 PAD x3 Set X.3 parameters on PAD rommon 1>
That’s entirely different and you have no capability of setting the boot parameters there. Miraculously though, you do have the “enable” command.
I’ll give you a hint, there’s no enable command in ROMMON mode! You aren’t really in ROMMON, you are just being punk’d by the router.
rommon 1>enable rommon 1>sh run | in rommon prompt "rommon 1>" rommon 1>
No matter what mode you are in, that’s the prompt. Let’s exit back out though, and assume that we’re doing things our “typical” configuration fashion.
rommon 1> rommon 1>en * * * *? ERR The "help" PAD command signal consists of the following elements: where is the identifier for the type of explanatory information requested *
What the heck is that??? Ahhhh… More entertainment. That would be the X28 Diagnostic Mode (helps if you are running an X.25 PAD, but since it’s highly unlikely that normal people today even know what that is, chances are you don’t want to run it! And yet here we are. Punk’d again.
The “exit” command will get you out. Only to be placed back to your fake ROMMON prompt! Try typing “enable” fully. (By the way, the prompt won’t change, so don’t believe everything you see!)
rommon 1>enable rommon 1>sh run | in en Current configuration : 3980 bytes no service password-encryption boot-end-marker enrollment selfsigned ip http authentication local Please change these publicly known initial credentials using SDM or the IOS CLI. alias exec en x28 end rommon 1>
Ahhhh… Aliases. Aren’t they exciting. If your proctor REALLY doesn’t like you, they’ll alias “en” and “enable” to something equally inane. But that’s a start. So let’s get rid of these things.
Don’t forget that “configure terminal” may be necessary to fully type out in case they aliased “conf” to “exit” or something fun like that!
rommon 1> rommon 1>conf t Enter configuration commands, one per line. End with CNTL/Z. exit rommon 1> *Feb 17 04:35:23.451: %SYS-5-CONFIG_I: Configured from console by console rommon 1>
Ummmm… Did someone eat the configuration mode?
rommon 1>conf t Enter configuration commands, one per line. End with CNTL/Z. interface fa0/0 ip address ? A.B.C.D IP address dhcp IP Address negotiated via DHCP pool IP Address autoconfigured from a local DHCP pool ip address % Incomplete command. ^Z rommon 1> *Feb 17 04:36:24.755: %SYS-5-CONFIG_I: Configured from console by console rommon 1>
Commands appear to work, but we can’t see anything. That would be another command!
rommon 1>conf t Enter configuration commands, one per line. End with CNTL/Z. do sh run | in config Building configuration... Current configuration : 4005 bytes no service prompt config Fix me! ^ % Invalid input detected at '^' marker. service prompt config TestRouter(config)#
Just put the service back on and we’re good. Interesting enough, the hostname shows up while in config mode. Once back in user mode, the prompt comes back.
TestRouter(config)#end rommon 1> rommon 1> rommon 1> *Feb 17 04:38:52.911: %SYS-5-CONFIG_I: Configured from console by consoleconf t Enter configuration commands, one per line. End with CNTL/Z. TestRouter(config)# TestRouter(config)# TestRouter(config)#default prompt TestRouter(config)#exit TestRouter# *Feb 17 04:39:04.775: %SYS-5-CONFIG_I: Configured from console by console TestRouter# TestRouter#
Other things you may have occur to your routers… Modification of the Exec-Timeout timers. Every 30 seconds may work without detection. Or if someone is really being amusing set “exec-timeout 0 1″ on the console port. Type one character every second or get kicked out.
This is a place for cut/paste if I’ve ever seen one!
There may be any number of other odd things appearing throughout your configurations. With a decent glance these (hopefully) will stand out like a sore thumb. Other than the show commands from the last two days to verify IP addressing and looking for basic “no” things in your config, you should do a “show run” on every device.
When you get into the lab exam, you really have no idea just how much configuration will be in place already. Just like any consulting engagement, you could have a completely blank greenfield deployment. Or you could walk into a semi-dysfunctional existing network to improve/fix/enhance throughout the day. Check out what you have to begin with. Make notes.
Things that are especially important as they may lead to future difficulties when you configure the tasks given to you:
1. Backup-interface configurations — These leave interfaces in a “standby” state which is most definitely not up!
2. Span or Remote-Span configurations — This may involve the copying of information from one port to another. While it’s one way, so OSPF peers won’t show up, RIP advertisements could!
3. “ip classless” command — This may have effects on your routing processes, or at least what is showing up versus expected!
4. kron jobs — I outlined this before regarding time-based redistribution, but anything pre-existing should be noted!
5. EEM (Embedded Event Manager) — Shouldn’t see these anyplace (or rarely!) – See below
Every once and a while, we hear ramblings from people insisting that the proctor got into their racks and changed configuration. Even if using Notepad to track your commands, or the scrollback buffer, there’s insistence that interfaces were configured one minute and had no configuration the next minute. There was no reboot, therefore it must have been the proctor.
While they do have a devlish glint in their eyes most of the time, and look like rather unsavory individuals, the Proctor’s job is not to interfere with anyone’s routers or switches. They have enough to do rather than resort to that level of torture! The Geneva Convention actually prohibits this type of behavior!
So ask yourself… If the proctor doesn’t get into my equipment… And I KNOW that I’ve configured things and they are working…. What could possibly be the cause of it? How about the last two things I mentioned above? Kron or EEM have the ability to execute commands, configure device changes and/or copy files from TFTP devices into the running config. You should be aware of what’s happening on a network at any point.
If you job is to evaluate and improve, it would be silly to rush off to do a list of tasks without understanding the impact along the way, wouldn’t it? Or what forces were working against you?
A simple check of the running configuration before hand can show these anomalies to you. Anything that looks strange needs to be investigated! Nothing worse than working through things in the middle of the day, then seeing:
Rack1R1(config)# *Feb 18 16:47:59.973: %OSPF-5-ADJCHG: Process 1, Nbr 126.96.36.199 on FastEthernet0/0 from FULL to DOWN, Neighbor Down: Interface down or detached *Feb 18 16:47:59.997: %PIM-5-NBRCHG: neighbor 188.8.131.52 DOWN on interface FastEthernet0/0 DR *Feb 18 16:48:00.009: %SYS-5-CONFIG_I: Configured from console by vty0 Rack1R1(config)# *Feb 18 16:48:05.473: %OSPF-5-ADJCHG: Process 1, Nbr 184.108.40.206 on OSPF_VL4 from FULL to DOWN, Neighbor Down: Interface down or detached Rack1R1(config)# *Feb 18 16:48:30.041: %SYS-5-CONFIG_I: Configured from console by vty0 Rack1R1(config)# Rack1R1(config)# Rack1R1(config)#
And man, I’d agree. That evil proctor just jacked my lab!
Rack1R1(config)#do sh run int fa0/0 Building configuration... Current configuration : 73 bytes ! interface FastEthernet0/0 no ip address duplex auto speed auto end Rack1R1(config)#
Well, that’s way not cool. My scrollback even tells me what I’ve done.
Rack1R1(config)#do sh run int f0/0 Building configuration... Current configuration : 115 bytes ! interface FastEthernet0/0 ip address 220.127.116.11 255.255.255.0 ip pim sparse-mode duplex auto speed auto end Rack1R1(config)#
Well, let’s put it back…
Rack1R1(config)# Rack1R1(config)#interface FastEthernet0/0 Rack1R1(config-if)# Rack1R1(config-if)# ip address 18.104.22.168 255.255.255.0 Rack1R1(config-if)# Rack1R1(config-if)# ip pim sparse-mode Rack1R1(config-if)# Rack1R1(config-if)# duplex auto Rack1R1(config-if)# Rack1R1(config-if)# speed auto Rack1R1(config-if)# Rack1R1(config-if)# Rack1R1(config-if)# *Feb 18 16:51:22.753: %PIM-5-NBRCHG: neighbor 22.214.171.124 UP on interface FastEthernet0/0 *Feb 18 16:51:22.757: %PIM-5-DRCHG: DR change from neighbor 0.0.0.0 to 126.96.36.199 on interface FastEthernet0/0 *Feb 18 16:51:27.501: %OSPF-5-ADJCHG: Process 1, Nbr 188.8.131.52 on FastEthernet0/0 from LOADING to FULL, Loading Done Rack1R1#do sh run int f0/0 *Feb 18 16:51:29.921: %OSPF-5-ADJCHG: Process 1, Nbr 184.108.40.206 on FastEthernet0/0 from FULL to DOWN, Neighbor Down: Interface down or detached *Feb 18 16:51:29.929: %PIM-5-NBRCHG: neighbor 220.127.116.11 DOWN on interface FastEthernet0/0 DR *Feb 18 16:51:30.005: %SYS-5-CONFIG_I: Configured from console by vty0 Rack1R1(config)#do sh run int f0/0 Building configuration... Current configuration : 73 bytes ! interface FastEthernet0/0 no ip address duplex auto speed auto end Rack1R1(config)#
Our neighbors come back, things are good again… But then it’s back down right away. Or it could be much later. Either way, the same frustration ensues!
Rack1R1(config)#do sh run | s event event manager applet NeverTrustTheseThings event timer watchdog time 300 action 1.0 cli command "enable" action 2.0 cli command "configure terminal" action 3.0 cli command "default interface FastEthernet0/0" Rack1R1(config)#
Well, that would certainly do it. And it may have been hidden before by a startup “logging console warnings” or something like that.
EEM is a POWERFUL tool. Check out the Network Management section of your Documentation. http://www.cisco.com/en/US/docs/ios/netmgmt/command/reference/nm_06.html#wp1157622
Rack1R1(config)#no event manager applet NeverTrustTheseThings Rack1R1(config)#do sh run | s event Rack1R1(config)#
So all during the first 30-45 minutes of your lab exam (after the Core Technology Q&A), you should be:
1. Reading through the whole exam
2. Taking notes to remind yourself of things later
3. Redraw the diagram quickly so you can write on it
4. Verify IP addresses quite simply
5. Identify major things in the configs beginning with “no” or altering the configuration register
6. Quickly do “show run” on all devices and scan through for anything that looks strange or out of place
7. Get ready to kick butt and take your number home!
We get stuck in ruts when going through practice labs. We have a process generally dictated by what level of preparation that we’ve done. While there are certainly a good number of tools and labs and documents out there to help you study, there is nothing that compares to the “personalized approach”. By that, I mean, make changes yourself. Or have a buddy make some additional tweaks, tasks, changes to labs for you.
Keep in mind that folks on the CCIE team probably have EVERYONE’s study materials. So while we are indeed pretty cool, they’ll go out of their way to find something we didn’t think of. So outsmart them! Process, process, process.
Good luck in your lab prep, and most certainly in the troubleshooting portion! Don’t forget, this addresses nothing about the self-induced troubleshooting. While going through your lab, you should be verifying things every step of the way with show or debug (or whatever) commands. If properly verified, you will have few, if any, surprises during your lab.
10 Responses to “CCIE Troubleshooting: Part 3 (Ummmm…. Houston, We Have a Problem)”
Leave a Reply