Jump to content






Photo - - - - -

CCIE RnS Lab Strategies

Posted by asasel , 06 December 2011 · 1227 views

CCIE RnS

Looking at the many CCIE vendors out there I find the netmasterclass method
of approaching the lab most closely resembles how I see it.

phase 0
Check your ip addressing and netmask and correct any error.

phase 1
Next check/build your layer 2 such to ensure that devices which must have direct
connectivity can indeed ping each other
(ppoe/ppp/vlans/l3-etherchannels/frame-relay/qinq/rspan).

phase 2
Next you have a couple of IGPs in your lab. Note which interfaces must be participating
in each IGP and be as precise as possible. Configure the IGP's one by one.
For each IGP configured check that on each participating device you see the expected
routes with show ip route "The-IGP" command. Build your TCL script for each IGP and test
the expected reachability in each IGP. If you have overlooked anything within an IGP
(eg split-horizon/adjusting interface network type/no auto-summary/virtual-link etc) it must be
apparent here because your TCL script will not run succesfully. Fix any IGP issue
(typo/overlooked feature etc) in this stage and rerurn the TCL script for the problem
IGP's.

phase 3
At this point administration is even more important. It is not always possible at this
point to test full reachability because there maybe some none participating interfaces
which at this point are not reachable and which later on will be reachable through for
instance the MBGP/MPLS combo.
The important step in this phase is to take care not to try to fix things which ain't
broken!
In phase 2 you already know how many participating interfaces you have per IGP.
A special case is the backbone router which prefixes will (must) be treated as external..
This may happen by setting an EBGP relationship or by setting two IGP's, one for the
backbone router and one for the internal network and next doing a mutual redistribution
between the two.
It is implied that you must have inter-IGP connectivity in your POD, but it is not
implied that all backbone prefixes must be reachable from all internal devices.
So this is the point where you have to read even more carefully so not to make things
more complex than they should be.
Once you have deducted which prefixes in this phase must be reachable from each device
in your pod, you can use a TCL script to test the reachability of each from every single
device.
The method to glue the information together is through mutual redistribution.
Know when a prefix will be redistributed into another IGP and when not and also know
when loops may occur and when not and also know how to prevent loops in the most
elegant way as possible.

phase 4
In this phase you will be doing the last reachability work in your pod. It could happen
that there are still some prefixes which must be reachable from certain points in your
network, probably through a MPLS VPN construction... or through setting EBGP connections
to some backbone routers or maybe even through GRE tunnels. When completed your next step
is to conduct the full reachability test as defined/specified by your lab requirements.
Use a TCL script to test this on each device in your pod.

phase 5
Do steps 0 through 5 but now for IPv6.

phase 6
Having L2/IGP/BGP/MPLS/IPv6 done and verified, you have now a base to build some other
stuff above.. Multicast being the obvious one.

phase 7
Be very precise with QoS and only manage the traffic which needs to be manipulated..
so make your selection criteria very precise.

phase 8
Be very precise with Security and only manage the traffic which need to be filtered..
This is the step which can make you fail if not done right. Be vey carefull with the
implied filtering! .. It is best to do security in a very early stage so that you can
have another reassurance that what you will be building afterwards will not be broken
by some kind of mistakes in filtering traffic.

phase 9
In this phase you have done and verified L2/IGP/BGP/MPLS/IPv6/Multicast/QoS/Security and
you should be ready to move on to the superficial stuff.

I believe that Cisco would do a great job if the CCIE lab test is centered on these
technologies and a candidate would have enough chance to pass the exam if he is proficient
in the 8 above areas.. meaning that the total point up to then should be at least 90.




agree  100%

Mysterioussyed
Dec 09 2011 07:24 PM
Thanks for sharing
great my friend!! I appreciate your help.
i appreciate it, thanks for sharing :)
Great structured approach,did you passed your lab with this approach ?
thank you mate
I learned a lot...
Thanks

October 2014

S M T W T F S
   1234
567891011
12131415161718
192021 22 232425
262728293031 

Recent Entries

Recent Comments

Tags

    Search My Blog

    Organization

    Community

    Downloads

    Test Providers

    Site Info


    Go to top