It begins.

rakem

Yep that makes sence... thanks guys.

rakem

Off work for a few days due to wisdom teeth being pulled out so trying to do some lab work.

Been working on different ways to assign local_pref, weight using different styles of route-maps.

Using prefix-lists and route-maps to manipulate the as_path length.

rakem

Played around using MED to influence routes.

A few key points is that routers will by default only compare MED if the AS-path is the same for the routes being compared. This can be changed by using bgp always-compare med command.

MED is propagated between iBGP peers but no other eBGP peers.

jason_lunde

rakem wrote: »

Played around using MED to influence routes.

A few key points is that routers will by default only compare MED if the AS-path is the same for the routes being compared. This can be changed by using bgp always-compare med command.

MED is propagated between iBGP peers but no other eBGP peers.

Good point. If you want to look a little more into the intricacies of this command, along with it's partner command look at this doc:
How the bgp deterministic-med Command Differs from the bgp always-compare-med Command [IP Routing] - Cisco Systems
BGP deterministic-med is a big one to know too. Like your thread man, keep it up.

rakem

jason_lunde wrote: »

If you want to look a little more into the intricacies of this command, along with it's partner command look at this doc:
How the bgp deterministic-med Command Differs from the bgp always-compare-med Command [IP Routing] - Cisco Systems

That's a good link, thanks.

Been doing some reading tonight from the CCIE Study Guide, BGP Chapter. That book just hits you with fact after fact after fact. I have found myself opening up TCP/IP routing Volume 2 for a better or more detailed explanation than what the study guide has to offer.

Still a good book thou, I like how it follows up its theory with a practical example.

rakem

More BGP reading and note taking tonight. Only an hour or so... big day at work.

rakem

No study last night... band practice.... got a show on saturday night.

rakem

Up early for some study before a day of house hunting and a gig tonight!

rakem

So been playing around with redistribution from eigrp into BGP, and checking out the MED behaviour, especially when the 'missing-as-worst' feature is used.

So BGP will advertise a prefix to an eBGP peer with eigrps FD as the metric.
Prefixes that are on a directly connected interface are advertised with a metric of 0
Aggregrated routes are advertised with a metric of 4294967295

Can't find much info on how MED reacts when summaries, redistribution and other funky stuff occurs.... interesting though.

rakem

not feeling like doing to much reading or anything in general today after a pretty big night last night.

Ended up watching a few CBT nuggets on BGP. The vids brush over most topics at a fairly high level i think they are still a good source of info, since Jeremy knows his stuff pretty well there are always a few little bits of gold that you can pick up.

For example, i didn't know this command existed...

R1#show tcp brief all
TCB       Local Address               Foreign Address             (state)
673B9050  10.123.1.1.179              10.123.1.2.49375            ESTAB
6633C654  10.123.5.1.18078            10.123.5.6.179              SYNSENT
6768B5D0  10.123.3.1.179              10.123.3.4.15849            ESTAB
6768AE1C  *.443                       *.*                         LISTEN
6768DA70  *.80                        *.*                         LISTEN
6768BB24  *.*                         *.*                         LISTEN
6761BFD8  *.179                       10.123.5.6.*                LISTEN
6761B54C  *.179                       10.123.4.5.*                LISTEN
6761A844  *.179                       10.123.3.4.*                LISTEN

Shows the TCP sessions currently active on the router.... handy.

rakem

Currently working through the BGP case study exercises from TCP/IP Routing V2. There are heaps to get through but this should help cement in all the theory.

rakem

no study tonight. spent all day in the datacentre and am wrecked.

deth1k

Why don't you dive into CCIP while you at it, will help towards you IE no doubt.

rakem

deth1k wrote: »

Why don't you dive into CCIP while you at it, will help towards you IE no doubt.

Haven't looked into that at all.. From the very brief look i just had, its basically the CCNP + BGP, MPLS and QoS exams.... is that correct?

Perhaps if i'm struggling to pass the CCIE written i will look into one of these to keep my CCNP valid.

rakem

BGP Communities.....
Community lists....
Route maps matching community lists...
Using community to set local pref....

Its Friday night! bah.

Haven't got much done this week. Work was pretty busy. Also in the process of buying my first home so my mind has been elsewhere....

There is so much to BGP.. it just keeps coming.

jason_lunde

rakem wrote: »

There is so much to BGP.. it just keeps coming.

Ya, every time I thought I had seen everything to do with BGP more kept appearing. Still does honestly.

rakem

Snuck in an hour or so of study this morning before i head out looking at houses again.

Still running through the case studies from TCP/IP routing. This morning was the aggregate routes case study (which goes on for about 16 pages!)

Some good stuff in there about using suppress-maps, advertise-maps, good old route-maps.

rakem

Got a good 4 or so hours of study in today. Think i'll spend another week or so on BGP then might move on the MPLS.

On Page 457 of the CCIE R&S Study guide it has a mnemonic for remembering the BGP decision making process

N
WLLA
OMNI

I though that was rubbish so i made up my own.

Never (Next hop reachable)
Walk (Weight)
Lonely (Local Pref)
Ladies (locally injected preferred over i/eBGP learned)
Around (AS_path length)
Ontario (Origin I > EGP > ?)
Monday ( MED, smallest is winner)
Night (Neighbour type Ebgp > iBGP)
Included ( IGP metric to next hop smaller is better)

hono

Hello,

I study with the same book,

It's :
N
WLLA
OMNI

Rgds

Turgon

rakem wrote: »

Got a good 4 or so hours of study in today. Think i'll spend another week or so on BGP then might move on the MPLS.

On Page 457 of the CCIE R&S Study guide it has a mnemonic for remembering the BGP decision making process

N
WALLA
OMNI

I though that was rubbish so i made up my own.

Never (Next hop reachable)
Walk (Weight)
Lonely (Local Pref)
Ladies (locally injected preferred over i/eBGP learned)
Around (AS_path length)
Ontario (Origin I > EGP > ?)
Monday ( MED, smallest is winner)
Night (Neighbour type Ebgp > iBGP)
Included ( IGP metric to next hop smaller is better)

The hours are ramping up on the study front. Good. 50 is a milestone! Hit it as soon as you can.

Good luck!

rakem

Thanks mate.

I have just purchased my first home as well so some exciting stuff is happening outside of work/cisco.

On another note, what does everyone use for practise questions? I installed the boson stuff that came with the CCIE cert guide, it said on the back of the book that you can focus your study on particular topics but i can't find how to do that.

What else is useful for CCIE level theory practise questions?

rakem

Tonight i went through the BGP multiple choice questions in the CCIE cert guide. Getting around 80% correct at the moment. So not bad, but need to brush up on a few things.

Also did some reading on prefix-lists since i had no idea what the ge and le commands meant... Its pretty simple really

something like:

ip prefix-list permit 10.0.0.0/8 ge 9 le 20


would match anything subnet starting with 10. that has a prefix length (subnet mask) between 9 and 20.

Turgon

rakem wrote: »

Thanks mate.

I have just purchased my first home as well so some exciting stuff is happening outside of work/cisco.

On another note, what does everyone use for practise questions? I installed the boson stuff that came with the CCIE cert guide, it said on the back of the book that you can focus your study on particular topics but i can't find how to do that.

What else is useful for CCIE level theory practise questions?

The Boson test that comes with the book allows you to just test on particular topics and you should do that. Concentrate on reading one chapter each day followed by the end of chapter questions. I think the Boson test with the book has those on the CD.

Boson also sell a test prep for the written which I found useful. For the moment though concentrate on reading the book patiently.

Turgon

rakem wrote: »

Tonight i went through the BGP multiple choice questions in the CCIE cert guide. Getting around 80% correct at the moment. So not bad, but need to brush up on a few things.

Also did some reading on prefix-lists since i had no idea what the ge and le commands meant... Its pretty simple really

something like:

ip prefix-list permit 10.0.0.0/8 ge 9 le 20


would match anything subnet starting with 10. that has a prefix length (subnet mask) between 9 and 20.

Is that not greater than or equal to 9? So that includes prefix length 9 and 20

You will find there are lots of things in written prep that you have no idea about. You need mastery of filtering and metric mechanisms so you can influence best path selection for many protocols. The lab tests this greatly.

rakem

Turgon wrote: »

Is that not greater than or equal to 9? So that includes prefix length 9 and 20

My understanding was greater than 9 and less that 20... so everything inbetween. Will look into it.

Had a slow week on the study, and have done none this weekend. Will pick back up again tomorrow.

rakem

Went through the BGP Community case study in Routing TCP/IP Volume 2. Quite enjoyed it.

Started getting into the Confederation case study which also looks good, but I'm to tired now so will hit it up tomorrow.

rakem

This link explains the ip prefix commands quite well...


IP Prefix List Operators In A Nutshell CCIE Pursuit Blog

rakem

Good session tonight. Finished off a confederation lab i was working with. Had a hell of a time getting routes to use the best MED, they kept using a sub-optimal route....

There were two routes with the same MED (50) to a destination, but the router was choosing the longer path (The longer confederation path(as_confed_seq is not considered as part of the decision making process)).

Turned out it was because the IGP i was using (EIGRP) had a lower metric for the sub-optimal routes next hop. Using the next-hop-self command on the advertising router fixed it up but it took me a while to figure that one out.

Hopefully will be done with BGP by this weekend. I have about 12 pages of notes that I have taken so I will consolidate them into a **** sheet then move on.

Turgon

rakem wrote: »

My understanding was greater than 9 and less that 20... so everything inbetween. Will look into it.

Had a slow week on the study, and have done none this weekend. Will pick back up again tomorrow.

Odom page 322 seems to clear it up Table 11-5

10.0.0.0/8 ge 9 - "The prefix length muct be between 9 and 32, inclusive"

So I take inclusive to include 9 and 32.

Ahriakin

Yup I believe the 'ge' is for Greater than or Equal.