It begins.

PsychoFin

Congrats on all the certs, the marriage and enjoy your honeymoon!

Cheers,
Fin

rakem

Passed my final F5 exam yesterday.
I have now got all the certs that my work wanted me to get so can seriously get back to CCIE study once i'm back from the Honeymoon.

Going to start some light reading again soon, aim to have written passed by early October.

Essendon

Nicely done, enjoy the break!

ITdude

Congrats and enjoy the honeymoon too!

CCIEWANNABE

rakem wrote: »

Yep 4 weeks.... Alaska, then Hawaii

Nice, Hope you had fun! I was in Alaska for almost 4 years, was a great place. I was stationed at Elmendorf AFB right outside of Anchorage. Snowboarding, Fishing, Ice Hockey and drinking lots of homebrew = Fun Times

rakem

Back from honeymoon, which was awesome.

Will start up the study soon.

rakem

I'm back!

Last few months have been hectic. Been working 60 to 70 hour weeks for the last 4 months to get our new DCs up and running. No energy or time to study.

That project is finished and i'm back to my normal role. Written date set for 1st November.

rakem

Two hours of QoS review tonight, just the basic theory behind ToS, IPP, CoS, and DSCP.

Turgon

rakem wrote: »

I'm back!

Last few months have been hectic. Been working 60 to 70 hour weeks for the last 4 months to get our new DCs up and running. No energy or time to study.

That project is finished and i'm back to my normal role. Written date set for 1st November.

hehehe..that's what they pay you for. I know the feeling! Glad you are back on the study horse now.

rakem

Went over CBWFQ, LLQ and WRED tonight.

Some key points:

bandwidth remaining percent command - allocates the remaining bandwidth to the class. Remaining bandwidth is a product of the interface configured bandwidth (using the bandwidth statement) and the max-reserved-bandwidth (which by default is 75%)
So if you have a 128Kb link and leave max-reserved-bandwidth at is default, and use bandwidth remaining precent 50 you would be allocating 48Kb.

Since 75% of 128Kb = 96. And 50% of 96 = 48.

The bandwidth percent command just allocates a percentage of the configured interface bandwidth. It still can't exceed the max-reserved-bandwidth value though.

WRED - Random drops based on IPP or DSCP (configurable)
minimum threshold - when will dropping begin
maximum threshold - at what point will 'full drop' start
mark probability denominator - how aggressed will WRED drop packets. e.g. if configured to 7 - 1 in every 7 packets will be randomly dropped.

This is all configurable per DSCP or IPP value as well.

So lots of reading - this stuff is all just revision but its good to refresh. The CCIE cert. guide is a bit light is some of these areas so i also use End to End QoS Network design by Tim Szigeti (Cisco press) its a bit old but still got some good info.

rakem

maybe its late and i'm missing something by can someone explain how CS3 = deciamal 24.

I know the formula of 8x + 2y = decimal. But thats only for AF values. So is there something similar for CS values?

argh. late night study is tough sometimes.

reaper81

6 bits are used for DSCP so the values can be from 0 to 63. On the right hand side will be the LSB (Least Significant Bit) and then move towards the MSB (Most Significant Bit) just like in any binary.

32 16 8 4 2 1 are the values. So now lets look at CS 3 (IP precedence 3)
0 1 1 0 0 0

011 is IP precedence 3 and CS3 corresponds to that. So we have 16+8 = 24. If you want to find the decimal value for EF.

101110 which is 46 in decimal. AF 33 is 011110 which is 30 in decimal and so on.

Turgon

rakem wrote: »

Went over CBWFQ, LLQ and WRED tonight.

Some key points:

bandwidth remaining percent command - allocates the remaining bandwidth to the class. Remaining bandwidth is a product of the interface configured bandwidth (using the bandwidth statement) and the max-reserved-bandwidth (which by default is 75%)
So if you have a 128Kb link and leave max-reserved-bandwidth at is default, and use bandwidth remaining precent 50 you would be allocating 48Kb.

Since 75% of 128Kb = 96. And 50% of 96 = 48.

The bandwidth percent command just allocates a percentage of the configured interface bandwidth. It still can't exceed the max-reserved-bandwidth value though.

WRED - Random drops based on IPP or DSCP (configurable)
minimum threshold - when will dropping begin
maximum threshold - at what point will 'full drop' start
mark probability denominator - how aggressed will WRED drop packets. e.g. if configured to 7 - 1 in every 7 packets will be randomly dropped.

This is all configurable per DSCP or IPP value as well.

So lots of reading - this stuff is all just revision but its good to refresh. The CCIE cert. guide is a bit light is some of these areas so i also use End to End QoS Network design by Tim Szigeti (Cisco press) its a bit old but still got some good info.

Going well. Get that counter over 100 hours for reading as soon as you can. I think Jimmy just crossed that milestone on his thread.

It means you are no longer beatable on reading over a two weekend mammoth reading spree by other candidates. They do try! Take your time as you have done, but read daily, the hours do add up!

48 hours reading without food, sleep, or use of the toilet is one weekend. Times two is 96 hours. It can't be done anyway and it means you can't be surpassed by a two weekend crammer. Many try to do this. It is foolish

jamesp1983

rakem wrote: »

Went over CBWFQ, LLQ and WRED tonight.

Some key points:

bandwidth remaining percent command - allocates the remaining bandwidth to the class. Remaining bandwidth is a product of the interface configured bandwidth (using the bandwidth statement) and the max-reserved-bandwidth (which by default is 75%)
So if you have a 128Kb link and leave max-reserved-bandwidth at is default, and use bandwidth remaining precent 50 you would be allocating 48Kb.

Since 75% of 128Kb = 96. And 50% of 96 = 48.

The bandwidth percent command just allocates a percentage of the configured interface bandwidth. It still can't exceed the max-reserved-bandwidth value though.

WRED - Random drops based on IPP or DSCP (configurable)
minimum threshold - when will dropping begin
maximum threshold - at what point will 'full drop' start
mark probability denominator - how aggressed will WRED drop packets. e.g. if configured to 7 - 1 in every 7 packets will be randomly dropped.

This is all configurable per DSCP or IPP value as well.

So lots of reading - this stuff is all just revision but its good to refresh. The CCIE cert. guide is a bit light is some of these areas so i also use End to End QoS Network design by Tim Szigeti (Cisco press) its a bit old but still got some good info.

I couldn't agree with you more. This book certainly is lighter than the QoS Exam Guide. Good luck with your studies!

rakem

Spend last night brushing up on Token buckets - Bc, Be, Tc.
bit of LLQ, and LAN QoS (SRR)

Turgon

rakem wrote: »

Spend last night brushing up on Token buckets - Bc, Be, Tc.
bit of LLQ, and LAN QoS (SRR)

Good stuff. Remember the formula. Particular attention to dual rate stuff, PIR etc. Also make a note of default Tc i.e .125ms and that some parameters are in bps and some in bytes. Also how to calculate Be.

rakem

Turgon wrote: »

Remember the formula

Tc = Bc/CIR - thats a key one.

its annoying how you need to configure some in Kb and other in bits. Where's the standard!!

jamesp1983

Any studying today?

Turgon

Its the weekend. Study hard.

rakem

yes plenty of reading this weekend. Heading out with the Mrs now for a bit. Plan to get at least 2 more hours in tonight.

Mainly been reviewing all the QoS stuff i have been reading for the past week.
Spent a bit of time this morning playing with QoS marking/shaping/policing in my lab.

Edit - new forum layout is cool!
Took me a while to figure out how to edit my signature thou!

rakem

Just another hour or so tonight. Moving on today L3 protocols tomorrow.

Turgon

Cool. Keep pumping and get upto 100 reading hours/50 lab hours. What you think you know now you will start to make more sense of by then. A lot of stuff to take in!

rakem

An hour or so of EIGRP tonight. Actually been a long time since i have looked at EIGRP so this is a good refresh.

The CCIE study guide chapter on EIGRP is very light and brushes over so much.
Will be spending most of my time in Routing TCP/IP Volume 1 SE

Turgon

Crossed 50 reading hours I see. Well done. Unbeatable by a weekend of cramming.

rakem

Day off from work today as i'm not feeling that well.
Managed 2 hours of more EIGRP review. Resting now and will come back to it tonight.

Turgon

Sorry to hear that. Get well soon and cross 100. We have a few candidates now that are over that number. There appears to have been a resurge in interest in CCIE prep and we have some solid candidates on the boards.

rakem

Had a frustrating time tonight with EIGRP in my GNS3 lab.
Not sure if it was something wierd with GNS3 or with the IOS version (12.4(11)) but i had to clear the eigrp process whenever i made a change to the a metric. Otherwise the routers would not update. I will run the same lab on my physical gear to see what happens.

key EIGRP stuff:

Reported distance - Metric of a route reported by neighbour
Feasible Distance - This routres best metric
Successor route - route with best FD
Feasible successor - RD must be lower than this routers FD.

bandwidth and delay are default metrics. (K values)
K1- Bandwidth
K2- Load
k3 -Delay
K4 - Reliability
K5 - MTU

Can be turned on or off using 'metric weights k1 k2 k3 k4 k5' command in router config mode.
K values must match to become neighbours.

Routers only run DUAL when there is no FS in the topology table

All queries sent to out must be replied to within 3 minutes (default) or route will be marked stuck in active (SIA)
successor is not chosen until all replies from queries are received.

Turgon

Not my favorite protocol. I often forget the metric calcs and roles. A couple of killer questions on the written. Learn it well.

rakem

I don't know why Cisco persist with it.... OSPF is so much better.

jovan88

Good to see a CCIE candidate representing Sydney! Good luck