"Why do we need VPNs?"
We need VPNs because enterprises need private connectivity, and they need that over a shared infrastructure to minimize their costs.
"Why do we need MPLS?"
This is because MPLS offers a good infrastructure for converged networks. Rather than using passwords, it offers hierarchy, data privacy, traffic engineering, and differentiated services.
"Why do we need BGP?"
We need BGP because it offers a common framework for providing all VPN types--Layer 2 VPNs, IP VPNs, and VPLS. In addition, it can do auto-discovery, VPN label exchange, and interprovider and carrier-to-carrier VPNs.
Why do we need all these kinds of VPNs?
We need Layer 2 VPNs because of their legacy; they are there and have been for a long time.
We need IP VPNs because traffic is increasingly all IP, and any-to-any connectivity has its benefits. Moreover, outsourcing routing can be a big cost saver.And we need VPLS, as this is an exciting new service that takes advantage of an exciting new medium, Ethernet.
Monday, February 23, 2009
Debugging my fault
I had taken a sat at JNCIP exam before. Its a remote lab, telnet-ing to physical lab in Amsterdam.
Its an 8-hour practical exam excluding one hour lunch time, start from 10 AM to 7 PM including one hour lunch time. I had build an ISP consisting of seven M-series routers and multiple EBGP neighbors. All 7 router's IGP, BGP route reflector and EBGP neighbors had been reach up. And then i jump into IGP/BGP customization task and routing policy, but i saw the clock already 5 PM, i hurried up finish my task, and by 7 PM i feel i had meet all the requirements but i did not had time to check and verified the configuration.
15 days after the exam, Gary Hausser, who is proctoring me send me an email, he said i failed the exam and i lost big points in IGP/BGP customization task and routing policy. I feel very sad to had receiving email like that. :( i realized that i had no enough time to verified again my configuration at that area. Its because i didn't wisely allocate my time wisely, the pitfall was my battle against unfamiliar interface at beginning of exam taking 3 hour to conquered them.
Debugging my fault
After receiving email from Gary, i check again configuration my m7i's logical router, yeah i found a lot of mistake because i didn't re-read the requirements, i just follow my assumption not following what the juniper want.
Now, i preparing myself to taking another exam by trying bunch of exam scenario and a lot of what if question...
Wish me luck okay :D
Its an 8-hour practical exam excluding one hour lunch time, start from 10 AM to 7 PM including one hour lunch time. I had build an ISP consisting of seven M-series routers and multiple EBGP neighbors. All 7 router's IGP, BGP route reflector and EBGP neighbors had been reach up. And then i jump into IGP/BGP customization task and routing policy, but i saw the clock already 5 PM, i hurried up finish my task, and by 7 PM i feel i had meet all the requirements but i did not had time to check and verified the configuration.
15 days after the exam, Gary Hausser, who is proctoring me send me an email, he said i failed the exam and i lost big points in IGP/BGP customization task and routing policy. I feel very sad to had receiving email like that. :( i realized that i had no enough time to verified again my configuration at that area. Its because i didn't wisely allocate my time wisely, the pitfall was my battle against unfamiliar interface at beginning of exam taking 3 hour to conquered them.
Debugging my fault
After receiving email from Gary, i check again configuration my m7i's logical router, yeah i found a lot of mistake because i didn't re-read the requirements, i just follow my assumption not following what the juniper want.
Now, i preparing myself to taking another exam by trying bunch of exam scenario and a lot of what if question...
Wish me luck okay :D
Thursday, February 05, 2009
show version and haiku
IS-IS screams,
BGP peers are flapping:
I want my mommy!
IS-IS sleeps.
BGP peers are quiet.
Something must be wrong.
Look, mama, no hands!
Only one finger typing.
Easy: commit scripts.
TTL down one
the end nearer with each hop
little packet, poof.
3am; darkness;
Maintenance window closing.
Safety net: rollback.
Weeks of studying,
Days of lab exercises:
JNCIE.
Juniper babies
The next generation starts
Gotta get more sleep
BGP peers are flapping:
I want my mommy!
IS-IS sleeps.
BGP peers are quiet.
Something must be wrong.
Look, mama, no hands!
Only one finger typing.
Easy: commit scripts.
TTL down one
the end nearer with each hop
little packet, poof.
3am; darkness;
Maintenance window closing.
Safety net: rollback.
Weeks of studying,
Days of lab exercises:
JNCIE.
Juniper babies
The next generation starts
Gotta get more sleep
Tuesday, February 03, 2009
General Networking Troubleshooting Tips
- You must know what is "NORMAL" for your system
- Start with a visual inspection
- A divide-and-conquer approach is ideal when multiple fault can lead to a common symptom
- Failure hypotheses should be testable-be definitive about parameter of the test
- Please do not blinded by subjectivity-open your eyes and mind
- Layered troubleshooting approach
- modern communication networks are modeled around architecture layer
- matching a symptom to the root-caused layer is a critical step in rapid diagnosis and restoration
- identifying the specific fault within the root-cause layer is icing on the cake!
Subscribe to:
Posts (Atom)