Wednesday, February 25, 2009

OSPF Fast Hello Packets - Hmm whats that ?

I know that many or should say most of people in Networking world knows about basics of OSPF hello packets and what they are meant for. I guess I came to know about ospf hello packets first during my ccna studies . But later during my ccnp studies I learned some more details of it like what information is carried within ospf hello packet and all. So there is nothing new in that.
 
But an year back when I started preparing for my CCIE certification , I came to know about interesting new IOS feature called OSPF Fast Hello Packets.  This feature was basically introduced to enhance fast convergence capabilities of ospf protocol. So lets discuss about it:

OSPF hello packets are packets that an OSPF process sends to its OSPF neighbors to maintain
connectivity with those neighbors. The hello packets are sent at a configurable interval (in seconds). The defaults are 10 seconds for an Ethernet link and 30 seconds for a non broadcast link. Hello packets include a list of all neighbors for which a hello packet has been received within the dead interval. The dead interval is also a configurable interval (in seconds), and defaults to four times the value of the hello interval. The value of all hello intervals must be the same within a network. Likewise, the value of all dead intervals must be the same within a network.
These two intervals work together to maintain connectivity by indicating that the link is operational. If a router does not receive a hello packet from a neighbor within the dead interval, it will declare that neighbor to be down.

OSPF fast hello packets refer to hello packets being sent at intervals of less than 1 second.

OSPF fast hello packets are achieved by using the ip ospf dead-interval command. The dead interval is set to 1 second, and the hello-multiplier value is set to the number of hello packets you want sent during that 1 second, thus providing sub second or “fast” hello packets.
When fast hello packets are configured on the interface, the hello interval advertised in the hello packets that are sent out this interface is set to 0. The hello interval in the hello packets received over this interface is ignored.
The dead interval must be consistent on a segment, whether it is set to 1 second (for fast hello packets) or set to any other value. The hello multiplier need not be the same for the entire segment as long as at least one hello packet is sent within the dead interval.

The benefit of the OSPF Fast Hello Packets feature is that your OSPF network will experience faster convergence time than it would without fast hello packets. This feature allows you to detect lost neighbors within 1 second. It is especially useful in LAN segments, where neighbor loss might not be detected by the Open System Interconnection (OSI) physical layer and data-link layer.

Example:
Router(config-if)# ip ospf dead-interval minimal hello-multiplier 5

Because the multiplier is set to 5, five hello packets will be sent
every second. That means one hello packet in every 200 milliseconds

Hmmm...lets verify this:

Router# show ip ospf interface ethernet 1/3
Ethernet1/3 is up, line protocol is up
Internet Address 172.16.1.2/24, Area 0
Process ID 1, Router ID 172.17.0.2, Network Type BROADCAST, Cost:1
Transmit Delay is 1 sec, State DR, Priority 1
Designated Router (ID) 172.17.0.2, Interface address 172.16.1.2
Backup Designated router (ID) 172.16.0.1, Interface address 172.16.1.1
Timer intervals configured, Hello 200 msec, Dead 1, Wait 1, Retransmit 5

Best Regards,
Deepak Arora

Tuesday, February 24, 2009

Some More Funny RFCs - gr8 collection :)

Local Proxy ARP - hmmm it's not ordinary ip proxy ARP

The local proxy ARP feature allows the Multilayer Switching Feature Card (MSFC) to respond to ARP requests for IP addresses within a subnet where normally no routing is required. With the local proxy ARP feature enabled, the MSFC responds to all ARP requests for IP addresses within the subnet and forwards all traffic between hosts in the subnet. Use this feature only on subnets where hosts are intentionally prevented from communicating directly to the Catalyst 6500 series switch on which they are connected.
Before the local proxy ARP feature can be used, the IP proxy ARP feature must be enabled. The IP proxy ARP feature is enabled by default.
Internet Control Message Protocol (ICMP) redirects are disabled on interfaces where the local proxy ARP feature is enabled.
Examples
The following example shows how to enable the local proxy ARP feature:
Router(config-if)# ip local-proxy-arp

Best Regards,
Deepak Arora

Wednesday, February 18, 2009

CCIE Certification - Few Words

The CCIE is Cisco's expert-level certification. Created in 1993, the CCIE program was the first certification program developed by Cisco. To achieve CCIE status you first must pass a comprehensive written qualification exam. A passing score on the qualification exam entitles you to register for the true challenge: an intensive, hands-on lab exam conducted at one of several Cisco locations around the world. The lab exam lasts a full day and will test your ability to quickly configure and troubleshoot a complex internetwork to exacting specifications using a variety of Cisco hardware components.

Monday, February 16, 2009

Simple but Interesting NAT question - Good to know for interview :-)

Around an year back I have been asked a question in interview - hey If a packet is coming to router for both NAT and Routing purpose than what will router do first - Routing & then Natting or Natting then Routing ?

Hmmm...interesting one...Isn't it

I said Routing first and then Natting.
Let me explain why...See if a packet is coming to router on which it has to perform both routing and natting functions than in that case routing function will take precedence. Reason is if a packet is not supposed to be forwarded after routing check than why would router perform Nat function first and later will drop packet because no routing is found for destination subnets. I mean if router will perform Natting first than we are just increasing it's over head. Simple if we are not going to route a packet than we need not to perform Natting on it first.

Best Regards,
Deepak Arora