Anthony Fajri

Misc

LDP: 646 (TCP/UDP)
TDP: 711 (TCP/UDP)
BGP: 179 TCP
MSDP: 639

MSDP: RFC 3618
Private IP: RFC 1918

RSVP: IP Protocol 46
GRE: IP Protocol 47

OSPF: 224.0.0.5 (hello) & 224.0.0.6
RIPv2: 224.0.0.9
EIGRP: 224.0.0.10
Multicast: 224.0.0.0/4
Administrative Scope: 239.0.0.0 – 239.255.255.255

RFC1918: 10.0.0.0/8, 172.16.0.0/12, 192.168.00/16

ATM: MTU 4470
Dot1Q: 4 bytes overheads
PPPoE: MTU 1492

IPv4 = 32bit
IPv6 = 128bit
RD = 64bit
VPNv4 = RD+IPv4
MPLS Header = 32bit (20 bit label, 3 bit exp qos, 1 s bit, 8 bit ttl)
TLV: 6 bit (24 bit for metric-style wide)

eBGP: AD 20
EIGRP: AD 90 (internal) & 170 (external)
OSPF: AD 110
ISIS: AD 115
RIP: AD 120
iBGP: AD 200

ISIS Auth: interface (hello) & ISIS Process (LSP, CNSP, PSNP)
OSPF Auth: null, clear text, md5

ISIS: L2->L1 need redistribute, L1->L2 automatic
O : LSA 1 & 2
O IA: LSA 3:
O E1 & O E2: LSA 5
Multicast: LSA 6
NSSA: LSA 7
Opaque (for TE): LSA 10

Inbound: route-map, filter-list, prefix-list
outbound: prefix-list, filter-list, route-map

BGP Well-known Communities: no-export (local-as only), no-advertise (local only), internet, none
BGP communities: 32 bit (AA:BB – 16 bit each)
extended communities for MP-BGP: RT & SoO
aggregation: summary-only (suppress all specific routes), suppress-map (suppress some specific routes), attribute-map (set bgp parameter on summary), advertise-map (conditional)

AToM: ethernet, ATM AAL5, FR, ATM Cell Relay, PPP, HDLC, Circuit Emulation

Forwarding traffic to TE: static, PBR, autoroute
FRR Protection: Path & Node
tail and head have to be in the same area

PE-CE routing protocol: RIPv2, EIGRP, eBGP, OSPF, ISIS, static
OSPF redistribution:
- default E2
- E1: configure metric-type
- IA: check (or configure) domain-id
- O: sham-link

ORF: outbound route filtering (to save memory), need to be enabled both side

Inter-AS VPN:
option 1: back to back (no tagging in between ASBR)
option 2: MP-eBGP in between ASBR
option 3: eBGP in between ASBR, MP-eBGP in between RR

February 23rd, 2010 Posted by Anthony Fajri| information technology | no comments

Traffic Shapping in Frame Relay

Imagine you have a router connected to ISP’s network. and you are using frame relay with the following specification:

the physical connection can support up to 256 kbps.

Your ISP is provisioning 128 kbps in average.

However, the ISP guarantee that the throughput will not lower than 32 kbps

To maintain the average bandwidth, every 50 millisecond the router send Bc bits

Then we have the following graph as illustration:

(picture taken from: http://linuxbrit.co.uk)

Based on the spefication above, we have some parameter:

CIR = 128 kbps

MINCIR = 32 kbps

access rate = 256 kbps

Tc = 50 ms

ps: in above illustration and picture, the maximum traffic should be CIR. but for some case, there are some traffic that exceed the provisioning bandwidth. I can not get a good picture to ilustrate that. The thoughput which reach access rate should be spike only.

Then we can calculate Bc and Be

Bc = Tc * Cir / 1000 = 6400

Be = (AR – CIR) * TC / 1000 = 6400

Bc is the amount of traffic which sent during Tc period.
Be is the amount of traffic which exceed the provisioning bandwidth during Tc Period.

For some cases, Bc is zero, which mean that the CIR is the same value as access rate. this is the ideal case.

June 7th, 2008 Posted by Anthony Fajri| information technology | no comments

Controlling BGP Route

There are many ways in controlling BGP Routes. In this post, I will discuss by using 4 method, which are:
1. Weight (cisco proprietary)
2. Local Preference
3. Metric
4. AS-Path Prepend

1. Weight
Topology:

Weight is Cisco Proprietary. It’s used to control outbound traffic. Weight is not propagated to the other BGP router, which mean that this configuration only affect local router. Path with higher weight wins.

2. Local Preference
Topology:

Local Preference is used to control outbound traffic. Local preference is propagated to the other BGP router in the same AS, which mean that local preference affect within the same AS. Default local preference is 100, and the path with higher local preference wins.

3. Metric
Topology:

Metric is used to control inbound traffic. This information will be exchanged inter-AS, but only for the peer BGP router. which mean in the topology in the left image, Router A will advertise the AS 100 Prefix to the other BGP peer with metric 0. Path with higher metric wins.

4. Prepend Path
Topology: