Friday, March 16, 2012

ICND 2 Flashcard: Frame Relay

Frame Relay allows the expansion of the WAN with less hardware by providing virtual circuits; less costly than running multiple leased lines like HDLC and PPP.

For example Router 1 (DTE) at your company site has to connect to Router 2 (DTE) at the other site. In between is the Frame Relay cloud. The Router 1 connects to one switch (DCE) in the cloud, and the other Router 2 connects to anothr swtich (DCE). Router 1 could also connect to a Router 3 and so on.

One router is an access linl which can support multiple virtual circuits to send data to multiple remote routers. Each link is a virtual circuit. The data link header and trailers on each frame. The header holds ad address filed called a DLCI.
DLCI - Data Link Connection Identifier

The notes for this section are taken from the first CCNA Bootcamp course I studied in 2003 by Marketbridge Technologies in Hull-Gatineau. http://www.marketbridge.com/#  The company has grown and changed alot over the years from offering courses to consulting services, but the owner remains the same so it's probably the same company.

Two Frame Relay encapsulations: Cisco and IETF
Cisco is the default, and it means that you have a Cisco router on each end of the Frame Relay network. If you don’t have a Cisco router on the remote end of your Frame Relay network, then you need IETF encapsulation.

Frame Relay is a cost efficicient technology, for intermittent connection from LAN or between endpoints to major backbones or a public WAN.  A permanent virtual circuit (PVC) is formed, enabling the customer to perceive a continous, dedicated connection without having to pay for a full-time leased line; the ISP determines the route each frame travels to its destination and can charge based on usage.  Think of it as shared bandwidth, a portion of the dedicated paid bandwidth to allotted to each user; allows the user to exceed the guaranteed bandwidth if resources are available

However for a truly private network, Frame Relay would run over leased lines over T-1 lines. A dedicated connection during the transmission period is required, but without a steady flow of transmisions, so it is not often used for voice or video.  The data units are in frames in variable sizes.

Packet based switching.
Frame relay is based on the older X.25 packet-switching technology which was designed for analog voice. Frame relay today is fast packet switching because it operates in the Layer 2 data-link layer and not so much Layer 3 network even though a frame can carry packets of Ethernet and X.25 Error checking or resending is up to to the endpoints to evaluate.

DLCI - Data Link Connection Identifier

I really like the description of a frame-relay map and how it joins an DLCI with an IP address much like ARP mapping MAC address to IP. See the IP-to-DLCI mappings with the command, show frame-relay map (IARP is default on Cisco routers).

Possible network topologies
Hub and Spoke: one hub many spokes used with sub interfaces
Partial Mesh or Hub and Spoke - routers do not have a VC to all other routers

Full Mesh: each router has a logical circuit to every other router

Hybrid: Two remote routers may have VC to each other providing full mesh connectivity between them and the hub. The other remote routers may have only one VC back to the hub.
Frame-Relay LMI
Frame-Relay used the Local Management Interface (LMI) protocol to generate keepalives and obtain a status on the virtual circuits.  LMI message formats: Cisco (DLCI 1023), ANSI or Annex D (DLCI 0), and Q.933A or Annex A (DLCI 0?) containing info regarding -
  • Keepalives
  • Multicasting - Multicasting uses the reserved DLCIs from 1019 through 1022.
  • Global addressing - This provides global significance to DLCIs, like a LAN
  • Status of virtual circuits - This provides DLCI status.
Possible states for the circuits
Active - the circuit is up and running
Inactive - possible cause, circuit to CO (frame switch) is find but remote end is down
Deleted - circuit to CO not working, interface down or cabling issue, no LMI

Congestion Control in Frame-Relay
DE - Discard Eligibility, mark for packets exceeding the CIR
BECN - Backward Explicit Congestion Notification, tell source to slow down transmission
FECN - Forward Explicit Congestion Notification, tell destination there was congestion in the cloud
An excellent summary of everything you need to know about Frame Relay for Cisco CCNA, written like a study blog with network diagrams. The style of writing is like an instructor speaking. Click here.

usage: conf t
encapsulation frame-relay

Scenario
There are seven remote sites to connect and only one serial port on the router. Instead of seven leased lines, frame-relay might be a good soluction to statistically multiplex multiple logical circuits over one physical interface to save money.

Access rate The maximum speed that the Frame Relay serial interface can transmit.

CIR The maximum bandwidth of data guaranteed to be delivered. In reality, it’s the average amount that the service provider will allow you to transmit.

For example if the T1 is carrying an access rate of T1 (1.544Mbps) and you're paying for a CIR of 256Kbps. The first 256Kbps of traffic you send is guaranteed delivery. Beyond that, it's a “burst”— a transmission that exceeds the guaranteed 256Kbps rate and up to the T1 access rate (if that amount is in your contract). I believe this sounds like the whole discussion about paying for bandwidth usage and doing away with unlimited with the CRTC ruling??

Obviously if the combined committed burst (the CIR) and excess burst sizes, known as the MBR or maximum burst rate when combined, exceed the access rate, then the packets will be dropped, depending on the service provider.

No comments:

Post a Comment