1. TCP Splicing for URL-aware Redirection
Objectives
What is a layer-7 switch
Why content-aware switching
Difference between layer-4 and layer-7 routing
TCP splicing
Case studies
Linux-based content aware switch
IBM Layer 5

2. Layer-7 switch
A layer-7 switch is also known as a URL-aware redirector or a content-smart switch
They all refer to “a switch having the capability to redirect HTTP requests based on the URL specified in the client’s GET request”
As we know that there are two types of HTTP request message: GET and POST

3. Layer-7 switch (Cont.) HTTP GET message format:
ASCII (human-readable format)
request line
(GET,
HEAD commands)
GET /somedir/page.html HTTP/1.1
Host:
User-agent: Mozilla/4.0
Connection: close
Accept-language:fr
header
lines
Carriage return,
line feed
indicates end
of message

4. Content-aware Switch Front-end of a web cluster
Internet
Image Server IP
TCP
APP. DATA
Application Server
GET /cgi-bin/form HTTP/1.1
Host:
Switch
HTML Server
Front-end of a web cluster
Route packets based on layer 5/7 (content) information

5. Why use Context-aware Switching
Servers can be specialized for certain types of request
Content segregation
Exploit locality
Affinity-based routing
Increase the performance because of the improved hit rate
Partial replication of server file set
Partition the server’s file set over different nodes

6. Layer 4 vs. 7 Routing

7. Content-aware Switch Architecture
Two way architecture
Both request and response pass through switch
One way architecture
Only the request passes through switch
Response goes directly to the client
Servers and switch may share a virtual IP (VIP)
Needs modification of ARP/ICMP

8. Content-aware Switch Architecture
Two way architecture
Server returns the
response to the switch
One way architecture
response to the client
client
switch
server

9. Layer-7 Two-way Architecture

10. Layer-7 Two-way Mechanisms
TCP gateway
An application level proxy running on the web switch mediates the communication between the client and the server
TCP splicing
reduce the overhead in TCP gateway. Packet forwarding occurs at network level between the network interface driver and the TCP/IP stack, is carried out directly by OS
user
kernel
user
kernel

11. URL Parsing is expensive!!
Performing content-aware routing implies that some kind of string searching and matching algorithm is required
Such a time-consuming function is expensive in a heavy traffic web site
Experience showed that the system performance would be severely degraded if we implement some URL parsing functions in the distributor

12. TCP splicing
Once the two TCP connections are established, they are spliced
IP packets are forwarded at the network layer
TCP splicing requires
Connection binding
Packet analyzer to rewrite packets
Appropriate address translation
Sequence number modifications to be performed on the packets
Basically, we are deploying connection re-use

13. Operation of Content-aware Distributor

14. Functional Overview of Linux-based content aware switch

15. IBM Layer 5 Switch Architecture
Switch core
Port controller:
Identify packets (layer 5) and send them to CPU
Processing all other packets
CPU: PowerPC 603e
Parse http request
URL based routing

16. Layer-7 one-way architecture

17. Layer-7 one-way mechanisms
TCP handoff
The switch hands off the TCP connection endpoint to the server
TCP connection hop
Software-based proprietary solution
encapsulating the IP packet in an RPX packet and sending it to the server

18. Layer-7 one-way mechanisms
Migrate the created TCP connection from the switch to the back-end sever
Create a TCP connection at the back-end without going through the TCP three-way handshake
Retrieve the state of an established connection and destroy the connection without going through the normal message handshake required to close a TCP connection
Once the connection is handed off to the back-end server, the switch must forward packets from the client to the appropriate back-end server

19. TCP Splicing: Summary Why URL-aware switching
Layer-4 versus layer-7 routing
TCP splicing
TCP gateway
Case studies
Linux-based content aware switch
IBM Layer 5