KnowledgeShop

Learn & Share

Web Application Architecture

Load Balancers

Overview

From a physical point of view, it can be plugged anywhere in the architecture:

  • in a DMZ
  • in the server LAN
  • as front of the servers, acting as the default gateway
  • far away in an other separated datacenter

Why load balancers?

  • HTTP is not a connected protocol and is stateless
  • When there are multiple app servers, the user may potentially hit a different server for every request and a different session on every server.

Possible solutions are

  1. Use a clustered web application server where the session are available for all the servers
    • Pros
    • Cons
      • Only certain products like Weblogic, Tomcat, JBoss, allow to create a cluster
      • very complex to set up and maintain
  2. Sharing user’s session information in a database or a file system on application servers
    • Pros
      • simple to share session via DB or shared file system
    • Cons
  3. Use IP level information to maintain affinity between a user and a server
    • An easy way to maintain affinity between a user and a server is to use user’s IP address: this is called Source IP affinity.
    • Pros
    • Cons
      • works only if a user use a single IP address or never change his IP address during the session.
      • wouldn’t work for multiple users who are behind a single proxy
  4. Use application layer information to maintain persistance between a user and a server
    • Store the user information details in a Session Cookie, either set by the load-balancer itself or using one set up by the application server.
    • Pros
    • Cons

Affinity vs. Persistence

  • Affinity: this is when we use an information from a layer below the application layer to maintain a client request to a single server

  • Persistence: this is when we use Application layer information to stick a client to a single server sticky session: a sticky session is a session maintained by persistence

  • The main advantage of the persistence over affinity is that it’s much more accurate, but sometimes, persistence is not doable, so we must rely on affinity.
  • Using persistence, we mean that we’re 100% sure that a user will get redirected to a single server.
  • Using affinity, we mean that the user may be redirected to the same server

Types

2 types of load balancers:

  • Hardware load balancer (Level 4 load-balancers)
    • Layer 4 (transport level). e.g.,: TCP and UDP protocols are transport level.
    • A layer 4 load-balancer takes routing decision based on IPs and TCP or UDP ports.
    • It has a packet view of the traffic exchanged between the client and a server which means it takes decisions packet by packet.
    • The layer 4 connection is established between the client and the server.
    • It is really fast but can’t perform any action on the protocol above layer 4.
    • The fastest layer4 load-balancers uses an ASIC to take routing decision.
    • Example: F5 BIG-IP, Citrix Netscaler, etc.
    • Expensive
  • Software load balancer (Level 7 load-balancers)
    • Layer 7 (application level) e.g.,: HTTP, FTP, SMTP, DNS protocols are application level.
    • A layer 7 load-balancer takes routing decision based on IPs, TCP or UDP ports or any information it can get from the application protocol (mainly HTTP).
    • The layer 7 load-balancer acts as a proxy, which means it maintains two TCP connections: one with the client and one with the server.
    • The packets are re-assembled then the load-balancer can take a routing decision based on information it can find in the application requests or responses.
    • Even if this kind of processing seems slow, it is not that much: less than the millisecond.
    • Cheaper

Hardware load balancer architectures

1. NAT or routed mode

2. Direct server return or Gateway mode

3. IP Tunnel mode

Software load balancer architectures

1. Proxy mode

2. Transport proxy mode

References

  • Sites
    • haproxy.com/blog