Load Balancing: DNS round robin
Some identified port list: SMTP 25,POP 110, DNS 53, Web 80, FTP 21 or 23, SSH 22.
IP tunneling,which is supported in router and switch software involves encapsulating a packet supporting one protocol with another packet that run the same or a different protocol.
Key components of tunneling:tunneled protocol, transport protocol, multiplexing identifier.
Load balancing feature: High availiability, Scalability.High availability can be defined as redundancy.Scalability is an application’s ability to support a growing number of users.Of the many methods available to balance a server load, the main two are DNS round robin and Hardware load balancers.Each of them has its own advantages and disadvantages.DNS round robin make several IP address directing to the same logic IP, in the scene,all the nodes combine as one to user transparently.This method only distributes IP to every node,but not distribute actural load to nodes because the load of each connect will not be same.
The problem is thart brower caches that server’s IP address.Once the cache expires the broweer make another request to the DNS server for the IP address associated with the domain name.If the DNS server returns a different IP address that of anther server in the cluster.The session information is lost.Changes to the cluster take time to propagate.One reason is that many large organizations,ISPs,coporations,agencies cache their DNS request to reduce netword traffic and request time.When a user within these organizations make a DNS request,it’s checked against the cache’s list of DNS names mapped to IP addresses.If if finds an entry,it returns the IP address to the user.If the entry is not found in its local cache,the ISP send this DNS request to the DNS server and caches response.
When a cached entry expires,the ISP updates its local database by contacting other DNS servers.When your list of servers changes,it can take a while for the cached entries on other organization’s network to expire and look for the updated list of servers.During that period,a client can still attempt to hit the downed server mode,if that client’s ISP still has an entry pointing to it.In such a case,some users of that ISP could not access your site on their first attempt,even if your cluster has rundundant servers up and running.
This is a bigger problem when removing a node than when adding one.When you drop a node,a user may be trying to hit a non-existing server.When you add one, the server may just be under-utilized until its IP address proprogates to all the DNS servers.Although this method tried to balance the number of users on each server,it doesn’t necessarily balance the server load.Some user could demand a high load of activity during their session than uses on another servers.The electrocircuit module suppling processor power in the mainbroad is called VRM.