Author: M Handley Date: 2006 Link: PDF
DNSwas rolled out to address a scaling problem and enable decentralized administration. Prior to
DNS, host name mappings were maintained in a hosts.txt file managed centrally and manually.
TCPflows can cause congestion.
TCP’scongestion control mechanism was introduced because it was a simple solution:
TCP’scongestion control mechanism is insufficient as it’s not the only protocol (albeit being the most used).
ISPfrom its competitors.
What do IP Multicast, Mobile IP, quality of service, and Explicit Congestion Notification (
ECN) have in common? They are all core network technologies that solve real problems that are not immediately pressing [and hence were not widely deployed].
NATs) is not a shortage of
IPv4addresses but tiered pricing — whereby ISPs charge more for additional IP addresses, even though IP addresses do not in fact cost the
ISPin any significant way.
NATsare frequently cited as a security solution.
NATsare a poor firewall but have an advantage over traditional firewalls because they fail closed.
IPv6eventually sees widespread deployment.
NATsimpedes the deployment of certain types of apps (e.g: Skype)
TCP’scongestion control mechanism is quite minimal, basically probing the network repeatedly to see how much traffic can be in flight at a time, and then backing off when overload is detected via packet loss. One issue with
TCP’scongestion control is
TCP’slimited dynamic range.
BGPhas provided inter-domain routing for the Internet. BGP is conceptually very simple — routes to subnets are advertised together with attributes of the path to that subnet.
BGPis slow to converge; error-prone and easy to misconfigure; and difficult to debug and insecure.
In the early 1990s it became clear that the Internet would run out of IP addresses.
CIDRwas an interim solution to this, and has been quite successful. The rise of
NATs, from being considered an ugly hack to being nearly ubiquitous, has also reduced the problem somewhat.
The Internet was never designed to be optimal for any particular problem — its great strength is that it is a general-purpose network that can support a wide range of applications and a wide range of link technologies. The Internet is also a cost-effective network — it does not make great promises about the quality of service that it provides. It is good enough for a wide range of applications, but anyone considering telesurgery or remote-control of a nuclear power station might well be advised to look somewhere else.