I have seen things about RealTime computer systems.
One of the finest descriptions I have seen from: http://www.ecrts.org/ is:
The technical committee on real-time systems promotes state-of-the-art
research and research for applications with temporal constraints,
real-time systems. Such computing systems have to provide results which
are not only correct, but also delivered in time. Instead of average
behaviour as for standard computing, real-time systems have to allow for
guarantees that the temporal requirements will be met.
Consequently, a range of areas in system development is affected:
Applications: distributed real-time information systems/databases;
embedded systems; animation and simulation; real-time control applications.
Software: design and analysis (tools); validation; languages;
operating systems; distributed systems; scheduling; monitoring; software
reuse; object oriented approaches; real-time components.
Quality-of-Service and Multimedia: video/audio streaming with
real-time constraints/limited resources; mobile communications; inhouse
entertainment networks.
Hardware: architectures; real-time oriented devices; coprocessors;
timing engines; energy aware scheduling.
Networks and Communication: communication protocols; protocol
engines; analysis tools; development tools; wireless networking.
Real-time and Reliability: combinations real-time and reliability;
trade-offs; adaptation
Squid is 100% one of the systems that tries and succeed on these
specific tasks.
I was wondering about very high load systems that do exists out there
serving a lot of clients.
I know from somebody that works a lot with CentOS that some major ISP in
united states do use squid since very long ago.
Can *you* provide us with some data about you squid machine?
To install squid and configure it takes lots of design of the network
but if properly configured and planned the results can be pretty amazing.
Here to help,
Eliezer
