Squid architecture is not realtime at all.
It is not made to handle "blocking" issues properly, as it is required
in realtime systems, and latency can easily grow up because of disk
overload for example.
On 2013-06-21 16:20, Eliezer Croitoru wrote:
On 06/21/2013 03:43 PM, babajaga wrote:
Squid is 100% one of the systems that tries and succeed on these
specific tasks. <
A bit too optimistic. I did a lot of assembler programming (incl.
device
drivers for special HW) for RT-systems (16bit/32bit), using OS, which
were
especially designed to handle RT-tasks using HW/SW-interrupts. I very
doubt,
that ANY SW (not only squid) running on LINUX really can be called
real-time
program. Because it should react during fractions of microseconds to
external events.
Depends on what the definition of RT is...
RT should be something reliable for a human to use in realtime and not
in the RealMathematicalTime.
There is no human that can think in a microsecond.
if there is a human that can think in a microsecond it means he can
think\calc faster then a computer.
What do you think?
Eliezer
--
View this message in context:
http://squid-web-proxy-cache.1019090.n4.nabble.com/Just-wondering-about-something-tp4660745p4660754.html
Sent from the Squid - Users mailing list archive at Nabble.com.
