Connection Refused Errors
Ifusers are experiencing connection refused errors on a heavily loaded server, you can tune the
useof network resources on the server.
Whena TCP/IP connection is closed, the port is not reused for the duration of
tcp_time_wait_interval(default value of 240000 milliseconds). This is to ensure that there
areno leftover segments. The shorter the tcp_time_wait_interval, the faster precious
networkresources are again available. This parameter is changed by executing the following
command(do not reduce it below 60000):
usr/sbin/ndd -set /dev/tcp tcp_time_wait_interval 60000
Toautomatically have this ndd command executed after each system reboot, place it in a le
called/etc/init.d/network-tuning and create a link to that le named
/etc/rc2.d/S99network-tuning.
Ifyour system is not exhibiting the symptoms mentioned, and if you are not well-versed in
tuningthe TCP protocol, it is suggested that you do not change the above parameter.
TuningTCP Buering
Ifyou are seeing unpredictable intermittent slowdowns in network response from a
consistentlyloaded server, you might investigate setting the sq_max_size parameter by adding
thefollowing line to the /etc/system le:
set sq_max_size=512
Thissetting adjusts the size of the sync queue, which transfers packets from the hardware driver
tothe TCP/IP protocol driver. Using the value of 512 allows the queue to accommodate high
volumesof network trac without overowing.
Usingthe Solaris Network Cache and Accelerator(SNCA)
TheSolaris Network Cache and Accelerator (SNCA) is a caching server that provides improved
webperformance to the Solaris operating system.
Itis assumed that SNCA has been congured for the system on which the Web Server is
running.For more information about SNCA and its conguration and tuning, refer to the
followingman pages on your system:
■ncab2clf(1)
■ncakmod(1)
SolarisPlatform-SpecicIssues
Chapter4 • Platform-Specic Issues and Tips 91