On 24/01/11 22:02, Saiful Alam wrote:
TRIED AUFS, but didn't get better performance, while
researching in the web, I read everywhere that AUFS is better than
UFS in terms of performance,but I don't know why I get bad
performance with this. Processor is Intel(R) Core(TM) i7 CPU
870 @ 2.93GHz Motherboard is Intel® Desktop Board DH55TC
UFS performs one file operation per cache_dir over many small steps in a
single series.
AUFS operates by using multiple threads per cache_dir to execute the
disk I/O. Taking better advantage of OS disk scheduling mechanisms etc.
RAM is 2x4GB=8GB DDR3 DISK 1 = 250GB Hitachi SATA HDP72502
{[USED ONLY FOR UBUNTU SYSTEM& BOOT]} DISK 2 = 500GB Hitachi SATA
HDS72105 {[USED FOR CACHE DIRECTORIES ONLY, MOUNTED ON /MEDIA/CACHE
FILESYSTEM EXT4]}
Having only one disk with multiple cache_dir on it is likely the source
of any slow disk problems you see. Squid uses its cache_dir in a
random-access mostly-write fashion with all cache_dir accessed in parallel.
When more than one UFS-based (ie UFS, AUFS or DiskD) dirs exists on the
same drive their I/O will collide with each other causing I/O signals
and other processor 'page faults' to slow the machine disk IO down.
Sometimes a lot.
Normally my Disk I/O never goes more than 15% and I would say the
average is about 3-4%. For bandwidth monitoring I usually see iptraf
which is also good, but surely I'll try ifstat next time. At peak
hours (which is between 10pm - 2am GMT +0600), we have around 500
clients connected (approx)
I have tried apt-get install squid3 (which is the default 3.1.6 in
apt repository) and found the performance of 3.1.10 (my custom
configuration) is better than the 3.1.6.
Regards, Saiful
Using webkit developer tools or firebug plugin to firefox you can get a
trace of all the objects on a particular page along with their fetch
times and total page rendering time.
In my experience most "slow" pages are composed of vast amounts of code
which needs to be run. The longer this takes to render the worse the
page appears to be in the eyes of clients.
The fetch time is the only part relevant to Squid. This should be also
shown by the "transfer time" column in your access.log. Once you have
identified the particular delayed requests you can look at why they are
delayed.
Amos
--
Please be using
Current Stable Squid 2.7.STABLE9 or 3.1.10
Beta testers wanted for 3.2.0.4
