hi all..
in my squid log i've found some line like this :
"squidaio_queue_request: WARNING - Queue congestion"
and i'm search in my squid source dir, i've
that error message in
src/fs/aufs/aiops.c
//begin
if (request_queue2.head) {
static int filter = 0;
static int filter_limit = 8;
if (++filter >= filter_limit) {
filter_limit += filter;
filter = 0;
debug(43, 1) ("squidaio_queue_request:
WARNING - Queue congestion\n");
}
}
//end..
could somebody englighten me.. what's meaning for..
am i doing something wrong in configure my squid ?..
eg.. configure with too many AUFS_IO_THREADS perhaps
?.
I'm using squid-2.5PRE7
and kernel 2.5.18 (I'm needed ide driver for
via vt8233a and promise pdc20276 so i'm
stuck with beta linux kernel)
my configure options as follow :
before that i'm setting environment vars
export CFLAGS="-O6 -s -DNUMTHREADS=50"
ulimit -HSn 16384
./configure --enable-dlmalloc --enable-async-io=50
--enable-poll --enable-arp-acl --enable-snmp
--enable-removal-policies --prefix=/usr
--exec_prefix=/usr --bindir=/usr/sbin
--libexecdir=/usr/lib/squid --localstatedir=/var
--sysconfdir=/etc/squid
--disable-ident-lookups
--enable-linux-netfilter
i'm using 2 cache dir each at separate hdd,
one at /dev/hde1 and second at /dev/hdg1
what's the most optimum setting (number of
aio_threads) for each hdd / cache_dir ?
thanks..
regards
Budi Aditya
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
Received on Wed Jun 05 2002 - 14:03:53 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:08:28 MST