x86 Linux bind9 -futex take much sys CPU time and cause errors

JINMEI Tatuya / 神明達哉 jinmei at isl.rdc.toshiba.co.jp
Fri Feb 11 08:26:45 UTC 2005


>>>>> On Tue, 18 Jan 2005 13:21:34 +0800, 
>>>>> <yakut at pku.edu.cn> said:

> I cannot think out what named threads are competing for, one 
> possibility is data cache, but I have a 100k records queryperf input 
> file and another 100k records domain data file, this might not make 
> sense, because in such read operation, I don't believe that one named 
> thread will lock all data cache; another concern is network, named use 
> udp for communication, at the time of running,  netstat shows below:

If the 100k records are in a single zone, the large number of records
do not help very much in terms of reducing lock contention, since
different records (more accurately, different RRsets) in a single zone
share a small number of locks.  See lib/dns/rbtdb.c and search for the
keyword "node_locks".

					JINMEI, Tatuya
					Communication Platform Lab.
					Corporate R&D Center, Toshiba Corp.
					jinmei at isl.rdc.toshiba.co.jp



More information about the bind-users mailing list