Hey guys,
I have two freshly installed directors, running Ultramonkey
on Debian Sarge with a 2.6.18.1 kernel - on very overspec'd hardware. All
is well except, I have noticed when enabling connection synchronisation both
machines slowly rise to a load average of 1.00 without any apparent cause.
These machines arent in production, and are idling in every other way -
infact there isnt even any state information to multicast. If I disable the
connection synchronisation daemon, and reboot. They stay around 0.00~ as
expected. Is this load obscurity normal ? or have I barfed something.
The two directors do their multicasting via cross-over on eth1. Other
config info below:
director01:~# w
17:34:04 up 58 min, 2 users, load average: 1.00, 1.00, 1.00
director01:~# cat /boot/config-2.6.18-1-686 |grep IP_VS
CONFIG_IP_VS=m
# CONFIG_IP_VS_DEBUG is not set
CONFIG_IP_VS_TAB_BITS=12
CONFIG_IP_VS_PROTO_TCP=y
CONFIG_IP_VS_PROTO_UDP=y
CONFIG_IP_VS_PROTO_ESP=y
CONFIG_IP_VS_PROTO_AH=y
CONFIG_IP_VS_RR=m
CONFIG_IP_VS_WRR=m
CONFIG_IP_VS_LC=m
CONFIG_IP_VS_WLC=m
CONFIG_IP_VS_LBLC=m
CONFIG_IP_VS_LBLCR=m
CONFIG_IP_VS_DH=m
CONFIG_IP_VS_SH=m
CONFIG_IP_VS_SED=m
CONFIG_IP_VS_NQ=m
CONFIG_IP_VS_FTP=m
director01:~# ipvsadm --version
ipvsadm v1.24 2003/06/07 (compiled with popt and IPVS v1.2.0)
director01:~# uname -a
Linux director01 2.6.18-1-686 #1 SMP Tue Oct 31 18:53:52 CET 2006 i686
GNU/Linux
Regards
__________________________________________
Dean Manners
|