|
Hi!
I have a small web-cluster with 2 directors and (for the
moment) 2 realservers.
The Directors are PIII 650 Machines.
RedHat 6.2 Kernel 2.2.16-4
Piranha 0.4.17-2
I am using NAT. and are directing port 80 and 443.
Loadmonitor Uptime.
Scheduler wlc.
When the directors have been "active" and redirecing my
test-traffic (5-40 hits/s) for about 6 hours the director
stop redirecting traffic...
But it still works in some way.. (The Backup doesn't take
over util I run "pulse stop" on the master).
Then the backup starts.. and runs for another ~ 6h .. then
it stops..
The heartbeat-monitoring (for the directors) are running on
the inside-network (the same subnet as my realservers).. is
this good or is it better to have it on the outside?
If I try to restart Pulse on the failing dierctor.. I got a
message like this.
IP Virtual Server version 0.9.14 (size=4096)
Prot LocalAddress:Port Scheduler Flags
-> RemoteAddress:Port Forward Weight ActiveConn
InActConn
[root@weblvs2 /root]# /etc/rc.d/init.d/pulse restart
Shutting down pulse:
[ OK ]
Starting pulse: pulse: cannot create heartbeat socket.
running as
root?
[FAILED]
Any suggestions on what I should do?
/ Johan Elmerfjord
Instadia, www.instadia.net
|
