Bad cluster interconnections

Question

Hi,
I've an Oracle Database 11g Release 11.1.0.6.0 - 64bit Production With the Real Application Clusters option.
Due to some check I was executing I noticed something wrong for the cluster interconnections
This is the oifcfg getif output:
eth0  10.81.10.0  global  public
eth1  172.16.100.0  global  cluster_interconnect
this is equal for both node, and it seems to be right as the 10.81.10.x is the public network and the 172.16.100.x is the private network.
But if I query the gv$cluster_interconnects, I get:
SQL> select * from gv$cluster_interconnects;
INST_ID NAME IP_ADDRESS IS_ SOURCE
2 bond0 10.81.10.40 NO OS dependent software
1 bond0 10.81.10.30 NO OS dependent software
It seems the cluster interconnections are on public network.
Another info that support this fact is the traffic I can see on network interface (using iptraf):
NODE 1:
lo: 629.80 kb/s
eth0: 29983.60 kb/s
eth1: 2.20 kb/s
eth2: 0 kb/s
eth3: 0 kb/s
NODE 2:
lo: 1420.60 kb/s
eth0: 18149.60 kb/s
eth1: 2.20 kb/s
eth2: 0 kb/s
eth3: 0 kb/s
This is the bond configuration (the configuration is the same on both nodes):
+[node01 ~]# more /etc/sysconfig/network-scripts/ifcfg-eth0+
DEVICE=eth0
USERCTL=no
BOOTPROTO=none
ONBOOT=yes
MASTER=bond0
SLAVE=yes
+[node01  ~]# more /etc/sysconfig/network-scripts/ifcfg-eth1+
DEVICE=eth1
USERCTL=no
BOOTPROTO=none
ONBOOT=yes
MASTER=bond1
SLAVE=yes
+[node01  ~]# more /etc/sysconfig/network-scripts/ifcfg-eth2+
DEVICE=eth2
USERCTL=no
BOOTPROTO=none
ONBOOT=yes
MASTER=bond1
SLAVE=yes
+[node01  ~]# more /etc/sysconfig/network-scripts/ifcfg-eth3+
DEVICE=eth3
USERCTL=no
BOOTPROTO=none
ONBOOT=yes
MASTER=bond0
SLAVE=yes
Why the oifcfg getif output is different than the gv$cluster_interconnects view?
Any suggestions on how to configure correctly the interconnections?
Thanks in advance.
Samuel

Answer 1

As soon as I reboot the database I'll check it out.
In the meantime, I snapshot the dabase activity during last hour (during which we suffered a little) and I extracted this top 15 waiting events (based on total wait time (s)):
Event / Total Wait Time (s)
enq: TX - index contention     945
gc current block 2-way     845
log file sync     769
latch: shared pool     729
gc cr block busy     703
buffer busy waits     536
buffer deadlock     444
gc current grant busy     415
SQL*Net message from client     338,421
latch free     316
gc buffer busy release     242
latch: cache buffers chains     203
library cache: mutex X     181
library cache load lock     133
gc current grant 2-way     102
Could some of those depends on that bad interconnection configuration?
And those one are the 15 top wait event based on %DB Time
Event / % DB time
db file sequential read     15,3
library cache pin     13,72
gc buffer busy acquire     7,16
gc cr block 2-way     4,19
library cache lock     2,64
gc current block busy     2,59
enq: TX - index contention     2,29
gc current block 2-way     2,04
log file sync     1,86
latch: shared pool     1,76
gc cr block busy     1,7
buffer busy waits     1,3
buffer deadlock     1,08
gc current grant busy     1
Thanks in advance
Edited by: Samuel Rabini on Jan 11, 2011 4:51 PM
Edited by: Samuel Rabini on Jan 11, 2011 4:53 PM