I've never had a chance to put my hands on a RAC system, so finally decided to try, using Tim Hall's instructions. Of course the fundamentals of setting up a machine under VBox is well understood by me, having done it scores if not hundreds of times. But the additional network config is beyond me at this point, and I'm finding the given instructions to be a bit vague at precisely the point I need extra guidance.
Here's my setup so far:
Host OS - Ubuntu 16.04 LTS
- adapter 1 (eth0) - NAT, dhcp
- adapter 2 (eth1) - hostonly, ip 192.168.56.103
- adapter 3 (eth2) - internal (intnet) , ip 192.168.1.101
- OS is Oracle Linux 6, 64-bit
on the vm, /etc/hosts looks like this:
[root@ol6-121-rac1 ~]# cat /etc/hosts
127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4
::1 localhost localhost.localdomain localhost6 localhost6.localdomain6
127.0.0.1 localhost.localdomain localhost
# Public
192.168.56.103 ol6-121-rac1.localdomain ol6-121-rac1
192.168.56.104 ol6-121-rac2.localdomain ol6-121-rac2
# Private
192.168.1.101 ol6-121-rac1-priv.localdomain ol6-121-rac1-priv
192.168.1.102 ol6-121-rac2-priv.localdomain ol6-121-rac2-priv
# Virtual
192.168.56.105 ol6-121-rac1-vip.localdomain ol6-121-rac1-vip
192.168.56.106 ol6-121-rac2-vip.localdomain ol6-121-rac2-vip
# SCAN
192.168.56.107 ol6-121-scan.localdomain ol6-121-scan
192.168.56.108 ol6-121-scan.localdomain ol6-121-scan
192.168.56.109 ol6-121-scan.localdomain ol6-121-scan
[root@ol6-121-rac1 ~]#
Now is where things get confusing.
First, this statement:
The SCAN address is commented out of the hosts file because it must be resolved using a DNS, so it can round-robin between 3 addresses on the same subnet as the public IPs. The DNS can be configured on the host machine using BIND or Dnsmasq, which is much simpler. If you are using Dnsmasq, put the RAC-specific entries in the hosts machines "/etc/hosts" file, with the SCAN entries uncommented, and restart Dnsmasq.
Hmm. OK, so I added all of the above entries to /etc/hosts on the host system, and restarted dnsmasq. However, on further investigation, on the host system, I see this:
ed@ed-Gazelle-00:~/.ssh$ ps -ef|grep dnsmasq|grep -v grep
nobody 13357 13319 0 May08 ? 00:00:00 /usr/sbin/dnsmasq --no-resolv --keep-in-foreground --no-hosts --bind-interfaces --pid-file=/var/run/NetworkManager/dnsmasq.pid --listen-address=127.0.1.1 --cache-size=0 --conf-file=/dev/null --proxy-dnssec --enable-dbus=org.freedesktop.NetworkManager.dnsmasq --conf-dir=/etc/NetworkManager/dnsmasq.d
ed@ed-Gazelle-00:~/.ssh$
According to the man page for dnsmasq, '--no-hosts' means do NOT read /etc/hosts.
And that's the sum total I know about dnsmasq.
Then he goes on to talk about /etc/resolv.conf, but it is not clear if that is on the host or the vm (in his setup, his host is also linux).
In any event, on my host os:
ed@ed-Gazelle-00:~/.ssh$ cat /etc/resolv.conf
# Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8)
# DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
nameserver 127.0.1.1
And on the vm:
ed@ed-Gazelle-00:~/.ssh$ cat /etc/resolv.conf
# Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8)
# DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
nameserver 127.0.1.1
And to address the reported overwriting of resolv.conf:
[
root@ol6-121-rac1 ~]# cat /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE=eth0
TYPE=Ethernet
UUID=1592164e-f052-438c-b235-c2d37cf9b26a
ONBOOT=no
NM_CONTROLLED=yes
BOOTPROTO=dhcp
HWADDR=08:00:27:82:8A:F8
DEFROUTE=yes
PEERDNS=yes
PEERROUTES=yes
IPV4_FAILURE_FATAL=yes
IPV6INIT=no
NAME="System eth0"
So as far as I can tell, at this point everything should be in place, but when tested, from the vm:
[root@ol6-121-rac1 ~]# nslookup ol6-121-scan
;; connection timed out; trying next origin
;; connection timed out; no servers could be reached
And on the host:
ed@ed-Gazelle-00:~/.ssh$ nslookup ol6-121-scan
Server: 127.0.1.1
Address: 127.0.1.1#53
** server can't find ol6-121-scan: NXDOMAIN