In this article, we will see how we can add local zone as a resource in Solaris cluster to make the zone highly available. In the past we have seen similar setup in veritas cluster. By configuring zone as a resource, if any one node fails ,automatically zone will fly to other node with minimal downtime.(Flying zone on Solaris). Once you have configured the below things, then we can proceed with bring the localzone under Solaris cluster.
Unlike veritas cluster, local zone IP will be managed from global zone as cluster resource . So let me create a IP resource before proceeding with local zone creation.
1. Login to Solaris cluster nodes and add the local zone IP & Host name information in /etc/hosts file.
UASOL1:#cat /etc/hosts |grep UAHAZ1 192.168.2.94 UAHAZ1 UASOL1:#ssh UASOl2 grep UAHAZ1 /etc/hosts Password: 192.168.2.94 UAHAZ1 UASOL1:#
Here My local zone IP is 192.168.2.94 and host name is UAHAZ1
2.Add the logical host name as resource in Solaris cluster.
UASOL1:#clreslogicalhostname create -g UA-HA-ZRG -h UAHAZ1 CLUAHAZ1 UASOL1:#
- Resource Group Name = – g UA-HA-ZRG
- Local zone Name = -h UAHAZ1
- Local zone IP resource Name = CLUAHAZ1
3.Check the solaris cluster resource status
UASOL1:#clresource status
=== Cluster Resources ===
Resource Name Node Name State Status Message
------------- --------- ----- --------------
CLUAHAZ1 UASOL2 Online Online - LogicalHostname online.
UASOL1 Offline Offline
CLUAZPOOL UASOL2 Online Online
UASOL1 Offline Offline
UASOL1:#
4.You test the resource by pinging the local zone IP.
UASOL1:#ping UAHAZ1 UAHAZ1 is alive UASOL1:#
5.You can see that local zone IP has plumbed by Solaris cluster .
UASOL2:#ifconfig -a
lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
e1000g0: flags=9000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4,NOFAILOVER> mtu 1500 index 2
inet 192.168.2.91 netmask ffffff00 broadcast 192.168.2.255
groupname sc_ipmp0
ether 0:c:29:e:f8:ce
e1000g0:1: flags=1001040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,FIXEDMTU> mtu 1500 index 2
inet 192.168.2.94 netmask ffffff00 broadcast 192.168.2.255
e1000g1: flags=1008843<UP,BROADCAST,RUNNING,MULTICAST,PRIVATE,IPv4> mtu 1500 index 4
inet 172.16.0.65 netmask ffffffc0 broadcast 172.16.0.127
ether 0:c:29:e:f8:d8
e1000g2: flags=1008843<UP,BROADCAST,RUNNING,MULTICAST,PRIVATE,IPv4> mtu 1500 index 3
inet 172.16.0.129 netmask ffffffc0 broadcast 172.16.0.191
ether 0:c:29:e:f8:e2
clprivnet0: flags=1008843<UP,BROADCAST,RUNNING,MULTICAST,PRIVATE,IPv4> mtu 1500 index 5
inet 172.16.2.1 netmask ffffff00 broadcast 172.16.2.255
ether 0:0:0:0:0:1
UASOL2:#
6. Fail-over resource group to UASOL1 and check the status.
UASOL2:#clrg switch -n UASOL1 +
UASOL2:#logout
Connection to UASOL2 closed.
UASOL1:#
UASOL1:#clrg status
=== Cluster Resource Groups ===
Group Name Node Name Suspended Status
---------- --------- --------- ------
UA-HA-ZRG UASOL2 No Offline
UASOL1 No Online
UASOL1:#clresource status
=== Cluster Resources ===
Resource Name Node Name State Status Message
------------- --------- ----- --------------
CLUAHAZ1 UASOL2 Offline Offline - LogicalHostname offline.
UASOL1 Online Online - LogicalHostname online.
CLUAZPOOL UASOL2 Offline Offline
UASOL1 Online Online
UASOL1:#
We have successfully created logicalhostname cluster resource and tested on both the nodes.
7.Create a local zone on any one of the cluster node and copy the /etc/zones/global & /etc/zones/zonename.xml file to other node to make the zone configuration available on both the cluster nodes.Create a local zone without adding network part.(Ex:add net)
UASOL1:#zoneadm list -cv ID NAME STATUS PATH BRAND IP 0 global running / native shared - UAHAZ1 installed /UAZPOOL/UAHAZ1 native shared UASOL1:#ssh UASOL2 zoneadm list -cv Password: ID NAME STATUS PATH BRAND IP 0 global running / native shared - UAHAZ1 configured /UAZPOOL/UAHAZ1 native shared UASOL1:#
You can refer this article for creating the local zone but do not configure network.
8.Halt the local zone on UASOl1 and failover the resource group to UASOL2 to test the zone on it.
UASOL1:#zoneadm list -cv
ID NAME STATUS PATH BRAND IP
0 global running / native shared
- UAHAZ1 running /UAZPOOL/UAHAZ1 native shared
UASOL1:#zoneadm -z UAHAZ1 halt
UASOL1:#
UASOL1:#clrg switch -n UASOL2 +
UASOL1:#ssh UASOL2
Password:
Last login: Tue Jul 1 00:27:14 2014 from uasol1
Oracle Corporation SunOS 5.10 Generic Patch January 2005
UASOL2:#clrg status
=== Cluster Resource Groups ===
Group Name Node Name Suspended Status
---------- --------- --------- ------
UA-HA-ZRG UASOL2 No Online
UASOL1 No Offline
UASOL2:#
9. Attach the local zone and boot it .
UASOL2:#zoneadm list -cv ID NAME STATUS PATH BRAND IP 0 global running / native shared - UAHAZ1 configured /UAZPOOL/UAHAZ1 native shared UASOL2:#zoneadm -z UAHAZ1 attach -F UASOL2:#zoneadm list -cv ID NAME STATUS PATH BRAND IP 0 global running / native shared - UAHAZ1 installed /UAZPOOL/UAHAZ1 native shared UASOL2:#zoneadm -z UAHAZ1 boot UASOL2:#
10. Login to local zone and perform the health check .If everything seems to be fine , then just halt the localzone.
UASOL2:#zlogin UAHAZ1 [Connected to zone 'UAHAZ1' pts/4] Oracle Corporation SunOS 5.10 Generic Patch January 2005 # bash bash-3.2# uptime 12:37am up 1 user, load average: 0.50, 0.13, 0.07 bash-3.2# exit # ^D [Connection to zone 'UAHAZ1' pts/4 closed] UASOL2:#zoneadm -z UAHAZ1 halt UASOL2:#
Click Page 2 to see how to create the resource for local zone and adding in to the resource group .
Gopi R says
Hi Lingeswaran,
I am facing a challange to go with setup, below is my scenario
1. Solaris 10 – GZ1, GZ2 Configurated with Sun cluster 3.3
2. Created zonepath and data mountpoint HA resource as zone-boot-hasp-rs , zone-has-rs
3. When i create zone-ora-rs and zone-lis-rs for automatic DB start during the failover for zone ( Solaris 9), The cluster switch over is fine and DB is getting started at GZ itself in place of solaris 9 zone.
My question here, If this is the situation how can i configure db automatic start for solaris 9 zone.
Even tried with SC_Network=true but no luck…
Really appreciate if you could help on this….
Gopi R
Yadunandan Jha says
Hi Lingeswaran,
I would like to configure Oracle database in SUN Cluster for HA availability for that which file system shall i use QFS, ZFS or SVM ? But don’t want to use RAC ?
I have installed sun cluster 3.3 in Solaris 10 ZFS and allocated shared space from storage as well on both the systems now i would like to create shared file system so that i can install DB in cluster so which file system shall i use for HA.
Regards,
Yadu
Lingeswaran R says
ZFS pools are better than SVM disk sets. ZFS is not a cluster filesystem.But you can use it for failover service group.
Vikrant Aggarwal says
Good Article.
Just to add here, Now the zone clusters are the way to go instead of zone fail over. This help to minimize the downtime associated in fail over of non-global zone.
Thanks
Vikrant Aggarwal
Lingeswaran R says
Yeah.we can make cluster between zones. Thank you vikrant for your comments.