Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

GCO configuration between 2 nodes cluster and 1node cluster.

Created: 07 Jun 2014 • Updated: 08 Jun 2014 | 8 comments
NBU_13's picture
This issue has been solved. See solution.

Hi,

I need help to configure GCO option between 2 sites.

I have 3 server, 2 node in site 1 and 1 server 1 site 2.

In site 1, I can able to configure 2 node cluster and in 1 server in site 2 , can anyone help me, how to configure GCO between site 1 and site 2.

Thanks in Advanced.

Operating Systems:

Comments 8 CommentsJump to latest comment

Gaurav Sangamnerkar's picture

Hello,

do you have replication established between two sites already ? If no, that would be the first step to take care of.

You must read below section to understand all the steps & prerequisites in order to create a GCO

https://sort.symantec.com/public/documents/sfha/6....

GCO conditions

https://sort.symantec.com/public/documents/sfha/6....

Adding a remote cluster

https://sort.symantec.com/public/documents/sfha/6....

 

G

PS: If you are happy with the answer provided, please mark the post as solution. You can do so by clicking link "Mark as Solution" below the answer provided.
 

NBU_13's picture

Hi Gaurav,

Thanks for your information,

I have few more question on GCO,

yes, I have VVR, between 2 site.

- we need to configure cluster service group for WAC, via GCO wizard - it is asking ip address and port number, can you please guide me, which ip address and port number, we need to provide to configure cluster service.

Edit - Add/delete remote clsuter

1. From site1, I tried to configure GCO, by adding remote clsuter, it is asking hostname/ip address of site2 and port number - so I need to give virtual name/virtual ip address and wac port number of site 2 ?

2. if I have 2 nodes cluster in site 1 and 2 nodes cluster in site2, then I will configure GCO by adding cluster name, but if I have 2 nodes in site 1 and 1 node in site 2, how can i configure GCO between 2 nodes in site 1 and 1 node in site 2 and also how can i relicate between site via VVR.

Please advise me, Thanks

 

Gaurav Sangamnerkar's picture

Hi,

First I would answer your second question, you would need to establish a single node cluster on site B in order to get it participated in GCO. Once you setup a single node cluster, all the HA processess will run where VCS would be able to make communication between sites using WAC.

In your site A, which ever is your primary node will make replication to node in site B. VVR agents in site A will control the ownership of RVG using VVR agents

Now answering your next question

You would need to point to "clusteraddress" IP of site B. Typically when you setup VCS, you would input a clusteraddress IP for VCS. This is the IP address you need to point to when using GCO wizard.

 

G

PS: If you are happy with the answer provided, please mark the post as solution. You can do so by clicking link "Mark as Solution" below the answer provided.
 

NBU_13's picture

Thank you so much Gaurav for detail information,

For confirmation -

1. we need to configure cluster service group with WAC for remote clsuter, via GCO wizard - while I configure it is asking ip address, subnet mask and port number, can you please guide me, which ip address (need seperate nic with ip adddress is required ? ) and (eg: 155615) port number, we need to provide to configure cluster service ?

2, "Now answering your next question

You would need to point to "clusteraddress" IP of site B. Typically when you setup VCS, you would input a clusteraddress IP for VCS. This is the IP address you need to point to when using GCO wizard."

In site 1, so I need to provide site 2's cluster name/ virtual ip address and port number (new port number or site's wac port number ? )

3, While I tried to configure cluster in site 2, with 1 node, it is not asking for any virtual name and ip address for clsuter ? so I need to provide physical ip address of node 3 in site 2

Thanks in Advanced.

mikebounds's picture

The GCO wizard adds a WAC resource to the ClusterService service group so this service group needs to exist before running the GCO wizard.  For new clusters you can create the ClusterService service group  by running "installvcs -configure" or you can create this manually by modifying main.cf and restarting vcs (hastop -all, followed by hastart on node you modified main.cf and then hastart on other node)

Example main.cf below:

include "types.cf"

cluster cluster1 (
    UserNames = { admin = ePQiPKpMQlQQoYQkPN }
    ClusterAddress = "192.168.56.15"
    Administrators = { admin }
    )

system sys1 (
    )

system sys2 (
    )

group ClusterService (
    SystemList = { sys1 = 0, sys2 = 1 }
    AutoStartList = { sys1, sys2 }
    OnlineRetryLimit = 3
    OnlineRetryInterval = 120
    )

    IP webip (
        Device = eth0
        Address = "192.168.56.15"
        NetMask = "255.255.255.0"
        )

    NIC csgnic (
        Device = eth0
        )

    webip requires csgnic

 

Note the IP address above here is separate to your application virtual IP.  The above IP is the Cluster IP and needs to be specified in the webip resource and the ClusterAddress cluster attribute (as shown in example above) and this IP also MUST resolve to a name (I would suggest it resolves to the cluster name which you can add to DNS or /etc/hosts).

Likewise for your 1-node cluster you can use "installvcs -configure" or you can create Cluster Service group manually by modifying main.cf as above, adjusted for 1 node (i.e only need one node in SystemList and AutoStartList) - i.e your 1-node cluster should also use a virtual IP

You can leave port as default for both clusters which is 14155

For VVR, for your 2-node cluster, you MUST use a vritual IP when you configure replication (i.e you must use virtual IP for "vradmin addsec" command), I would recommend you use a virtual IP for your 1-node cluster also, but this will work with a physical IP.

See chapter 8 "Configuring the agents for high availability" from the  "Symantec Storage Foundation and High Availability Solutions Replication Administrator's Guide" for more information

Mike

UK Symantec Consultant in VCS, GCO, SF, VVR, VxAT on Solaris, AIX, HP-ux, Linux & Windows

If this post has answered your question then please click on "Mark as solution" link below

NBU_13's picture

Hi Mike,

Thanks for your information,

1. As per your example : the ip address 192.168.56.15 is a virtual ip address for cluster, we need to use same ip address for wac also, right ?, we need to use default port is 14155 (while Adding remote clsuter)

2. For example, if I configure service group for application (Netbackup) in cluster, I need to give another virtual ip for Netbackup application (eg 192.168.56.25), not a cluster ip (192.168.56.15) in site 1, 2 node cluster, same like, I need to assign another virtual ip in site 2, 1 node cluster for application(eg 192.168.56.35), right ?

Please correct me, If Iam wrong.

cluster cluster1 (
    UserNames = { admin = ePQiPKpMQlQQoYQkPN }
    ClusterAddress = "192.168.56.15"
    Administrators = { admin }
 IP webip (
        Device = eth0
        Address = "192.168.56.15"
        NetMask = "255.255.255.0"
mikebounds's picture

The WAC doesn't really have an IP;

When you run GCO wizard, this will ask for Virtual IP (VIP) of other cluster and this will create a "remote cluster" object in your main.cf which contains the VIP of the other cluster and a heartbeat object (global heartbeat) which contain the local VIP of the cluster.  You don't HAVE to use default port (obviously, as it wouldn't ask you for port, if you couldn't change it), but it is easier to use the default port.  The wac resource will use the "remote cluster" definition in the main.cf to communicate with the other cluster.

NBU will need to have a separate VIP as the Cluster VIP is for communication between clusters so each cluster needs to have its own VIP, where as the NBU VIP is for communication between NBU clients and the clusters so, if you have a single subnet for your 2 clusters, then this NBU VIP should be the same for each cluster so that NBU clients can always contact NBU in the cluster.

If you had different subnets for the clusters, then you need to use different VIPs for NBU, and the NBU clients would HAVE to connect via a single DNS name and the VCS DNS agent would update DNS so that the NBU virtual DNS name would resolve to the right NBU VIP.

Mike

UK Symantec Consultant in VCS, GCO, SF, VVR, VxAT on Solaris, AIX, HP-ux, Linux & Windows

If this post has answered your question then please click on "Mark as solution" link below

SOLUTION
NBU_13's picture

Hi Mike,

Thank you for your detail information.

Now, I have configured - site 2 with 1 node cluster - I have install the SFHA6.1 software and created dynamic disk group, volume and configured cluster with vip address, service group for wac and I tried to install Netbackup software in 1 node cluster, it is asking vip, I gave new vip address and it asked for path for shared disk. I mention manually as F: drive and clicked cluster configuration, it was not successful. but service group and resources are created with offline, 

error - cluster configuration  is unsuccessful, some of the resources are not come online

Please guide to fix it.