How to deploy and configure EMC RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster

By | March 21, 2015

In this part of deploying RecoverPoint for Virtual Machines 4.2, I will show how to create and configure vRPA Cluster. Before creating a vRPA Cluster, please ensure that the following prerequisites are met:

  1. Installing the RecoverPoint for VMs splitter on each ESXi host in the ESXi cluster.
  2. Configuring Software iSCSI Adapters on every ESXi host.
  3. Deploying vRPAs using OVA.

The above prerequisites are covered in my another post here.

There are some prerequisites to create vRPA Cluster:

  • RecoverPoint Deployment Manager (DM) 2.2 or later.
  • Fully-licensed RecoverPoint for VMs 4.2 or later and licenses for each replicating VM.
  • Datastore for the virtual repository must be mapped to every ESXi capable of running
    vRPAs. An additional minimum of 3 GB of storage is required for the repository.
  • The vRPA has ongoing validation algorithms. A failure of any validation causes the vRPA to disconnect from the vRPA cluster (meaning that this vRPA will not function as an active RPA).
  • Sharing vRPA iSCSI interfaces (iSCSI1 and iSCSI2) with the vRPA regular networks (LAN and WAN) may result in performance degradation and unexpected behavior of the vRPA.

To create a vRPA Cluster please follow the following steps:

  1. Launch the Deployment Manager and select Install RecoverPoint for Virtual Machines. Click Login.
  2. In the RecoverPoint Deployment Manager Wizard screen, select the RecoverPoint Installer Wizard.
  3. In the Prerequisites screen, read the prerequisites. Once you have ensured that the conditions are met, select the I have fulfilled the conditions for installing the RecoverPoint cluster checkbox.
  4. In the Configuration file screen, select whether to create a new installation configuration file or to continue installation from a saved file. Type the appropriate file location.
  5. In the Environment settings screen, define the Cluster name. Define the environment settings.How to deploy and configure RecoverPoint for Virtual Machines 4.2 Part 2: Create a vRPA Cluster
  6. In the RPA discovery screen, define how you want to set the IP addresses of the vRPAs.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 2
  7. In the IP and connectivity settings screen, define the vRPA cluster IP configurations and vRPA IP settings.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 3
  8. In the Login credentials screen, enter the default login credentials. (pass: boxmgmt)How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 4
  9. In the Connectivity results screen, review the results of the connectivity test.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 5
  10. In the Cluster iSCSI settings screen, define the vRPA cluster iSCSI settings. If you want to use CHAP, configure the credentials for the iSCSI ports.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 6
  11. In the RPA iSCSI ports screen, define the vRPA cluster iSCSI IP configuration.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 7
  12. Select 3rd option (Do not update. Continue with current RecoverPoint release on RPAs).How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 8
  13. In the Apply configuration results screen, the results of applying the configuration settings to all vRPAs are displayed.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 9
  14. In the vCenter Server registration screen, enter the vCenter server credentials, RecoverPoint administrator credentials (login: admin, password: admin), and the location of the vCenter certificate. The vCenter certificate can be found on the vCenter server at the following location:
    C:\Program Files\VMware\VMware Virtual Center\SSL\rui.crt
    If using vCSA, the vCenter certificate can be found at the following location:
    /etc/vmware-vpx/ssl/rui.crt
    EMC RecoverPoint for Virtual Machines: Deploy vRPA cluster
  15. In the Datastores screen, select a datastore from the Available datastores table to be the vRPA repository.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 11
  16. In the Storage configuration - cluster summary screen, review the vRPA information.How to deploy and configure RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster 12
  17. In the Summary screen, verify that the installation was completed successfully.

Please follow the next post about connecting vRPA clusters.

Note: You should deploy and create the second vRPA cluster (following above steps) in the second site to be able to replicate VMs between sites.

17 thoughts on “How to deploy and configure EMC RecoverPoint for Virtual Machines 4.2? Part 2: Create a vRPA Cluster

  1. SBaker

    Hello, I'm back again. This time I've read your entire article. (didn't realize there were 2 parts at first.)
    I've been successful up to step 11 in the installer wizard. Under 11.2.Datastores it keeps failing and giving me an error that states "Internal Error, please contact Customer Support"
    I am just trying to demo this, so I don't believe customer support is going to help me. The utility shows my datastore has the following properties - 1.35 TB capacity, 102GB provisioned, 1.25TB free, type=VMFS.
    So obviously I have plenty of space. Did you get any errors in your initial attempts at configuring the cluster?

    Reply
    1. Mariusz Post author

      Hi,

      yes, I go this error when I started the installer wizard again without deleting files (on datastore) created by earlier "wizard attempts".

      Have you tried to delete vRPAs and also some additional files on datastore and start again since zero? I think it would be the easiest way to solve this problem.

      Reply
  2. SBaker

    Hello,

    I have started from scratch on a few occasions now with the same result. One thing I may not be understanding is this - in the 'prerequisites ' section of the Deployment Manager the first thing it says is that the RPAs are connected to SAN and Ethernet Network. I just took that to mean that they had connectivity for later during the Deployment Manager setup. The Prerequisites section also lists a few things involving VNX arrays, but RP for VMs is supposed to be storage agnostic. So I guess I'm asking if there is work you did to connect to storage before even start the Deployment Manager utility. You screenshots look very similar to my screenshots during the Deployment Manager phase, and then mine fails at step 11.2
    I've posted on the EMC community forums and I got a few responses, but then they just stopped. Kinda frustrating

    Reply
    1. SBaker

      Amazingly, my install at our remote site went much smoother. The hardware was almost identical, so I'm not sure why I am having such a hard time with the install at our main site. I guess for anyone who tries this if you don't succeed on your first attempt you may want to try it on a different test host if you have one. Now I still have to go back and figure it out on our main test host.

      Reply
      1. Mariusz Post author

        Hi, thanks for info! I'm waiting for a new version of RP4VM...and I will upload a new post about it.

        Reply
        1. SBaker

          Well today I got everything squared away. 1 thing I did differently was use admin/admin at step 11.1 where you are setting up the vCenter credentials. EMC support had mentioned to use admin/admin at step 6, but it defaults to the boxmgmt user and cannot be changed. Now I'm not sure if getting the remote site setup somehow helped the main site work, but the splitter for the remote site is entered as the splitter for the main site as well. So if you are planning on testing this across a WAN and you are having trouble setting up 1 site, try switching to the other and see if you have better luck. Then go back to the first site.
          Thanks for this blog!

          Reply
          1. Mariusz Post author

            Good info. admin/admin is not documented as well as I remember. Very interesting... I updated this post.

            Reply
  3. SBaker

    One last thing I noticed is that my RPforVMs plugin is not working. I can go to the web address of my vSphere Web Client and look at the /mob page and see extensionList["com.emc.recoverpoint.vwc"] in there. I can also see it as an 'available' plugin when I check the vCenter fat client, but have no option to enable it.
    However when I SSH into my vCenter appliance I do not see any EMC plugins at /usr/lib/vmware-vsphere-client/plugin-packages.
    I would copy them over but I'm sure I need to name the directories accurately. If you have your vCenter running as an appliance could you please post the directory for the EMC Recoverpoint plugin? I'm sure this could come in handy to other people as well.
    I hope you don't mind - I have put links to your site on a couple forums. There is very little info on Recoverpoint for VMs out there

    Reply
    1. SBaker

      In regards to my last reply = the location of the vCenter plugin is /var/lib/vmware/vsphere-client/vc-packages/vsphere-client-serenity/
      Mine is still not working after restarting the vSphere Web Client services, so now I need to figure out what to do next. That plugin is required for the administration of RecoverPoint for VMs.

      Reply
  4. SBaker

    Ok, now for another tidbit. Apparently my RPAs installed the client with incorrect permissions.
    If your plugin is not showing up in the vCenter web console, go to the above directory and make sure the files are marked executable. If any aren't, just do 'chmod +x
    My plugin is now working.

    Reply
    1. Mariusz Post author

      Thanks for your valuable comments! What I can add more.. EMC RecoverPoint for VM is a new software...

      Reply
  5. SBaker

    You are welcome. Hopefully it can save some other people the headaches I had. I've also discovered that it uses a good amount of bandwidth if you are setting up 2 separate clusters over a WAN. Our connection isn't very good right now, and trying to manager RecoverPoint for VMs through the vSphere Web plugin is almost impossible. It is so bad we have discontinued any testing right now until we get our new circuit in place. So be warned if your WAN connection to the remote site is not very robust.

    Reply
  6. Jus

    How do I check whether a splitter is enabled. VIB version 4.3 installed successfully but replication fails with splitter errors

    Reply
    1. Mariusz Post author

      Hi,

      The splitter is installed/enabled because the create vRPA cluster wizard checks it on hosts.

      Reply
  7. Jus

    splitter is installed on esxi server through command line. I don't think the vRPA deployment checks this.
    When deploying OVA for vRPA, I have no option to select the correct network for the iscsi networks. May be this is problem?

    Reply
    1. Mariusz Post author

      An installer checks whether splitter is installed because I forgot to install the splitter on one of my esxi host and I received an error during the vRPA cluster deployment. When you deploy vRPA via OVA you specify just a temporary IP address. Later during creating a vRPA cluster, you specify iSCSI/WAN networks for each vRPA. I hope this helps you.

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.