Topic
13 replies Latest Post - ‏2014-10-16T12:26:45Z by LeonelSAA
LeonelSAA
LeonelSAA
18 Posts
ACCEPTED ANSWER

Pinned topic IP Replication V7000

‏2014-02-25T16:56:16Z |

Hi all,

I have a trouble with V7000 partnership IP replication. So, I try to create partnership with other V7000 on remote site but return error message and I can't create partnership.

When I put the IP of remote Storage return the error, how can see on the picture appear the error.

How can solve this issue?

 

Many thanks

  • AndersLorensen
    AndersLorensen
    156 Posts
    ACCEPTED ANSWER

    Re: IP Replication V7000

    ‏2014-02-25T18:07:05Z  in response to LeonelSAA

    Does both V7000's run the same firmware? If the remote site runs an older firmware that doesnt support IP replication that could explain it. (IP Replication was added very recently)

     

    Is the gateway on both V7000's set correct? And Network mask? And are you sure there is no firewall that Blocks trafic in between?

     

    /Anders

    • LeonelSAA
      LeonelSAA
      18 Posts
      ACCEPTED ANSWER

      Re: IP Replication V7000

      ‏2014-02-26T06:50:18Z  in response to AndersLorensen

      Yes, both runs 7.2 version. Gateway and mask it's correct. And there are no firewall blocking traffic, because I making test before install secondary V7000 on remote site to put at service.

      Look at the picture below what I did yesterday. I set IP address on both storage at group 1 interface ethernet 1. What I don't made? what remains to be done?

      • AndersLorensen
        AndersLorensen
        156 Posts
        ACCEPTED ANSWER

        Re: IP Replication V7000

        ‏2014-02-26T08:12:36Z  in response to LeonelSAA

        Are both systems in the same layer (Storage or replication)?

        Not sure what else could be the issue. I havent played around with the IP replication myself yet.

         

        /Anders

        • LeonelSAA
          LeonelSAA
          18 Posts
          ACCEPTED ANSWER

          Re: IP Replication V7000

          ‏2014-03-04T17:20:17Z  in response to AndersLorensen

          Yes are in the same layer.

           

          I need your help

          • AndersLorensen
            AndersLorensen
            156 Posts
            ACCEPTED ANSWER

            Re: IP Replication V7000

            ‏2014-03-04T21:39:17Z  in response to LeonelSAA

            I'd start by removing the gateways for the iSCSI IP's. They are wrong and not needed, and might confused it.

             Then try doing a ping the remote storwize on both storwizes. (done from the CLI)

            svctask ping <remote ip address>

            If there is no ping, you want to look at your router configuration. If there is a ping, I'm out of good ideas, except contacting IBM support.

            /Anders

            • LeonelSAA
              LeonelSAA
              18 Posts
              ACCEPTED ANSWER

              Re: IP Replication V7000

              ‏2014-03-05T09:14:05Z  in response to AndersLorensen

              I already made all ideas which you give me and don't solved.

              I'm contacting IBM support now.

  • M.Vieselman
    M.Vieselman
    2 Posts
    ACCEPTED ANSWER

    Re: IP Replication V7000

    ‏2014-04-14T12:45:02Z  in response to LeonelSAA

    Hi LeoneISAA id you eventually find the solution? 

    I have the same problem went to IBM support and got the answer "sorry this is a network problem" but i can ping default gatway, can ping myself but can't ping the partner

     

    • LeonelSAA
      LeonelSAA
      18 Posts
      ACCEPTED ANSWER

      Re: IP Replication V7000

      ‏2014-04-15T07:16:13Z  in response to M.Vieselman

      Hi,

      I solved this issue without ibm support, because our warranty expired.

      I had that error because the network configuration on the nodes wasn't good. Was configured the same IP with the cluster and IP service from nodes for services. Generally can be configured 2 or 3 IP for the Cluster failover. with 2 IPs, one for first node and other for second node, the IP from cluster always will be changed when one of the nodes goes down "This is the issue which returns that erro: Unreachable Cluster IP address when be create IP Partnership".

      the best practice is configuring 3 IP for the cluster failover, 2 IPs for nodes services and 1 for cluster failover on the enclosure. With this settings will not have the error "Unreachable Cluster IP address" when create IP Partnership.

      many thanks.

      regards.

      • M.Vieselman
        M.Vieselman
        2 Posts
        ACCEPTED ANSWER

        Re: IP Replication V7000

        ‏2014-04-15T07:34:04Z  in response to LeonelSAA

        Thanks for the reply.

        I was wondering what goes wrong with my configuration in this case.

        We have node 1 and 2 port 1 of both storwizes in a replication vlan 

        with cluster ip 10.25.40.33 and on the other side ip 10.25.40.36

        on node 1 port 1  on both storwizes are ip addresses 10.25.40.34 and 10.25.40.37

        and node 2 port 1 on both storwizes are ip addresses 10.25.40.35 and 10.25.40.38 defined.

        Port 2 of both storwizes are defined in seperate vlan's because of management and cannot be connected to each other.

        but when i delete the secondary mgmt management ip addresses so technicly disable the 2nd nic on both controllers it stil doesnt work.

         

        So i wonder if you build up replication with port 1 so port 2 makes a connection

         

        • LeonelSAA
          LeonelSAA
          18 Posts
          ACCEPTED ANSWER

          Re: IP Replication V7000

          ‏2014-05-09T11:17:47Z  in response to M.Vieselman

          Hi M.Viesel,

          Can you describe in details your scenario? If you are using both storages in local or remote site? Are you zoned both storages in same zone(in local site case)? If you are use in local site and same sanswitch, separates zones for both storages making invisibles. are you defined/changed Cluster IP and port by CLI, because you are using two IP Management.etc. Please describe your scenario in details.

          waiting.

          regards.

  • Azizio
    Azizio
    1 Post
    ACCEPTED ANSWER

    Re: IP Replication V7000

    ‏2014-04-29T08:44:43Z  in response to LeonelSAA

    Hi Leonel,

    I tried to configure the way you suggested but still no luck.

    This is my Configuration:

    Storage1 - Cluster IP (10.1.250.17)
                        Node1 Service IP - 10.1.250.19            Node2 Service IP - 10.1.250.20

    Storage2 - Cluster IP (10.1.250.170)
                        Node1 Service IP - 10.1.250.172           Node2 Service IP - 10.1.250.173

    Ethernet port config is shown in the screenshots below.

    Updated on 2014-04-29T08:45:28Z at 2014-04-29T08:45:28Z by Azizio
  • LeonelSAA
    LeonelSAA
    18 Posts
    ACCEPTED ANSWER

    Re: IP Replication V7000

    ‏2014-05-09T10:49:59Z  in response to LeonelSAA

    Hi Azizio,

    Sorry for the delay.

    I was out in a project!

    To solve your issue you need follow these steps:

    1. configure IP Cluster: you already configured. ( In what port you configured?)

    2. configure two IP remote copy one for each node ( in same port that ip cluster) on the same copy group ID (e.g. if ID 1, put the same Id on node 1 and 2) for the redundancy.

    3. If you are zoned both storages, you should unzoning the storages because the partnership FC has priority to listen partnership. separate both storages to different zone invisible if is on the local san or same sanswitch to make tests.

    4. Make a Ping partnership before create "partnership".

    Give me a feedback soon as possible.

    regards
     

  • LeonelSAA
    LeonelSAA
    18 Posts
    ACCEPTED ANSWER

    Re: IP Replication V7000

    ‏2014-10-16T12:26:45Z  in response to LeonelSAA

    Hi all,

    I implemented IP Replication in lab environment between V7000 and V3700, worked good without trouble, now I implemented in production environment I have big issue with connectivity, and every time I get notification saying " Connection to a configured remote cluster has been lost ". And replication running very slowly.

    Our network team guaranteed me that issue is not from network because they tested connectivity of dedicated link with 65000000 pings and reply from 99% success.

    Can anybody say me in fact what's happening?

    I don't know what can i do now, because software seems be good.

     

    Best Regards