Topic
  • 3 replies
  • Latest Post - ‏2013-03-19T09:59:51Z by chriscanto
SystemAdmin
SystemAdmin
4779 Posts

Pinned topic Redundancy ethernet connection Storwize V7000

‏2013-03-18T16:56:33Z |
Hello everyone. I have Storwize V7000: 2 file modules, 1 control enclosure and 2 extension enclosure. Also two G8124E Rackswitch. I'd like to setup redundancy configuration and done ethernet connectivity like on a picture.
The diagram shows the ethernet-ports and connections between they. Also screenshot with IP-addresses attached. I think that everything is properly connected and configured, but it seems that not all.

Problems:

1. When I turn off SW1, the file modules can't see control enclosure, although both port 2 of canisters control enclosure are connected to SW2 and both port 4 on file modules are connected to SW2 too. May be I must configure IP-address on both port 2 of canisters control enclosure? But How? There is no such in web-gui.

2. On servers (Win 2012 Server) enabled teaming on interfaces. When I detach both cables form file modules (port 3) to SW1, there is a problem to connect server to file modules. May be I must setup some like VLAG on switches?

Help please! What am I doing wrong?
Updated on 2013-03-19T09:59:51Z at 2013-03-19T09:59:51Z by chriscanto
  • chriscanto
    chriscanto
    292 Posts

    Re: Redundancy ethernet connection Storwize V7000

    ‏2013-03-18T21:33:48Z  
    To respond to the first problem:

    You're right, with that configuration then the second 1Gb Ethernet ports on the V7000 node canisters are not being used for management access. My understanding is that to configure the system to use failover is a two step process:

    1. Configure a management IP address on port 2 of the V7000 block storage system, e.g.
    
    chsystemip -clusterip 192.168.7.200 -gw 192.168.2.4 -mask 255.255.248.0 -port 2
    


    2. Update the configuration on the file modules to make use of this second management address, e.g.
    
    chstoragesystem --ip2 192.168.7.200
    


    Try that and see how it goes.

    Hopefully someone familiar with the NAS side of a Unified configuration will be able to comment on the second problem you mention.
  • SystemAdmin
    SystemAdmin
    4779 Posts

    Re: Redundancy ethernet connection Storwize V7000

    ‏2013-03-19T09:38:09Z  
    To respond to the first problem:

    You're right, with that configuration then the second 1Gb Ethernet ports on the V7000 node canisters are not being used for management access. My understanding is that to configure the system to use failover is a two step process:

    1. Configure a management IP address on port 2 of the V7000 block storage system, e.g.
    <pre class="jive-pre"> chsystemip -clusterip 192.168.7.200 -gw 192.168.2.4 -mask 255.255.248.0 -port 2 </pre>

    2. Update the configuration on the file modules to make use of this second management address, e.g.
    <pre class="jive-pre"> chstoragesystem --ip2 192.168.7.200 </pre>

    Try that and see how it goes.

    Hopefully someone familiar with the NAS side of a Unified configuration will be able to comment on the second problem you mention.
    Thanks for recommendations, I tried on control enclosure
    > lssystemip
    cluster_id cluster_name location port_id IP_address subnet_mask gateway IP_address_6 prefix_6 gateway_6
    00000200A0A24BCE subt-stor local 1 192.168.7.211 255.255.248.0 192.168.2.4
    00000200A0A24BCE subt-stor local 2 192.168.7.223 255.255.248.0 192.168.2.4

    on file modules:

    $ lsstoragesystem
    name primaryIP secondaryIP id
    StorwizeV7000 192.168.7.211 192.168.7.211 00000200A0A24BCE

    ok, change secondary IP:

    $ chstoragesystem --ip2 192.168.7.223

    $ lsstoragesystem
    name primaryIP secondaryIP id
    StorwizeV7000 192.168.7.211 192.168.7.223 00000200A0A24BCE

    System is good (degraded network ignore):

    $ lshealth
    Host Sensor Status Value
    mgmt001st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    mgmt002st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    MGMTNODE_REPL_STATE OK OK
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    V7000 CLUSTER OK The component cluster is running OK
    IO_GRP OK The component io_grp is running OK
    MDISK OK The component mdisk is running OK
    NODE OK The component node is running OK

    Then I detached cables form port 1 both canisters:

    $ lshealth
    Host Sensor Status Value
    mgmt001st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    mgmt002st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    MGMTNODE_REPL_STATE OK OK
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    V7000 CONNECTION ERROR Failed to open the SSH channel

    $ lsstoragesystem
    name primaryIP secondaryIP id
    StorwizeV7000 192.168.7.211 192.168.7.223 00000200A0A24BCE
    EFSSG1000I The command completed successfully.
    http://7803272.ibm$ ping 192.168.7.211
    PING 192.168.7.211 (192.168.7.211) 56(84) bytes of data.
    From 192.168.7.215 icmp_seq=1 Destination Host Unreachable
    From 192.168.7.215 icmp_seq=2 Destination Host Unreachable
    ^C
    • 192.168.7.211 ping statistics ---
    4 packets transmitted, 0 received, +2 errors, 100% packet loss, time 3759ms
    pipe 2
    http://7803272.ibm$ ping 192.168.7.223
    PING 192.168.7.223 (192.168.7.223) 56(84) bytes of data.
    64 bytes from 192.168.7.223: icmp_seq=1 ttl=64 time=0.472 ms
    64 bytes from 192.168.7.223: icmp_seq=2 ttl=64 time=0.094 ms
    ^C
    • 192.168.7.223 ping statistics ---
    3 packets transmitted, 3 received, 0% packet loss, time 2064ms
    rtt min/avg/max/mdev = 0.094/0.221/0.472/0.177 ms

    I tried to reboot, did not help. What else can you recommend?
  • chriscanto
    chriscanto
    292 Posts

    Re: Redundancy ethernet connection Storwize V7000

    ‏2013-03-19T09:59:51Z  
    Thanks for recommendations, I tried on control enclosure
    > lssystemip
    cluster_id cluster_name location port_id IP_address subnet_mask gateway IP_address_6 prefix_6 gateway_6
    00000200A0A24BCE subt-stor local 1 192.168.7.211 255.255.248.0 192.168.2.4
    00000200A0A24BCE subt-stor local 2 192.168.7.223 255.255.248.0 192.168.2.4

    on file modules:

    $ lsstoragesystem
    name primaryIP secondaryIP id
    StorwizeV7000 192.168.7.211 192.168.7.211 00000200A0A24BCE

    ok, change secondary IP:

    $ chstoragesystem --ip2 192.168.7.223

    $ lsstoragesystem
    name primaryIP secondaryIP id
    StorwizeV7000 192.168.7.211 192.168.7.223 00000200A0A24BCE

    System is good (degraded network ignore):

    $ lshealth
    Host Sensor Status Value
    mgmt001st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    mgmt002st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    MGMTNODE_REPL_STATE OK OK
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    V7000 CLUSTER OK The component cluster is running OK
    IO_GRP OK The component io_grp is running OK
    MDISK OK The component mdisk is running OK
    NODE OK The component node is running OK

    Then I detached cables form port 1 both canisters:

    $ lshealth
    Host Sensor Status Value
    mgmt001st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    mgmt002st001 HOST_STATE OK OK
    SERVICE OK All services are running OK
    CTDB OK CTDBSTATE_STATE_ACTIVE
    GPFS OK ACTIVE
    SCM OK SCM system running as expected
    MGMTNODE_REPL_STATE OK OK
    TB_STATE OK OK
    NETWORK WARNING Network interfaces are degraded
    V7000 CONNECTION ERROR Failed to open the SSH channel

    $ lsstoragesystem
    name primaryIP secondaryIP id
    StorwizeV7000 192.168.7.211 192.168.7.223 00000200A0A24BCE
    EFSSG1000I The command completed successfully.
    http://7803272.ibm$ ping 192.168.7.211
    PING 192.168.7.211 (192.168.7.211) 56(84) bytes of data.
    From 192.168.7.215 icmp_seq=1 Destination Host Unreachable
    From 192.168.7.215 icmp_seq=2 Destination Host Unreachable
    ^C
    • 192.168.7.211 ping statistics ---
    4 packets transmitted, 0 received, +2 errors, 100% packet loss, time 3759ms
    pipe 2
    http://7803272.ibm$ ping 192.168.7.223
    PING 192.168.7.223 (192.168.7.223) 56(84) bytes of data.
    64 bytes from 192.168.7.223: icmp_seq=1 ttl=64 time=0.472 ms
    64 bytes from 192.168.7.223: icmp_seq=2 ttl=64 time=0.094 ms
    ^C
    • 192.168.7.223 ping statistics ---
    3 packets transmitted, 3 received, 0% packet loss, time 2064ms
    rtt min/avg/max/mdev = 0.094/0.221/0.472/0.177 ms

    I tried to reboot, did not help. What else can you recommend?
    Unless anyone else on the forum has a better idea, I'd suggest opening a software problem ticket with IBM to ask why the file modules don't appear to be making use of the available/redundant management interface configured on the V7000 control enclosure :-/