Storage2011. 9. 19. 09:37

Note: RLM is required.

If filer is in a clustered state, go to one of the filers and issue the  command to check the current status:
 

>cf hw_assist status

Local Node(netapp1) Status:
  Inactive: filera not monitoring alerts  from partner(filerb)
  Fail to bind UDP socket to port 4444 IP address  10.61.69.125
 

 

To configure hw_assist, you will need to determine  which interface you will use for the configuration. Interface e0m is preferred  when setting up hw_assist to separate traffic, but e0a is the default if e0m is  not configured and online. If you will be using e0a, you will need to make sure  name resolution is working. No additional configuration is needed to use  e0m.

  1. Determine the interface to use:
     

        filera> ifconfig  -a
        e0M: flags=948043 mtu 1500
        inet 10.61.69.122 netmask 0xffffff00  broadcast 10.61.69.255
      partner e0M (not in use)
      ether  00:a0:98:0c:3d:a6 (auto-100tx-fd-up) flowcontrol full
        e0a: flags=948043  mtu 1500
        inet 10.61.69.120 netmask 0xffffff00 broadcast 10.61.69.255
       ether 00:a0:98:0c:3d:a4 (auto-1000t-fd-up) flowcontrol full

         filerb> ifconfig -a
        e0M: flags=948043 mtu 1500
        inet  10.61.69.127 netmask 0xffffff00 broadcast 10.61.69.255
      partner e0M (not in  use)
      ether 00:a0:98:0c:40:3e (auto-100tx-fd-up) flowcontrol full
         e0a: flags=108042 mtu 1500
        inet 10.61.69.125 netmask 0xffffff00 broadcast  10.61.69.255
      ether 02:a0:98:0c:40:3e (auto-1000t-fd-up) flowcontrol  full
     
  2. Turn on hw_assist on for both heads (if needed):
      > options  cf.hw_assist.enable on
     
  3. Set the partner address for hw_assist:
      filera> options  cf.hw_assist.partner.address 10.61.69.127
      Validating the new hw-assist  configuration. Please wait...
      filera>
      filerb> options  cf.hw_assist.partner.address 10.61.69.122
      Validating the new hw-assist  configuration. Please wait...
      filerb>
     
  4. Set the hw_assist port for both heads (if needed):
      > options  cf.hw_assist.partner.port 4444
      Validating the new hw-assist configuration.  Please wait...
      >
     
  5. Check hw_assist status for both heads (may take a few minutes):
        filera> cf hw_assist status
      Local Node(filera) Status:
      Active:  filera monitoring alerts from partner(filerb)
      port 4444 IP address  10.61.69.122
      Partner Node(filerb) Status:
      Active: filerb monitoring  alerts from partner(filera)
      port 4444 IP address 10.61.69.127

         filerb> cf hw_assist status
      Local Node(filerb) Status:
      Active:  filerb monitoring alerts from partner(filera)
      port 4444 IP address  10.61.69.127
      Partner Node(filera) Status:
      Active: filera monitoring  alerts from partner(filerb)
      port 4444 IP address 10.61.69.122
     

The status will not change to monitoring alerts from the partner until the  next keep alive ping has been received. Once the ping has been received, the  status will change to active for both the local and partner node.

As an alternative, test the configuration with the test command:
   > cf hw_assist test
    cf hw_assist: test command successful


 

Posted by [TheWon]