Please login or register. June 23, 2017, 10:08:29 AM

Author Topic: Swapping Persistent IP netmask  (Read 6135 times)

0 Members and 1 Guest are viewing this topic.

seamus

  • New Member
  • *
  • Posts: 3
  • Karma: +0/-0
Swapping Persistent IP netmask
« on: September 16, 2011, 03:29:53 PM »
I wonder if anyone has a quick solution to this.

I set up a simple 2 node cluster which all worked ok. Howerer, shortly after I noticed that I had configured the wrong subnet masks ugh.
I had configured 255.255.255.0
Should be          255.255.255.192
In order to rectify, i brought down the cluster on both sides and altered the subnet mask of the bott addresses on both sides using smitty tcpip. When bringing the cluster back up i now see.

en0: flags=1e080863,480<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,CHECKSUM_OFFLOAD(ACTIVE),CHAIN>
        inet 192.168.32.1 netmask 0xffffffc0 broadcast 192.168.32.63
        inet 10.160.32.33 netmask 0xffffff00 broadcast 10.160.32.255        inet 10.160.32.39 netmask 0xffffffc0 broadcast 10.160.32.63
         tcp_sendspace 262144 tcp_recvspace 262144 rfc1323 1
en1: flags=1e080863,480<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,CHECKSUM_OFFLOAD(ACTIVE),CHAIN>
        inet 192.168.33.1 netmask 0xffffffc0 broadcast 192.168.33.63
        inet 10.160.34.37 netmask 0xffffff00 broadcast 10.160.34.255        inet 10.160.34.38 netmask 0xffffffc0 broadcast 10.160.34.63
         tcp_sendspace 262144 tcp_recvspace 262144 rfc1323 1
en2: flags=1e080863,480<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,CHECKSUM_OFFLOAD(ACTIVE),CHAIN>
        inet 10.160.36.35 netmask 0xffffffc0 broadcast 10.160.36.63
         tcp_sendspace 262144 tcp_recvspace 262144 rfc1323 1
lo0: flags=e08084b,c0<UP,BROADCAST,LOOPBACK,RUNNING,SIMPLEX,MULTICAST,GROUPRT,64BIT,LARGESEND,CHAIN>
        inet 127.0.0.1 netmask 0xff000000 broadcast 127.255.255.255
        inet6 ::1%1/0
         tcp_sendspace 131072 tcp_recvspace 131072 rfc1323 1


ie the persistant addresses have not changed. Is there any way to change these witout rebuilding the cluster???

Michael

  • Administrator
  • Hero Member
  • *****
  • Posts: 1039
  • Karma: +0/-0
Re: Swapping Persistent IP netmask
« Reply #1 on: September 18, 2011, 01:42:19 PM »
yes, there is sure to be a way. Low level is to bring the cluster down and use odmchange commands to change the cluster definitions.
However, I expect the easier solution will be to remove the persistent addresses (as they are not required), synchronize, and then add them back.

I do not have the smit dialogs memorized anymore, so I cannot do better than this right now. However, it is an interesting question and I'll research it "soon".

seamus

  • New Member
  • *
  • Posts: 3
  • Karma: +0/-0
Re: Swapping Persistent IP netmask
« Reply #2 on: September 20, 2011, 08:38:13 AM »
nope removing the addresses from the HA config and performing a sync does not remove them from the interface. the only way i have found so far is to trash the interface after the above and use cfgmgr to bring it back in....messy!

would be interested in the odm change method though if its proved reliable..would need to make sure that HA entries and TCP entries dont conflict i guess.

Michael

  • Administrator
  • Hero Member
  • *****
  • Posts: 1039
  • Karma: +0/-0
Re: Swapping Persistent IP netmask
« Reply #3 on: October 09, 2011, 09:03:15 PM »
Sorry I took so long to get back on this. I'll have to setup a test cluster to learn why the normal way is not working. I asked a colleague who does these things "regularly" and his immediate response was to remove the persistent definition, and rebuild.

re: ODM method: his suggestion is to make a snapshot, copy it (using cp), edit the copy - removing all references to the persistent "interface", and apply the copy you just edited. I have been more "cowboy" performing odmchange commands with the nodes down.

Advantage of snapshot method is that you should be able to make the changes live. If is just a snapshot after all. ;)