Sun Open Telecommunications Platform 1.1 Installation and Administration Guide

Pair+N Topology Administration

This section provides the procedures for adding a new OTP host to a Pair+N clustered OTP system, and for repairing an OTP host within a Pair+N clustered OTP system.

The following topics are discussed:

To Add an OTP Host That Is Not Connected to Shared Storage

Reinstall the Solaris 10 Update 2 operating system on the OTP host as described in Installing Solaris 10 Update 2 and the Remote Agent on the OTP Hosts.

Add the OTP host back into the cluster configuration as described in To Add a Host to the Existing Cluster

Note –
As the OTP host will not be part of the resource group, steps to add the OTP host to the resource group need not be performed.

Install OTP on the OTP host.
- To install OTP on the OTP hosts using the command line, see Chapter 4, Installing the Open Telecommunications Platform For the First Time Using the Command Line
- To install OTP on the OTP hosts using the graphical user interface, see Chapter 5, Installing the Open Telecommunications Platform For the First Time Using the Graphical User Interface
- To install OTP on the OTP hosts using a production standalone or clustered OTP system, see Chapter 6, Installing the Open Telecommunications Platform Using the Provisioning Service On an Existing OTP System

To Repair an OTP Host That Is Not Connected to Shared Storage

In this procedure, pcl8-ipp2 is the OTP host that is being repaired. Substitute your own host information.

Check the disks connected to the OTP host that needs to be repaired.

# scconf -pvv | grep pcl8-ipp2 | grep Dev
(dsk/d10) Device group node list:                pcl8-ipp2
(dsk/d9) Device group node list:                 pcl8-ipp2

Remove the disks connected to the OTP host.

# scconf -c -D name=dsk/d10,localonly=false

# scconf -c -D name=dsk/d9,localonly=false

# scconf -r -D name=dsk/d10

# scconf -r -D name=dsk/d9

Place the OTP host to be repaired in maintenance state.

# scconf -c -q node=pcl8-ipp2,maintstate

Remove the transport information of the OTP host from cluster configuration.
1. Check the transport information.
  # scconf -pvv | grep pcl8-ipp2 | grep Transport Transport cable: pcl8-ipp2:bge1@0 switch1@3 Enabled Transport cable: pcl8-ipp2:ce1@0 switch2@3 Enabled
2. Remove the related transport.
  
  # scconf -c -m endpoint=pcl8-ipp2:bge1,state=disabled
  
  # scconf -c -m endpoint=pcl8-ipp2:ce1,state=disabled
  
  # scconf -r -m endpoint=pcl8-ipp2:bge1
  
  # scconf -r -m endpoint=pcl8-ipp2:ce1
  
  # scconf -r -A name=bge1,node=pcl8-ipp2
  
  # scconf -r -A name=ce1,node=pcl8-ipp2

Remove the OTP host from authentication list.

# scconf -r -T node=pcl8-ipp2

Remove the OTP host from the host list.

# scconf -r -h node=pcl8-ipp2

Make sure that the OTP host is completely removed from the cluster configuration.

If you see any output for the following command, revisit the above steps to make sure all the steps are executed properly.

# scconf -pvv | grep pcl8-ipp2

Add the OTP host back into the cluster configuration as described in To Add an OTP Host That Is Not Connected to Shared Storage for more information.

Note –
As the OTP host will not be part of the resource group, steps to add the OTP host to the resource group need not be performed.

To Repair an OTP Host That Is Connected to Shared Storage

In this procedure, pcl8-ipp3 is the OTP host that is being repaired. Substitute your own host information.

Move all the resource groups to another OTP host in the resource group list.

# scswitch -z -g otp-system-rg -h otherotphost

Remove the disks connected to the OTP host.

Check the disks connected to the OTP host.

# scconf -pvv | grep pcl8-ipp3 | grep Dev
(dsk/d8) Device group node list:                 pcl8-ipp3
(dsk/d7) Device group node list:                 pcl8-ipp3
(dsk/d6) Device group node list:                 pcl8-ipp1, pcl8-ipp3
(dsk/d5) Device group node list:                 pcl8-ipp1, pcl8-ipp3
(dsk/d4) Device group node list:                 pcl8-ipp1, pcl8-ipp3
(dsk/d3) Device group node list:                 pcl8-ipp1, pcl8-ipp3

Remove the local disks.

# scconf -c -D name=dsk/d8,localonly=false

# scconf -c -D name=dsk/d7,localonly=false

# scconf -r -D name=dsk/d8

# scconf -r -D name=dsk/d7

Remove the shared disks.

# scconf -r -D name=dsk/d5,nodelist=pcl8-ipp3

# scconf -r -D name=dsk/d6,nodelist=pcl8-ipp3

# scconf -r -D name=dsk/d4,nodelist=pcl8-ipp3

Shut down the OTP host.

# shutdown -y -g 0 -i 0

Place the OTP host in maintenance mode.

# scconf -c -q node=pcl8-ipp3,maintstate

Remove the transport information.
1. Check the transport information.
  
  # scconf -pvv | grep pcl8-ipp3 | grep Transport
  Transport cable: pcl8-ipp3:bge1@0 switch1@2 Enabled Transport cable: pcl8-ipp3:ce1@0 switch2@2 Enabled
2. Remove the transport information.
  
  # scconf -c -m endpoint=pcl8-ipp3:bge1,state=disabled
  
  # scconf -c -m endpoint=pcl8-ipp3:ce1,state=disabled
  
  # scconf -r -m endpoint=pcl8-ipp3:bge1
  
  # scconf -r -m endpoint=pcl8-ipp3:ce1
  
  # scconf -r -A name=bge1,node=pcl8-ipp3
  
  # scconf -r -A name=ce1,node=pcl8-ipp3

Remove the quorum disk.

# scconf -r -D name=dsk/d3,nodelist=pcl8-ipp3

# scconf -r -q globaldev=d3

Note –
If you perform this procedure on a three-host cluster, you will need to establish quorum before running the above procedure. Otherwise, you will get the following error:

# scconf -r -h node=pcl8-ipp3
scconf: Failed to remove node (pcl8-ipp3) - quorum could be compromised. scconf: All two-node clusters must have at least one shared quorum device.

Remove the host from the authentication list.

# scconf -r -T node=pcl8-ipp3

Remove the host from the host list.

# scconf -r -h node=pcl8-ipp3

Make sure that the OTP host is completely removed from the cluster configuration.

If you see any output for the following command, revisit the above steps to make sure all the steps are executed properly.

# scconf -pvv | grep pcl8-ipp3

Add the OTP host back into the cluster configuration as described in To Add a Host to the Existing Cluster for more information.