Jump to content United States-English
HP.com Home Products and Services Support and Drivers Solutions How to Buy
» Contact HP
More options
HP.com home
HP OpenCall SS7 platform Operations Guide: For Release 3.1 on Linux > Chapter 3 Configuring the Platform

Configuring Relocatable IP Addresses using PINS

» 

Technical documentation

Complete book in PDF
» Feedback
Content starts here

 » Table of Contents

 » Index

This IP network high availability feature is integrated in HP OpenCall SS7 by means of PINS (Plug-In Network Sensor).

Description of PINS

PINS manages a single “floating” IP address and re-maps the address to one of the four LAN ports on one of the hosts in a 2-host platform. The floating IP address is used by remote applications communicating with a local application. This enables the IP network to view the 2-host platform as a single IP host. PINS monitors the state of the LAN port associated with the virtual IP address. If the LAN where the virtual IP address is mapped fails, PINS re-assigns the IP address to another LAN port on the same front-end host when available (local LAN switch), or on the other host after a PINS process switchover (remote LAN switch).PINS provides high availability for an IP address over any LAN except M3UA and OpenCall LANs (also referred to as HA LANs). PINS uses ARP (Address Resolution Protocol) to broadcast an update request (of the Internet address translation tables) to the network neighborhood. Whereas a local LAN switch is transparent to remote applications, a remote LAN switch requires the remote application to reconnect to the platform and to re-initiate any on-going transactions.

HA Functionality of PINS

PINS is an active/standby process managed by the Fault Tolerance Controller (FTC). Within the HA mechanism supported by the platform, the behavior of the PINS process includes:

  • Re-assignment of the virtual IP addresses to a PINS configured LAN port on the host where the PINS is active.

  • Respawning of the PINS process by the FTC when the PINS process has failed.

  • Triggering a local or remote LAN switchover upon LAN failure.

  • Broadcast of re-ARP requests to the network to overwrite existing ARP tables on a remote machine that are still referring to the MAC address of the previous LAN interface used by the PINS process.

Figure 3-1 PINS on an HP OpenCall SS7 Platform

PINS on an HP OpenCall SS7 Platform

Requirements for PINS

For PINS to function properly, you must take account of the following:

  • The LANs controlled by PINS should be plugged on independent switch hubs to avoid Single Point Of Failures.

  • Two LAN interfaces must be available and configured for PINS usage.

  • Neighborhood machines must be able to receive and accept re-ARP requests. Therefore, network elements (such as switch hubs) that mask such requests should not be used.

PINS Configuration

You configure PINS as follows:

  1. On the Platform Manager, edit the file /etc/opt/OC/PINS/pins.conf (for the running configuration) by following the template provided in this file.

    In the section [IP_1], add the following parameters:

    • Nic = <ethx> [,priox]

      where <ethx> is the Linux distribution name (eth0, eth1, ...) of LAN access on which a static IP address has been configured.

      This is optionally followed by its priority (value 1 for the highest priority and 2 for the lowest one).

      The Nic parameter MUST be set at least one time, and may be repeated:

      Nic = ethx [,priox]
      Nic = ethy [,prioy]

    • IpAddress = <relocatable IP> <floating Ip netmask>

      A relocatable IP address associated to the LAN access, together with its subnet mask.

    • LanSwitch = <value>

      where <value> can be:

      • 'L' for local LAN switch. (1-host platform only).

      • 'LR' for both local and remote LAN switch. (2-host platform only).

      • 'R' for only remote LAN switch (2-host platform only).

      CAUTION: All other parameters and sections are reserved for internal use. Do not change or remove them.
      NOTE: For a 2-host platform, the PINS configuration should be installed in the same way, and at the same time on both hosts, to ensure that the dual NICs have the same Linux distribution name. For example, if on one host the 2 NICs are named eth12 and eth14, the 2 NICs on the second host must be named eth12 and eth14.
  2. To activate the PINS process, you must edit the ftc.conf file in the directory:

    • - for the running configuration:
      /etc/opt/OC/HA

    • - for saved configurations:
      /var/opt/OC/working_config/<platform_id>/HA

    Add the following lines to the [pins] section:

    [pins]
    RunDirectory = /var/opt/OC/logs
    RunString = /opt/OC/lbin/pinsrun
    ProcessType = HA
    ProcessGroup = 0

  3. Run cfgPropagate to propagate the configuration.

  4. Stop and restart the Front-End platform(s).

Deactivating the PINS Process

Remove the [pins] section from the ftc.conf file then stop and restart the platform.The PINS feature is still configured and can be re-activated by repeating Steps 2 to 4 in the above procedure.

Monitoring and Operating PINS

Local Switch

The command:

PINS_localSwitch -h<host> IP_1 <NIC>

requests a local LAN switch to the PINS running on the specified host (the default is the local host), on the specified LAN access to the specified NIC. The LAN access and NIC names must be the same as in the pins.conf file.

Remote Switch

A remote switch can be forced from the Platform Monitor by requesting a PINS process switch.

PINS HA Status

The floating IP address is available only for the active PINS process as seen from the Platform Monitor. It can be seen as the alias of a PINS LAN using the netstat -rn shell command on the host where the active PINS is located.

The command:

PINS_getState -h<host> IP_1

displays on the standard output the configured NIC states on the specified host (the default is the local host). The active NIC is marked with a star ('*').

Printable version
Privacy statement Using this site means you accept its terms Feedback to webmaster
© Hewlett-Packard Development Company, L.P.