Browse Prior Art Database

Automatic network path failover for systems management of a service processor

IP.com Disclosure Number: IPCOM000196868D
Publication Date: 2010-Jun-18
Document File: 3 page(s) / 84K

Publishing Venue

The IP.com Prior Art Database

Abstract

In order to provide network path redundancy when managing Power System via it's Flexible Service Processor (FSP) it is necessary to either provide two management entities (one on each network path), or dual network paths from one management entity. In the case of DPSM (Director Power System Manager is a project where the legacy Hardware Management Console Software system management logic is repackaged into a Director plugin to extend and enhance the IBM Director functionality) the latter is required - that is, provide two network paths from one management entity due to an Active/Passive HA (high availability) model for the management entity itself. To further complicate the dual network path from one management entity problem, the management entity (DPSM) requires an active socket to the service processor at all times. However, the Service Processor can only accept one active network connection from a particular management entity at a given point in time, so it is not possible for the management entity to maintain 2 active connections. In addition, the IP addresses on a given FSP must be on separate subnets.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 53% of the total text.

Page 1 of 3

Automatic network path failover for systems management of a service processor

In order to provide network path redundancy when managing Power System via it's Flexible Service Processor (FSP) it is necessary to either provide two management entities (one on each network path), or dual network paths from one management entity. In the case of DPSM (Director Power System Manager is a project where the legacy Hardware Management Console Software system management logic is repackaged into a Director plugin to extend and enhance the IBM® Director functionality) the latter is required - that is, provide two network paths from one management entity due to an Active/Passive HA (high availability) model for the management entity itself.

To further complicate the dual network path from one management entity problem, the management entity (DPSM) requires an active socket to the service processor at all times. However, the Service Processor can only accept one active network connection from a particular management entity at a given point in time, so it is not possible for the management entity to maintain 2 active connections. In addition, the IP addresses on a given FSP must be on separate subnets.

It is possible for the management entity to obtain both valid IP addresses of a given service processor through an SLP (Service Location Protocol) query. The management entity will store both IP addresses, and establish an active connection to one of the addresses. When the connection goes down, the management entity will automatically attempt to reconnect using an algorithm to establish the connection through an available network path.

Rather than relying upon complex network hardware and in order to work within existing customer setups, the fail-over algorithm will exist in the DPSM software. The following mechanism is used:

Using an SLP query, DPSM is able to get a list of all IP addresses configured on a given FSP, and will

store both of these away.

The management entity will then try and establish a connection through the IP address on which it

received the SLP query response.
If this fails, it will try the alternate one. If this further fails, it will stop trying and wait for user


1.

this h...