Skip to content

Support_multiple_hardware_control_points

ligc edited this page Jul 30, 2015 · 6 revisions

Table of Contents

{{:Design Warning}}

Support for multiple hardware control points

The existing xCAT hardware control structure does not support more than one hardware control point for each node, the multiple hardware control points support is needed in some configuration scenarios, such as the HMC redundancy and FSP redundancy, and the Direct Attach is also requiring to try different methods(fsp-api, ASMI and HMC) before indicating that the operation failed. This mini design will describe how to support multiple hardware control points for each node.

The main idea is to specify comma separated hardware control points with the node attribute ppc.hcp, then xCAT will try the hcps one by one until find a hardware control point that can succeed. If any of the hardware control point fails due to any reason, then sequentially try the next hcp specified in the hcp attribute. If all of the hcps failed, then tell the users why it fails on different hardware control points. When trying the hcps, the type of hcp can be determined by the hardware type of the hcp itself, the hcp has to be defined as a node in xCAT database anyway. Please notice that: if the hcp's type is fsp or bpa, it will run the command through fsp-api firstly; If it succeeds, the command will return; If it fails, the command will try the ASMI method secondly.

This feature can be used by several xCAT components and features:

1. fsp-api and HMC: the fsp-api does not support all necessary features for now, so we need HMC to perform some operations are not supported by fsp-api.

2. The fsp/bpa redundancy support that we plan to implement later this year can take advantage of this solution.

3. This can be treated as a new feature that xCAT can support multiple hardware control points for each node, system p does support two HMCs connected to the FSP/BPA, then xCAT can be used in this configuration scenario with this new feature. I believe this new general feature can be used in some other scenarios such as the blade management module redundancy.

Question: Will performance be a problem if xCAT needs to try the hardware control points one by one?

There are some restrictions in the support for multiple hardware control points:

1.Multiple hardware control points only support for system P machine.

2.For rcons, the value of the node's cons attribute should be set to "multiple"

3.specify comma separated hardware control points with the node attribute ppc.hcp.

 NOTE: When running command with noderange, it's required that the hcps' types of the nodes in the noderange are in the same order. Four examples as the following are correct: 

  node     hcp 
 node1  fsp1,hmc1
 node2  fsp2,hmc2
 node3  fsp3
 node4  fsp4,hmc4

,

 node      hcp 
 node1  hmc1,fsp1
 node2  hmc2,fsp2
 node3  hmc3
 node4  hmc4,fsp4

,

  node     hcp 
 node1  fsp1,fsp12
 node2  fsp2,fsp22
 node3  fsp3
 node4  fsp4,fsp42

and

 node      hcp 
 node1  hmc1,hmc12
 node2  hmc2,hmc22
 node3  hmc3
 node4  hmc4,hmc42

but the following two examples are wrong:

 node      hcp 
 node1  fsp1,hmc1
 node2  hmc2,fsp2
 node3  hmc3
 node4  hmc4,fsp4

and

 node     hcp
 node1  hmc1,fsp1
 node2  hmc2,fsp2
 node3  fsp3
 node4  hmc4,fsp4

4.For hierarchy environment, it's required that the service node for one node's hcps should be the same one.

News

History

  • Oct 22, 2010: xCAT 2.5 released.
  • Apr 30, 2010: xCAT 2.4 is released.
  • Oct 31, 2009: xCAT 2.3 released. xCAT's 10 year anniversary!
  • Apr 16, 2009: xCAT 2.2 released.
  • Oct 31, 2008: xCAT 2.1 released.
  • Sep 12, 2008: Support for xCAT 2 can now be purchased!
  • June 9, 2008: xCAT breaths life into (at the time) the fastest supercomputer on the planet
  • May 30, 2008: xCAT 2.0 for Linux officially released!
  • Oct 31, 2007: IBM open sources xCAT 2.0 to allow collaboration among all of the xCAT users.
  • Oct 31, 1999: xCAT 1.0 is born!
    xCAT started out as a project in IBM developed by Egan Ford. It was quickly adopted by customers and IBM manufacturing sites to rapidly deploy clusters.
Clone this wiki locally