Sun Microsystems, Inc.  Sun System Handbook - ISO 3.4 June 2011 Internal/Partner Edition
   Home | Current Systems | Former STK Products | EOL Systems | Components | General Info | Search | Feedback

Asset ID: 1-71-1003487.1
Update Date:2010-12-22
Keywords:

Solution Type  Technical Instruction Sure

Solution  1003487.1 :   Installing a Second Switch and System Controller (SSC) in the Sun Fire[TM] B1600 Intelligent Shelf  


Related Items
  • Sun Fire B1600 Blade System Chassis
  •  
Related Categories
  • GCS>Sun Microsystems>Servers>Blade Servers
  •  

PreviouslyPublishedAs
204905


Description

This document is intended to provide a comprehensive action plan for the installation of a second switch and system controller (SSC) module into a B1600 intelligent shelf which was supplied with only one SCC.

It includes an explanation of how to upgrade the system controller firmware as this may be required in order to complete the installation procedure.

Documentation for the �B1600 Intelligent Shelf is available at the following URL.

http://www.sun.com/products-n-solutions/hardware/docs/Servers/Workgroup_Servers/Sun_Fire_Blade_Platform/Sun_Fire_b1600

This includes the "Sun Fire B1600 Blade System Chassis Administration Guide" which contains some useful diagrams.

Information on the switch and system controller itself, including part numbers, may be found in the "Sun System Handbook" at the following URL.

https://support.oracle.com/handbook_private/Systems/SunFireB1600/SunFireB1600.html

Only the system controller (SC) component of an SSC has failover capability, the additional internal switch provides a redundant network path.

Note: During this procedure the system controller (SC) will need to be reset but this will not affect access to the production blades.



Steps to Follow
Checking the Existing System Controller Firmware
Connect to the SC using the serial management port which is located to the left of the ethernet network management (NETMGT) port on the currently installed SSC module as you look at the back of the shelf.

You need to login on the SC with a user that has been granted reset permission. If you are unsure you can check this using the 'usershow' command, e.g.,

sc>usershow
username         permissions  password
--------         -----------  ---------
admin            aucr         Assigned
useradm          -uc-         Assigned

This shows the letter "r" in the permissions column for those users who have been granted reset permission.

Version 1.2 of the SC firmware is the earliest version which supports SC failover so you must make sure that you have this or a later version installed.

You can check the version running on the currently installed SC using the 'showsc' command, e.g.,

sc>showsc
Sun Advanced Lights Out Manager for Blade Servers 1.0
Copyright 2003 Sun Microsystems, Inc. All Rights Reserved.
ALOM-B 1.0
Release: 1.0.67
Parameter                               Running Value      Stored Value
-------------------------------------------------------------------------------
Bootable Image :                   1.0.67 (Feb 10 03)
Current Running Image :              1.0.67 (Feb 10 03)
SC IP address:                      129.156.195.236    129.156.195.236
SSC0/SC (Active) IP private address:    129.156.195.237    129.156.195.237
SSC1/SC (Standby) IP private address:   0.0.0.0            0.0.0.0

Note that output from this command has been truncated for the sake of brevity and it is the value shown as "Release" that we are interested in here, i.e., 1.0.67 in this example.

If the SC firmware needs to be upgraded you can download the latest "Sun Fire B1600 Blade Platform System Controller 1.2 Firmware" by logging in at:

http://wwws.sun.com/software/download/network.html

At the time of writing version 1.2.7 is available from this location.

Alternatively, contract customers can look at the 115314 patch readme or obtain patch 115314 by logging in and selecting

the "MOS Patch Portal"

i.e.,

Patch-ID# 115314-01
Synopsis: Sun-Fire B1600 System Controller (1.2) Firmware Update

At the present time version 1.2.8 is contained in this patch.

You will need to download the firmware into the /tftpboot directory of a TFTP server which is capable of communication with the SSC NETMGT port. This will be assigned the "SC IP address" reported by the showsc command and is the address which will failover between the system controllers once the new SCC has been installed.

If you need to setup a TFTP server for this purpose you should consult the following Infodoc.

Document ID: ID19272
Synopsis: How to setup TFTP and verify that it works

Once the TFTP server has been setup you need to carry out the steps outlined in the following example.

# cd /tftpboot
# unzip SunFireB1600-sc-v1.2.7.zip
Archive:  SunfireB1600-sc-1.2.7.zip
 inflating: SunFireB1600-sc-v1.2.7.flash
inflating: copyright

or, if you have downloaded patch 115314,

# unzip 115314-01.zip
Archive:  115314-01.zip
creating: 115314-01/
inflating: 115314-01/copyright
inflating: 115314-01/SunFireB1600-sc-v1.2.8.flash
inflating: 115314-01/README.115314-01
# ls 115314-01
README.115314-01              SunFireB1600-sc-v1.2.8.flash copyright

It is important to understand that the TFTP transfer of the firmware will take place over the ethernet, and not via the serial port, so you need to check that you can access the system controller's NETMGT port from the TFTP server, e.g.,

# ping 129.156.195.236
129.156.195.236 is alive

Carry out the following steps on the existing system controller,

Copy over the new firmware.

For example, using the firmware contained in patch 115314-01,

sc>flashupdate -v -y -s 129.156.87.175 -f 115314-01/SunFireB1600-sc-v1.2.8.flash ssc0/sc SSC0/SC: Preparing to flashupdate. flashupdate: verifying image, please wait flashupdate: erasing segment 37 programming address ffeeffed SSC0/SC: flashupdate complete. The system must be reset (using resetsc) for the new image to be loaded

Note: The IP address 129.156.87.175 is assigned to an interface on the TFTP server and the path specified (-f) is relative to the /tftpboot directory.

Reset the existing SSC to run the new firmware.

sc> resetsc -y

Shutting down SSC0/SC ... Shutting down SSC0/SC ... this may take several seconds, please wait. SSC0/SC shutdown. Resetting ...

Once the system controller has reset, log back in and verify that the SSC is running the new firmware version, e.g,

username: admin password: ********* sc>showsc Sun Advanced Lights Out Manager for Blade Servers 1.2 Copyright 2003 Sun Microsystems, Inc. All Rights Reserved. ALOM-B 1.2 Release: 1.2.8

Installing the Additional Switch & System Controller

The unused SCC slot will contain a filler panel which needs to be removed, as follows, before you can install the additional SSC.

Remove the locking pin, squeeze the green ejector handle and gently pull it away from the center of the panel in order to unlatch the filler. Then remove the filler panel by sliding it out of the chassis using the handles provided. Insert the new SCC into the empty slot and secure by closing the adjacent ejector handle and replacing the locking pin.

You will need to power on the new SSC from the existing system controller, e.g.,

sc>poweron ssc1
SSC1: Powered on.
Jul 06 07:56:20: MINOR: SSC1: Active LED state changed to ON.
Jul 06 07:56:21: MINOR: SSC1: Powered on.
Jul 06 07:56:37: MINOR: Peer SC SSC1 is now online
Jul 06 07:56:37: MINOR: SSC1/SWT: Environmental monitoring enabled.
sc>

In order to check its status you can use the 'showlogs' and/or 'showenvironment' commands, e.g.,

sc>showlogs ssc1
Jul 06 07:56:20: MINOR: SSC1: Active LED state changed to ON.
Jul 06 07:56:21: MINOR: SSC1: Powered on.
Jul 06 07:56:37: MINOR: Peer SC SSC1 is now online
Jul 06 07:56:37: MINOR: SSC1: Environmental monitoring enabled.
sc>showenvironment ssc1
============ Environmental Status ============
System Temperatures (Celsius)   Current   Status
-----------------------------------------------------------------------------
SSC1       /TSensor/ambient     52        OK
System Voltages (Volts)         Current   Status
------------------------------------------------------------------------------
SSC1       /VSensor/12VInput    11.75     OK
SSC1       /VSensor/BOD         8.96      OK
SSC1       /VSensor/AOD         8.68      OK
SSC1       /VSensor/VCore       1.75      OK
System Fans (RPM)               Current   Status
------------------------------------------------------------------------------
SSC1       /Tacho/Fan0          9854      OK
SSC1       /Tacho/Fan1          10150     OK

Having upgraded the original system controller you should now also upgrade the firmware on the system controller which you have just installed. In order to do this it must become the 'active' system controller.

Type 'setfailover' at the system controller prompt and 'y' in response to the subsequent prompt, e.g.,

sc>setfailover
SSC0 is in Active Mode.
SSC1 is in Standby Mode.
Are you sure you want to failover to SSC1
All connections and user sessions will now be lost on SSC0 (y/n)  y

You will then see the system controller display the following.

Shutting down SSC0/SC ...
SSC0/SC shutdown.
Resetting ...

Further messages are displayed but not shown here, however eventually you will see the following.

System Controller in SSC0 is now in Standby Mode.

You can now connect to the serial port on SSC1, just as you did for SSC0, and repeat steps 1 to 3 above replacing "ssc0" with "ssc1".

Once you have completed this you will probably want to make SSC0 the active controller again, i.e.,

sc>setfailover
SSC0 is in Standby Mode.
SSC1 is in Active Mode.
Are you sure you want to failover to SSC0
All connections and user sessions will now be lost on SSC1 (y/n)  y

and eventually you will see:

System Controller in SSC1 is now in Standby Mode.

To confirm which system controller is active at any given time you can telnet to the failover "SC IP address" and use the 'showplatform' command, e.g.,

# telnet 129.156.195.236
Trying 129.156.195.236...
Connected to 129.156.195.236.
Escape character is '^]'.
Sun Advanced Lights Out Manager for Blade Servers 1.2
Copyright 2003 Sun Microsystems, Inc. All Rights Reserved.
ALOM-B 1.2
username: admin
password: *********
sc>showplatform
Domain    Status
--------  -------------------
SSC0/SWT  OS Running
SSC1/SWT  OS Running
SSC0/SC   OS Running (Active)
SSC1/SC   OS Running

Note that, once again, the command output is not shown in full.

Plumbing the B100s Additional GigaSwift[TM] Interface

Each B100s blade has a GigaSwift 'ce0' interface which will connect to the switch on SSC0. Now that a second switch is installed you can plumb up the 'ce1' interface in order to provide dual pathing through switch SSC1.

For example,

b1600-s0# ifconfig -a
lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
ce0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
inet 129.156.106.160 netmask ffffff00 broadcast 129.156.106.255
ether 0:3:ba:29:f7:72
b1600-s0# ifconfig ce1 plumb 129.156.106.161 netmask 255.255.255.0 broadcast + up
b1600-s0# ifconfig -a
lo0: flags=1000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
ce0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
inet 129.156.106.160 netmask ffffff00 broadcast 129.156.106.255
ether 0:3:ba:29:f7:72
ce1: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 4
inet 129.156.106.161 netmask ffffff00 broadcast 129.156.106.255
ether 0:3:ba:29:f7:73

Note: You will need to configure the switch component in SSC1 appropriately as the configuration does not get copied across from the original switch during the installation procedure.



Product
Sun Fire B1600 Blade System Chassis

Internal Comments

Although, as stated above, you should be able to carry out this
procedure without disrupting access to the production blades you
should be aware of the following bug report which is dated 30th
March 2004.



Bug Id: 4877079
Synopsis: Blades power down when SC is reset



Author: Neil Cheney




May 2004


configure, SC, SCC
Previously Published As
76622

Change History
Date: 2004-12-08
User Name: 31620
Action: Approved
Comment: Verified Metadata
Verified still correct for audience - contract
Checked review date 2006-05-28
Checked for TM - no new ones needed
Fixed a straange "underlining" problem in last half of the document
Publishing
Version: 4
Date: 2004-12-07
User Name: 130090
Action: Approved
Comment: Author requested Andy MacBain to perform final review
Version: 0
Date: 2004-12-06
User Name: 113400
Action: Rejected
Comment: Returning to John Adams for review/follow-up aciton. See Steve Gilliss comments from 12/06/04.
Version: 0
Date: 2004-12-06
User Name: 113400
Action: Add Comment
Comment: John,

To assign this article to Andy M. for review, refer to the following:

Q. Can I assign an article to be reviewed by a specific person? If so, how?

A. Voyager allows a specific person to review an article provided, of course, that the person is a member of the queue associated with the
article's Tech Team.

To assign an article to a single tech team queue member, open the article in Voyager and select "Workflow" from the All Actions drop-down
menu. At the resulting screen, select the appropriate "Workflow Stage" and add adequate instruction and/or comments to the "Comments" text
box. Finally, choose "Select Recipient".

From the "Select Recipient" screen, choose the desired reviewer from the "User" drop-down menu. Next, ensure Comments are present and then
hit "Submit". Voyager will send an email, with your comments, to the person you chose to review your article.

Note: If the person you would like to review your article does not appear on the User drop-down menu, then that individual is not a
registered member of the queue or you may have indadvertently assigned your article to the wrong Tech Team.

Thanks,

Steve Gilliss
KE
x29024
Version: 0
Date: 2004-12-06
User Name: 113400
Action: Accept
Comment:
Version: 0
Date: 2004-12-06
User Name: 130090
Action: Approved
Comment: I had previously approved this document.
Only change is author's name has been added.

At author's suggestion request that Andy MacBain
handles next stage.
Version: 0
Date: 2004-12-06
User Name: 130090
Action: Accept
Comment:
Version: 0
Date: 2004-11-22
User Name: 116519
Action: Add Comment
Comment: Document has detail and conveys information quite good, however documentation has syntax errors. Please review the following suggestions:

1. "...It includes an explanation of how to upgrade the system controller firmware, as this may be required in order to complete the installation procedure...." (adding comma to delineate ).

2. at the following URL.
Suggest using 'URL; ' denoting continuation of line and not termination by adding semi-colon.

3."... ethernet, and not via the serial port, so you need to check that you can access the system controller's NETMGT port from the TFTP server, e.g., ..."
Suggest using, "ethernet and not via the serial port, therefore; one must verify that you can access system controller's ....."
Version: 0
Date: 2004-11-22
User Name: 121361
Action: Approved
Comment: Added my name to the internal section as the original document author.
Version: 0
Date: 2004-11-22
User Name: 121361
Action: Update Started
Comment: Need to add my name to the internal section.
Version: 0
Date: 2004-07-14
User Name: C139400
Action: Approved
Comment: added tmark, keywords, tech area, corrected title, spelling, format, usage, punctuation and spacing, applied KE procedures
Version: 0
Date: 2004-07-14
User Name: C139400
Action: Accepted
Comment:
Version: 0
Date: 2004-07-08
User Name: 130090
Action: Approved
Comment: passed tech review
Version: 0
Date: 2004-07-06
User Name: 121361
Action: Approved
Comment: John, I have now made your suggested changes.
Version: 0
Date: 2004-06-15
User Name: 130090
Action: Add Comment
Comment: One thing that I should have added to my comments was the need to
power on the new ssc after it is installed

poweron ssc1
showenvironment or showlogs ssc1 (to check)
Version: 0
Date: 2004-06-15
User Name: 130090
Action: Rejected
Comment: Hi Neil,

As I said in my comments it looks pretty good to me.
Let me know if you agree with my points.
Version: 0
Date: 2004-06-15
User Name: 130090
Action: Add Comment
Comment: Hi Neil - looks pretty good to me. -just a few thoughts.

1) I would prefer the term second SSC to " redundant SSC " The system controller
component is redundant but strictly speaking the switch is not. The customer does
not have to employ the switch as a redundant configuration. Indeed in some of the
documents I saw referenced a so called "sandwhich config "for the blades

redundant references are in title and this line.

Only the System Controller (SC) component of an SSC has failover capability, the additional internal switch provides a redundant network path.



2) When mentioning plumbing in ce1 interfaces it may be an idea to also state that
customer has to configure the switch component of SSC1 to the desired configuration. ie the switch configuration does not get copied across and may indeed be different.
Version: 0
Date: 2004-06-04
User Name: 121361
Action: Approved
Comment: John, a first attempt. Let me know your thoughts.
Version: 0
Date: 2004-05-28
User Name: 121361
Action: Created
Comment:
Version: 0
Product_uuid
10bec5e4-5865-11d6-9ffc-c65b6cd3fd7d|Sun Fire B1600 Blade System Chassis

Attachments
This solution has no attachment
  Copyright © 2011 Sun Microsystems, Inc.  All rights reserved.
 Feedback