StorageTek Virtual Storage Manager System 6 (VSM6)
 
StorageTek Virtual Storage Manager System 7 (VSM7)
 

PLA-Support>Sun Systems>Sun_Other>Sun Collections>SN-OTH: TAPE-CAP VCAP
 

In this Document

Goal

Solution

References

Oracle Confidential PARTNER - Available to partners (SUN).
Reason: Internal field process

Applies to:

Goal

 Field procedure to replace a VSM6 or VSM7 SSD drive.

Solution

 DISPATCH INSTRUCTIONS
   WHAT SKILLS DOES THE FIELD ENGINEER/ADMINISTRATOR NEED?: VSM6 trained, T4 server, Solaris 11
   TIME ESTIMATE: 30-120 minutes
   TASK COMPLEXITY: 3

This process address replacing a Persistent Storage SSD.  A ZIL (ZFS Intent Log) SSD is part of stpool0100 and is replaced using the stpool drive replacement script as documented in the service guide and doc ID 1533980.1.  Please review the attached document How to Determine ZIL SSD

FIELD ENGINEER/ADMINISTRATOR INSTRUCTIONS:  

PROBLEM OVERVIEW: How to replace a persistent storage SSD drive in a VSM6 or VSM7 server.

   There are two types of procedures to replace a Persistent Storage SSD on a VSM6 or VSM7.

Scripted Process:
--Requires minimal Field Engineer input.
--Takes 20-30 minutes.
--Requires both nodes to be operational
--Requires the VSM to be offline to the customer in order to reboot.

Command Line Process:
--Requires more Field Engineer input to determine the failed SSD and process the replacement.
--Takes 60-90 minutes.
--Requires only the node with the failed SSD to be offline.

WHAT STATE SHOULD THE SYSTEM BE IN TO BE READY TO PERFORM THE RESOLUTION ACTIVITY?:

   Prepare affected VSM6 or VSM7server for servicing:

If only a single node must be taken offline:
--1. The Host CHPIDs attached to the node need to be taken offline to the host.
--2. The RTDs attached to the node need to be taken offline to VTCS.
--3. The Clinks and Rlinks for the entire VSM need to be taken offline.

If the entire VSM needs to be offline:
--1. The customer should follow their own process to take the VSM offline to all host activity.
--2. This would include varying all RTDs, Clinks/Rlinks, and the VTSS offline.

WHAT ACTION DOES THE FIELD ENGINEER/ADMINISTRATOR NEED TO TAKE?:

   Prior to the scheduled maintenance:
     1. Collect the latest or daily Support File Bundle (SFB) from each node at /var/opt/StatusService/output/archive.
     2. Review vsm_Healthcheck.log.txt for any previously unknown problems.
         The Healthcheck would be from a time after the SSD failure but before the repair is initiated.
     3. Run the command scstat and verify that all required resources are Online and Operational.
         The exception would be the failed SSD.

SSD GFR Script Process

For all microcode 7.1.1.xx.000 and higher, review the instruction and information README_gfrSsd 7.1.1.xx

--To initiate the SSD GFR script, enter the command:

     /opt/vsm/bin/gfrSsd.pl

For all microcode 7.0.0.12.000 and lower, the script needs to be installed.  Otherwise, use the command line process.

--Download the file gfrSsd_700xx.zip.  This file contains the script called gfrSsd.pl and instructions and information README_gfrSsd 700xx. 

--The README can be extracted separately on a PC, but the entire zip file needs to be moved to the VSM to be unzipped to prevent corrupting the script file.   

***WARNING***

 If the failed SSD is on a VSM6 Node 2, and running any 7.0.0.xx.000 code, workaround procedures must be followed.  

--Please download the zip file attachment GFR ssd02xx script workaround to collect the workaround instructions.

 NOTE: If the gfrSsd.pl script fails due to insufficient information, it will be necessary to try the Command Line process.

SSD GFR Command Line Process

The attached process uses the same command structure as is documented in the VSM Service guides and is reformatted to help prevent user errors.

--Download the zip file Persistent Storage SSD Replacement Command Line. The file can be unzipped on a PC to extract the instruction file.

--The enclosed file ssd.fomat is a text file to be copied to the VSM home directory.

***WARNING***

If the failed SSD is on a VSM6 Node 2 and running any 7.0.0.xx.000 code, workaround procedures must be followed.

--Please download the zip file attachment VSM6 GFR ssd020x command line workaround to collect the workaround instructions.

OBTAIN CUSTOMER ACCEPTANCE

WHAT ACTION DOES THE FE/CUSTOMER NEED TO TAKE TO RETURN THE SYSTEM TO AN OPERATIONAL STATE:

--Before returning the VSM to the customer, the FE should run the command scstat and verify that all resources are Online and Fully Operational.

--Have the customer vary the VSM online to the host. This would include the VTSS, Channel paths, RTDs, Clinks/Rlinks taken offline prior to the maintenance.

REFERENCE INFORMATION:

NOTE: When replacing an SSD with one of a larger size microcode prerequisites must be met.  Please reference the following table for details.

*NOTE:

If the VTSS is not at supporting code for the available SSD, it will be necessary to upgrade microcode before replacing the SSD.
Do NOT attempt to replace the SSD if not at supporting microcode.

This solution has no attachment