During deployment, step 11 fails with ERROR : No SSD device partition seen

Asset ID:	1-72-1952122.1
Update Date:	2016-10-31
Keywords:

Solution Type Problem Resolution Sure

Solution 1952122.1 : During deployment, step 11 fails with ERROR : No SSD device partition seen

Applies to:

Oracle Database Appliance X4-2 - Version All Versions and later
Information in this document applies to any platform.
This doc only applies to ODA X4-2 systems with the HGST Sunset Cove SSD's installed in them.

Symptoms

DO NOT run the following steps for an existing ODA installation UNLESS you are RE-IMAGING the ODA (re-deploying).

During the initial ODA deployment, the following error is encountered in Step 11:

SetupStorage
ERROR : No SSD device partition seen
in /dev/mapper on hostname
Refer to the log files for more details

INFO : Step 11 SetupStorage
INFO : Setting up storage devices...
INFO : Verifying storage topology, it may take few minutes...
SUCCESS: Verification of storage topology on ODA01 ...
SUCCESS: Verification of storage topology on ODA02 ...
INFO : Verifying nickname setting, it may take few minutes...
SUCCESS: Verification of nickname setting on default storage shelf...
...
INFO : Wait for 20 seconds for oakd resource to clean up before starting oakd.
..
INFO : Going to start oakd...
INFO : oakd should have started on ODA01...
...
INFO : Getting list of multipath disks...
INFO : Constructed diskstring = <(/dev/mapper/HDD_E0_S08_1133526648p1,/dev/map
per/HDD_E0_S05_1134385736p1,/dev/mapper/HDD_E0_S07_1134488428p1,/dev/mapper/HDD_
E0_S04_1134676900p1,/dev/mapper/HDD_E0_S09_1134676916p1,/dev/mapper/HDD_E0_S00_1
...
...
...
INFO : Running as root: /usr/bin/ssh -l root ODA01 /opt/oracle/oak/onecmd
/tmp/multipathdreload.sh
INFO : Running as root: /usr/bin/ssh -l root ODA02 /opt/oracle/oak/onecmd
/tmp/multipathdreload.sh
INFO : Checking multipath devices creation on ODA01 with 1 JBOD...

SUCCESS: Correct number(20) of HDD multipath devices seen in /etc/multipath.conf on oda01
ERROR : No SSD multipath devices seen in /etc/multipath.conf on oda01 <<<<<<<<<<<

INFO : Time in SetupStorage is 690 seconds.
Exiting...

Even though they physically exist
---------------------------------------------

[root@ODA01 bin]# ./oakcli show storage

==== BEGIN STORAGE DUMP ========
Host Description: Xen:SUN SERVER X4-2
Total number of controllers: 2
        Id         = 0
        Serial Num = 500605b001234123
        Vendor     = LSI Logic
        Model      = SGX-SAS6-EXT-Z
        FwVers     = 11.05.03.00
        strId      = mpt2sas:00:15.0

        Id         = 1
        Serial Num = 500605b001234123
        Vendor     = LSI Logic
        Model      = SGX-SAS6-EXT-Z
        FwVers     = 11.05.03.00
        strId      = mpt2sas:00:16.0

Total number of expanders: 2
        Id         = 0
        Serial Num = 500605b001234123
        Vendor     = ORACLE
        Model      = DE2-24P
        FwVers     = 0018
        strId      = Secondary
        WWN        = 5080020001ba7b3e

        Id         = 1
        Serial Num = 500605b001234123
        Vendor     = ORACLE
        Model      = DE2-24P
        FwVers     = 0018
        strId      = Primary
        WWN        = 5080020001ba803e

Total number of PDs: 24
        /dev/sda        HITACHI           HDD 900gb slot: 0 exp: 0
...
...
        /dev/sdar       HITACHI           HDD 900gb slot: 19 exp: 0
       /dev/sdas    HGST         SSD    200gb slot: 20 exp: 0 <<<
        /dev/sdat      HGST         SSD      200gb slot: 21 exp: 0 <<<
        /dev/sdau     HGST         SSD      200gb slot: 22 exp: 0 <<<
        /dev/sdav        HGST         SSD      200gb slot: 23 exp: 0 <<<
==== END STORAGE DUMP =========

Changes

Deployments / Installation

Cause

Systems were shipped with Software version 12.1.2.0, which is the initial release that provides support for the Sunset Cove HGST SSD's, part number 7094126.

If customers install an older version than 12.1.2.0, during deployment, step 11 fails, which is the step the SetupStorage part of deployment. It fails to recognize the new SSD part, and does not allow deployment to continue.

Solution

Download and install the latest iso image, 12.1.2.x which contains information for newer Disks and FW versions.

SW version 12.1.2.0 and above will have support for the HGST 200GB Sunset Cove SSDs.

If there is a special sitaution where the customer MUST install a lower version than 12.1.2.X and above, the customer will need to replace the existing /opt/oracle/oak/conf/validation_props.xml on both nodes before starting the configurator for deployment. This will allow the deployment to pass step 11, which is the step to SetupStorage. It is still not recommended to re-image the ODA nodes to an older version than what they were shipped with.

If you must retain 2.10 or lower please open an SR with support.

Pre 12.1 Steps

1) Download the validation_props.xml.new file -- Attached to the note 1952122.1

2) On the ODA as root make a copy of the current validation_props.xml file found in the following directory

cd /opt/oracle/oak/conf/

and rename the copy validation_props.xml.original = Do this on Both Nodes

/opt/oracle/oak/conf/cp validation_props.xml validation_props.xml.original

3) Copy the new downloaded validation_props.xml.new file to the same directory on BOTH Nodes

4) remove the current validation_props.xml and copy validation_props.xml,new to be validation_props.xml

/opt/oracle/oak/conf/cp validation_props.xml.new validation_props.xml

5) re-run the deployment using as root in the following directory on only Node0

cd /opt/oracle/oak/onecmd

Once in the directory run the following command as root:

Continue the Deployment installation from Step 11 = SetupStorage

For Bare Metal installs

./GridInst.pl -s 11

If you are using ODAVP run

./GridInst.pl -o -s 11

6) If this completes successfully then run the rest of the steps on only Node0

From /opt/oracle/oak/onecmd directory

For Bare Metal installations / deployments

./GridInst.pl -r 12-23

For ODAVP installations / deployments

./GridInst.pl -o -r 12-23

Attachments

This solution has no attachment