Friday, October 12, 2012

Upgrading Netapp Data Ontap 7.3.4 to Ontap 8.1


Use My AutoSupport to improve the Fitness of your NetApp Storage Systems - http://support.netapp.com/NOW/asuphome/

Hostname: FBTFAS01 Serial No: 123456789 Model No: FAS2040 Upgrade plan generated on: 14-Feb-2013 02:27:47 GMT  Version: 5.1.6 1

1. Upgrade Advisor Plan

This upgrade plan is based on AutoSupport received on Feb 30 16:34:53 2012

1.1. Warnings for Current OS 7.3.4 to Target OS 8.1

1.2. Upgrade Plan for upgrading from Current OS 7.3.4 to Target OS 8.1 Serial No Hostname Model No 123456789 FBTFAS01 FAS2040

Related to Warnings Description

Upgrade You should run AggrSpaceCheck Tool to verify that the system has the requisite
available free space needed to upgrade successfully to 7.3 or later (without
disabling space guarantees or offlining aggregates/volumes).

Upgrade If you are running SnapDrive software on Windows hosts connected to this filer, the
following versions are supported with ONTAP 8.1:

Note: If you are using SnapDrive with SnapManager for
Exchange/SQL/Oracle/SAP/SharePoint/Domino, please check the compatibility
between your SnapDrive and SnapManager before implementing the upgrade.

Upgrade NetApp Global Services recommends running perfstat during a typical usage time to
save a performance baseline prior to an upgrade in case it is needed. This will take
about 30 mins of run time.

Upgrade For each HA pair, you should plan for approximately 30 minutes to complete
preparatory steps, 60 minutes to perform the upgrade, and 30 minutes to complete
post-upgrade steps.

Upgrade You must ensure that CPU utilization does not exceed 50% before beginning a
NDU upgrade

Upgrade Please check the risk details for your system here that might impact your upgrade.

Upgrade This system may experience the issue described in Burt 455754, 'Boot Device may
be reformatted during Data ONTAP upgrades': 455754.

Upgrade Please check the risk details for your system here that might impact your upgrade.
Downgrade If you are running SnapDrive software on Windows hosts connected to this filer, the
following versions are supported with ONTAP 8.1:

Note: If you are using SnapDrive with SnapManager for
Exchange/SQL/Oracle/SAP/SharePoint/Domino, please check the compatibility
between your SnapDrive and SnapManager before implementing the upgrade.

Downgrade This system may experience the issue described in Burt 455754, 'Boot Device may
be reformatted during Data ONTAP upgrades': 455754.

Steps Upgrade Plan

Use My AutoSupport to improve the Fitness of your NetApp Storage Systems - http://support.netapp.com/NOW/asuphome/

Hostname: FBTFAS01  Serial No: 123456789 Model No: FAS2040 Upgrade plan generated on: 14-Feb-2013 02:27:47 GMT Version: 5.1.6 2

1 Be sure to read the Release Notes as well as the Upgrade Guide of the Data ONTAP 8.1 for
important information and technical detail before beginning your upgrade.
Before upgrading Data ONTAP, monitor CPU and disk utilization for 30 seconds by
entering the following command at the console of each storage controller:
sysstat -c 10 -x 3

The values in the CPU and Disk Util columns are strongly recommended not to exceed
50% for all ten measurements reported. Ensure that no additional load is added to the
storage system until the upgrade completes.

Download perfstat and run it on a client as follows:

perfstat -f filername -t 4 -i 5 > perfstatname.out

Save this output file for a couple of weeks after the upgrade is complete

2 Ensure all SnapDrive for Windows clients are running one of the supported versions for 8.1:

3 Download the system files for 8.1 (81_e_image.zip) from the Support Site. Be sure to
download the system files that match your node model.

If you are performing a Data ONTAP NDU (or backout), you must perform this step on
both nodes before performing the takeover and giveback steps.

4 Verify the checksum of the image file with the value on the Support site.

5 Connect to the console of the node and trigger an AutoSupport:
FBTFAS01> options autosupport.doit "starting_NDU 8.1"

6 Contact NetApp Support and check /etc/messages for any obvious errors; e.g. disk
errors, firmware errors, etc.
Using Windows: Map C$ to the Windows Host
Browse the etc folder
Open file 'messages' using WordPad. If failed disks are found, it is recommended they
be removed before upgrading.

7 Back up the etc\hosts and etc\rc files in Windows to a temporary directory.

8 Copy the system image file (81_e_image.zip) to the /etc/software directory on the node.
From a Windows box as an Administrator:
o Map the C$ share to a Windows drive letter (for example X:).
o Copy the image file to X:\etc\software

9 Install the system file via the software command:

FBTFAS01> software update 81_e_image.zip -r

If you are performing a Data ONTAP NDU (or backout), you must perform this step on
both nodes before performing the takeover and giveback steps.
10 Check to see if the boot device has been properly updated:

FBTFAS01> version -b

The primary kernel should be 8.1.

11 Terminate CIFS on the node to be taken over ( FBTFAS02 ):

FBTFAS02> cifs terminate

Use My AutoSupport to improve the Fitness of your NetApp Storage Systems - http://support.netapp.com/NOW/asuphome/

Hostname: FBTFAS01 Serial No: 123456789 Model No: FAS2040 Upgrade plan generated on: 14-Feb-2013 02:27:47 GMT Version: 5.1.6 3

12 Takeover and giveback sequence between nodes

Terminate CIFS on the node to be taken over (FBTFAS02):

FBTFAS02> cifs terminate

From FBTFAS01:, take over the data service from the partner node FBTFAS02:
FBTFAS01> cf takeover

Wait 8 minutes before proceeding to the next step.

Doing so ensures the following conditions:
- The node that has taken over is serving data to the clients.
- Applications on the clients have recovered from the pause in I/O that occurs during
takeover.
- Load on the storage system has returned to a stable point.
- Multipathing (if deployed) has stabilized.

After FBTFAS02: reboots and displays "waiting for giveback", give back the data
service:

FBTFAS01> cf giveback
Terminate CIFS on the node to be taken over ( FBTFAS01 ):

FBTFAS01> cifs terminate
From the newly upgraded node FBTFAS02, take over the data service from

FBTFAS01:

FBTFAS02cf takeover -n
Halt, and then restart the first node:

FBTFAS01> halt

FBTFAS01> bye
Wait 8 minutes before proceeding to the next step.
Doing so ensures the following conditions:
- The node that has taken over is serving data to the clients.
- Applications on the clients have recovered from the pause in I/O that occurs during
takeover.
- Load on the storage system has returned to a stable point.
- Multipathing (if deployed) has stabilized.
After the first node reboots and displays "waiting for giveback", give back the data
service:

FBTFAS02> cf giveback
Attention: The giveback is not initiated and an error message is returned if any
conditions such as the following are detected:
- open client sessions (such as CIFS sessions)
- long-running operations
- operations that cannot be restarted (such as tape backup or SyncMirror
resynchronization)
- error conditions (such as disk connectivity mismatch between the nodes)

If giveback is not initiated, complete the following steps:
Use My AutoSupport to improve the Fitness of your NetApp Storage Systems - 

http://support.netapp.com/NOW/asuphome/
Hostname: FBTFAS01 Serial No: 123456789 Model No: FAS2040 Upgrade plan generated on: 14-Feb-2013 02:27:47 GMT Version: 5.1.6 4

1.3. Backout Plan for upgrading from Current OS 7.3.4 to target OS 8.1

1. Address the condition described in the error message, ensuring that any
Identified operations are terminated gracefully.

2. Enter the cf giveback command with the -f option:
cf giveback -f
For more information about the -f option, see the cf(1) man page.

13 Verify the upgrade completed successfully to 8.1:

FBTFAS01> version

14 Upgrade FBTFAS02 (if not already done)

15 Connect to the console of the node and trigger an AutoSupport

FBTFAS01> options autosupport.doit "finishing_NDU 8.1"

Steps Backout Plan Description

1 Ensure all SnapDrive for Windows clients are running one of the supported versions for
7.3.4:

2 Download the system files for 7.3.4 (734_setup_e.exe) from the Support Site. Be sure to
download the system files that match your node model.
If you are performing a Data ONTAP NDU (or backout), you must perform this step on
both nodes before performing the takeover and giveback steps.

3 Verify the checksum of the image file with the value on the Support site.

4 Connect to the console of the node and trigger an AutoSupport using the following
command:

FBTFAS01> options autosupport.doit "Reverting to 7.3.4"

5 Contact NetApp Support and check /etc/messages for any obvious errors; e.g. disk
errors, firmware errors, etc.
Using Windows: Map C$ to the Windows Host

Browse the etc folder
Open file 'messages' using WordPad. If failed disks are found, it is recommended they
be removed before upgrading.

6 Back up the etc\hosts and etc\rc files in Windows to a temporary directory.

7 Copy the system image file (734_setup_e.exe) to the /etc/software directory on the
node. From a Windows box as an Administrator:
o Map the C$ share to a Windows drive letter (for example X:).
o Copy the image file to X:\etc\software

8 Install the system file via the software command:

FBTFAS01> software update 734_setup_e.exe -r
If you are performing a Data ONTAP NDU (or backout), you must perform this step on
both nodes before performing the takeover and giveback steps.

9 Check to see if the boot device has been properly updated:

FBTFAS01> version -b
The primary kernel should be 7.3.4.

10 Disable the data service:

FBTFAS01> cf disable

Use My AutoSupport to improve the Fitness of your NetApp Storage Systems - http://support.netapp.com/NOW/asuphome/

Hostname: FBTFAS01 Serial No: 123456789 Model No: FAS2040 Upgrade plan generated on: 14-Feb-2013 02:27:47 GMT Version: 5.1.6 5

1.4. Anomalies Plan for upgrading from Current OS 7.3.4 to target OS 8.1

11 Run the revert_to command:

FBTFAS01> revert_to 7.3.4

At the end of revert, the system will automatically halt itself and drop to a loader prompt.
Once at a loader prompt, you can issue either bye or reboot to bring the system back up

12 Reboot the node:

FBTFAS01> reboot

13 Enable the cluster (if both sides are downgraded):

FBTFAS01> cf enable

14 Verify the upgrade completed successfully to 7.3.4:

FBTFAS01> version

* Anomalies Description

1 Find the latest disk firmware for disk model X410_S15K7288A15 on Support site.

2 Find the latest disk firmware for disk model X287_HVIPC288A15 on Support site.

2 comments:

  1. Hi Rasool,

    I have couple of scenarios here,

    1) Node A and Node B are in HA pair, I have upgraded Node A and just before starting upgrade on Node B there is a disk failure on Node B. What is the course of action here? - Should I downgrade Node A version?

    2) Node A and Node B are in HA pair, I have upgraded Node A and just before starting upgrade on Node B, there is a disk failure on Node A as well as Node B. What is the course of action?

    Regards,
    Viswanath Kotte

    ReplyDelete
  2. Hi Viswa,

    It depends upon the type of upgrade you are performing whether Non-Disruptive or Disruptive , because If you are doing non-disruptive upgrade, the requirement is that the failed disks needs to be replaced before beginning the upgrade, this is because failed disk may prevent giveback operations and can introduce loop instability throughout the storage system.

    Also in this case have to check the RAID GROUP status and also check if spare disks are available, If it says rebuilt in progress actually you can proceed ( Risking it on your own, If i were you i will replace disk and then proceed )

    In case of disk failure on both the nodes still apply the same as above.

    Hope it helps !

    Thanks,
    Nayab

    ReplyDelete