Technical documentation
CRITICAL PRE-
UPGRADE HOST
RECOMMENDATIONS
     © Copyright 2021 Hewlett Packard Enterprise Development LP. The information contained herein is subject to
     change without notice. The only warranties for Hewlett Packard Enterprise products and services are set forth in the
     express warranty statements accompanying such products and services. Nothing herein should be construed as
     constituting an additional warranty. Hewlett Packard Enterprise shall not be liable for technical or editorial errors or
     omissions contained herein.
     This document contains confidential and/or legally privileged information. It is intended for Hewlett Packard
     Enterprise and Channel Partner Internal Use only. If you are not an intended recipient as identified on the front
     cover of this document, you are strictly prohibited from reviewing, redistributing, disseminating, or in any other way
     using or relying on the contents of this document.
     Trademark acknowledgments if needed. All third-party marks are property of their respective owners.
     Sept 2024
Document type                                                                                     Page
TECHNICAL DOCUMENT
CRITICAL PRE-UPGRADE HOST RECOMMENDATIONS
Note:
       Please be informed that it is critical to follow the pre upgrade host recommendations
        to avoid service disruption
       The below SPOCK link aids in checking the compatibility of the drivers:
        https://www.hpe.com/Storage/Spock/
       Multipathing policy should be set to round robin
       For arrays attached to AIX and Solaris hosts, it is mandatory to disable port failover
        feature during upgrade from any versions to 3.3.1 MU5.
       AIX host-persona with AIX legacy should be updated to AIX-Alua before upgrading to
        target 3.3.2EMU1.
        Refer link : https://spock.corp.hp.com/SPOCK/Content/ExportPDFView.aspx?
        Id=137635&typeId=2
       Refer below customer advisory for Brocade switch with FOS version 9.X. ‘HPE 3PAR
        StoreServ 8000, 9000, and 20000 Storage - Arrays With 16Gbps FC Host Ports Might Report
        TE_DIF_CRC and “Logical Block Address out of range” Errors’
        https://support.hpe.com/hpsc/doc/public/display?docId=emr_na-a00135503en_us
 Document type                                                                                           Page
 TECHNICAL DOCUMENT
1. Provide health check summary:
CheckUpgrade for System 2M284900HR
== showversion -b ==
Release version 3.3.1.648 (MU5)
Patches: P126,P132,P135,P140,P146,P150,P151,P155,P156,P170,P172,P173,P175,P177,P180
Component Name                     Version
CLI Server                         3.3.1.728   (P177)
CLI Client                         3.3.2.159
System Manager                     3.3.1.725   (P173)
Kernel                             3.3.1.648   (MU5)
TPD Kernel Code                    3.3.1.648   (MU5)
TPD Kernel Patch                   3.3.1.723   (P170)
== checkupgrade ==
Check Touch Files passed, but returned with a warning:
2025-02-21 15:40:57 CET check_touchfiles: Node 0 Touchfile count = 1
2025-02-21 15:40:57 CET # 1: checkhealth_skip_admin_link_check
2025-02-21 15:40:57 CET check_touchfiles: Node 1 Touchfile count = 1
2025-02-21 15:40:57 CET # 1: checkhealth_skip_admin_link_check
2025-02-21 15:40:57 CET check_touchfiles: Node 2 Touchfile count = 1
2025-02-21 15:40:57 CET # 1: checkhealth_skip_admin_link_check
2025-02-21 15:40:57 CET check_touchfiles: Node 3 Touchfile count = 1
2025-02-21 15:40:57 CET # 1: checkhealth_skip_admin_link_check
2025-02-21 15:40:57 CET
check_touchfiles: WARNING - above touchfiles / files / directories exist on
/var/opt/tpd/touchfiles.
2025-02-21 15:40:57 CET Contact your support organization to verify above are applicable for
upgrade.
The system can proceed with an online upgrade, although please check the above warnings before
proceeding.
  --> The system cannot undergo an online upgrade. See checkupgrade output
Return Code = 1
== upgradesys -status ==
System is not   currently undergoing an online upgrade.
First node to   be rebooted is node 2
   Which will   be followed by node 1
   Which will   be followed by node 0
   Which will   be followed by node 3
2. Follow the Critical Pre-Upgrade Host Recommendations below (reply with Yes/No):
                         Hotfixes might be required for the support of specific Windows functionality.
   For Windows           See the Microsoft website for any required packages, hotfixes, and updated
   Hosts                 drivers.
                         If running Windows 2012 hosts, please confirm the installation of the
                         following MPIO hotfixes. Not installing the hotfixes can lead to the host losing
                         access to LUNs during the OS Upgrade.
 Document type                                                                                        Page
 TECHNICAL DOCUMENT
                         Windows Server 2012 or 2012 R2 - Install KB3046101 to resolve issue:
                         The Server may freeze during startup when ALUA-capable storage is used in
                         Windows 2012 R2 or Windows Server 2012.
                         For 2012 R2: Please note that KB3046101 is superseded by KB3121261. Any
                         Monthly Rollup (2018/03 or later) also includes the fix to resolve this issue.
   For VMware            The Host disconnects due to ATS timeout while running VMware vSphere 5.5
   Hosts                 Update 2 and later. If applicable, follow this Customer Advisory Link
                         VMware 6.5 - Random PSOD with lpfc 11.1 FC/FCoE Driver Series. If
                         applicable, follow this Customer Advisory Link
                         For ESXi 7.X hosts with Emulex HBA models 5330C & SN1200E1P having
                         Driver version below 14.0.622.0, would suggest getting the HBA driver
                         upgraded to latest supported version as per SPOC
  NOTE : HPE recommends updating host drivers to a supported level of the targeted HPE 3PAR OS
  upgrade version before attempting an online update. This information is available on the HPE Single
  Point of Connectivity (HPE SPOCK) website.
 3. Specify the preferred mode of communication. (Email/HPE Myroom/MS Teams):
 Please specify your preferred modes of communication. To maximize
 time and value, we recommend use of Phone.
                                                                              Phone/Email
 Provide the contact details of the person who would be available during
 the upgrade to approve the Controller Node Reboots
 Please specify your preferred collaboration platform/tool. To maximize       MyRoom/Teams/other
 time and value, we recommend use of MyRoom or Teams. Provide an              (please specify)
 alternative only if you don’t have access to these two platforms.
4. Follow the below requirement if target OS is 3.3.1 MU5 or above and LDAP using SSL
   (reply back with Yes/No):
 << The below are not applicable if the target OS is not 3.3.1>>                            Yes/No
   For target OS code 3.3.1 XX.
   3.3.1 XX needs CA certificate defined in order to use LDAP via SSL. Please ensure
   the valid CA bundle for the LDAP server is imported to the array prior to the
   upgrade. If not, once the array is at 3.3.1 XX the array will not connect via SSL to
   the LDAP servers as the CA bundle was not imported into the trust store.
5. Consider the below requirements for target SP 5.0.X Upgrades:
    For Virtual machines.
 Document type                                                                                       Page
 TECHNICAL DOCUMENT
   Virtual Service Processors are supported only on ESXi Version 5.5/6.0/6.5/6.7/7.0 (ESXi 7.0 is only
   supported from 5.0.7) or Hyper-V 2012/2012 R2/2016/2019 & reconfigure the VM to meet the SP 5.x
   requirements (Memory >= 4GB, Virtual CPU cores >= 4)
  Note: To Note: For Hyper-V, do not reconfigure the virtual memory until after the upgrade to SP 4.5 is
    complete.
 Change of Login credentials. After SP upgrade from 4.4.XX to 5.0.X, User credentials have changed
 as below:
                            Username prior to      Username post the          HPE standard
  Service Processor         the upgrade            SP OS 5.x upgrade          password
  TUI /CLI /3PAR
  Service Console           3parcust               admin                      3parInServ
 Note: The passwords for the accounts have been preserved during the upgrade. Use the current
 password with the new username to login into SP 5.x
 Network port change. The network port through which the Service Processor’s webserver listens to
 has changed from 443 to           8443. Please ensure to connect from your web browser to the
 Service Processor (Service Console) on port 8443, which may require firewall rule updates in your
 environment.
 Connect to Service Console. After the upgrade to SP 5.0.X, type in the following URL in your web
 browser to connect to the 3PAR Service Console: https://<SP IP Address>:8443
 Pre requisite for updating SP 5.1.2. New Firewall configuration to allow access to ports and IPs to
 be updated as per the release notes. Click here - Refer Page 194.
 Local notifications’ settings are not ported to 5.x. This must be configured after the SP 4.5 to 5.x
 update has been completed.
6. Verify remote connectivity:
                                                                                         Enabled/
 If the Remote Connection to the Array is enabled, we will need your permission to       Disabled
 perform the upgrade remotely.
 If the Remote Connection to the Array is disabled, we will need to use one of the
 collaboration tools recommended above to perform the Upgrade.
7. Download software packages as a backup for the Code Staging (reply back with Yes/No):
                                                                                         Yes/No
       Pre-Requisite Patch:
   N/A – Already Installed
       Inform OS:
   HP 3PAR Operating System HP_3PAR_ 3.3.2.159 (EMU1)
          Post Patches:
 Document type                                                                                           Page
 TECHNICAL DOCUMENT
   HP 3PAR Operating System HP_3PAR_ 3.3.2.159 (EMU1); P08
   HP 3PAR Operating System HP_3PAR_ 3.3.2.159 (EMU1); P15
      SP OS:
   HPE 3PAR Physical Service Processor 3PAR_PHYSICAL_SP 5.1.4.0-29682
       Upgrade tool kit:
   OS-4.2.0.24-U070
   Note: To ensure you have access to your software updates in My License Portal,
   your support agreements (SAID) must be linked to your HPE passport through HPE
   Support Center. Your software updates are now downloadable from My License
   Portal in a new section named “Software Updates”.
   Download the software packages from My License Portal as a backup in case we
   face any challenges in deploying the codes remotely.
   For additional information about recent My License Portal changes, please access
   the “What’s New” section of the Frequently Asked Questions (FAQs).”
8. Refer to the HPE 3PAR Operating System Update Planning Guide to complete the
   additional steps. You can also download it from here.
9. Refer to the documents below as appropriate:
  Applicable for Option B. When changing upgrade version family, e.g. 3.2.2 XX to 3.3.1 XX, refer
  to the blank
  Host Worksheet
  Host Worksheet and fill in the Host details.
10. Check if the system is registered on HPE InfoSight:
 To register your HPE 3PAR StoreServ system(s), please log into the Cloud-Based Analytics Portal using
 your HPE Passport credentials. Once on HPE InfoSight, review the portal bulletins (My Account > Past
 Bulletins), then select the Register Systems link and complete the registration steps. In case of any
 questions or issues, please use the Portal Support Help Desk link found under Resources > HPE
 InfoSight > General Help. View Technical White Paper here.
 11. HPE recommends updating host drivers to a supported level of the targeted HPE 3PAR
    OS upgrade version     before attempting an online update. This information is available
    on the HPE Single Point of Connectivity (HPE SPOCK) website:
    https://www.hpe.com/Storage/Spock/
12.Upgrade Rollback Plan:
 In case of any issues, the Array can be rolled back to the original version following the below plan:
 Document type                                                                                            Page
 TECHNICAL DOCUMENT
 Rollback when all the Nodes are not upgraded (one or more nodes pending            20 Minutes per Node to
 for the upgrade or the upgrade is not finalized/committed)                         revert to the old version
                                                                                    (online)
 Rollback when all the Nodes are upgraded (upgrade is finalized/committed)          1hr – The Array has to
                                                                                    be offline during this
                                                                                    rollback.
                                                                                    Note: A downgrade
                                                                                    from 3.2.2.x or
                                                                                    3.2.1.x to lower
                                                                                    versions on
                                                                                    production 3PAR
                                                                                    systems is not
                                                                                    recommended by
                                                                                    HPE.
 13. In case of unexpected delay during the Upgrade activity, refer to the handover
    Information below:
  – In scenarios where the OS Upgrade activity is delayed due to reasons such as network connectivity
     issues or delay in staging the codes, please be informed that pending tasks of the upgrade will be
     handed over to one of our available engineers.
  – In case there is a ‘significant delay’ in starting the upgrade or a resource availability issue to take
    over, you would be requested to consider rescheduling the upgrade to another date.
14. CLX & 3PAR CLI Recommendations for 3PAR OS upgrades:
 CLX depends on the 3PAR Remote Copy functionality in the backend. Hence it is always recommended
 to have the same version of the HPE 3PAR OS on the Primary and Secondary 3PAR Arrays.
       Pre-Upgrade recommendations:
    -   CLX and CLI software versions must be upgraded prior or post the 3PAR OS upgrade.
    -   Check the CLX Config matrix for the supportability details under HPE SPOCK Matrix.
    -   https://h20272.www2.hpe.com/SPOCK/Pages/spock2Html.aspx?htmlFile=sw_array_3par.html
    -   When one array is upgraded and the other array upgrade is pending, it is not recommended to
        do any CLX failovers, as there will be a mismatch of the 3PAR CLI versions and hence the CLX
        failover may fail.
    -   Hence, please plan to upgrade both the arrays within a minimal gap between the upgrades.
    -   Prior to the upgrade, it is recommended to failover to the other Cluster Node which will not be
        affected by the 3PAR Upgrade.
       Post Upgrade Recommendations:
    -   Post the 3PAR OS upgrade. The 3PAR CLI must be upgraded to the same version as of the 3PAR
        OS version.
    -   After Installing the new 3PAR CLI, the new Array certificate must be accepted by logging into the
        3PAR CLI. This is required for the CLX to communicate with the 3PAR Array.
        NOTE: Unless the array certificate is accepted in the 3PAR CLI, the CLX connection test will not
        be successful.
Document type                                                                                         Page
TECHNICAL DOCUMENT
   -    Also, the CLX Application must be upgraded to the supported version. Ensure that the CLX
       dependency is removed on all Storage disks configured in the MS Failover Cluster, prior to the
       CLX upgrade.
   -   Care must be taken to pause the Failover of the Roles in the MS failover cluster, when the CLX
       software is being upgraded. Failing to do so, may cause an unexpected Cluster failover to be
       triggered causing Host outages.
   -   The CLX Software must be upgraded on all the Nodes of the Cluster. All nodes should have the
       same version of the CLX software installed, otherwise, the CLX failover will fail.
   -   Please raise a case with the 3PAR Host Applications team (3PAR 3SC Breakfix), in case of any
       queries related to the CLX upgrade procedure.
15.Upgrade Related Information:
Activity        Remote GDC Engineer Name & Contact                    Bharath Kumar P
Information     Information                                (
[email protected])
                Scheduled Time & Date for Upgrade
                                                           2025-03-11 20:00(Europe/Madrid)
                (online/offline)
                Source & Target OS Code + Patches          3.3.1.648 (MU5) to 3.3.2.159 (EMU1)
                Array Serial Number                        2M284900HR
                HPE Case no.                               RR-1418822
                Number of Hops:                            1
                 st
                1 Hop Details:                             3.3.1.648 (MU5) to 3.3.2.159 (EMU1)
                2nd Hop Details:                           N/A
                Pre-requisite Patch(es)                    N/A
                Pre-requisite Script/s                     U070
                Post Upgrade Patch(es)                     P08 & P15
                Post Upgrade Script/s                      U070
Upgrade         Estimated Upgrade Duration (excluding      4hrs
Information &   Cage & PD Firmware Upgrade) In
Timelines for   Hours:
each Hop
                SP Update with the latest code (non-       15min
                intrusive) In Minutes:
Note: In
                Pre-Upgrade Checks (non-intrusive) In      30min
case of 2
                Minutes:
Hops / Multi
Hops            Installation of Pre-requisite Patches In   15min
Upgrade,        Minutes:
these
timelines       Running pre-requisite Scripts In           15min
will vary       Minutes:
accordingly     Controller Node Upgrade (reboot)           Number of nodes *10min +5 min pause
                performed one at a time with Host          time
                validation (5min default pause
                between each Node Reboot. Max
                pause= 30min). In Minutes:
                The Customer to validate Host Access       30min
                and approve Node Reboots
Document type                                                                                                Page
TECHNICAL DOCUMENT
                Installation of Post Upgrade Patches In      15min
                Minutes:
                Running Post Upgrade Scripts In              15min
                Minutes
Post Upgrade    Drive Cage Firmware Upgrade In              10min /Cage
Information     Minutes:
                Health Check and other post checks In       10mins
                Minutes:
                Physical Drives Firmware Upgrade            Depends on the number of drives
                (background task)
                Final Host Verification by the Customer     4 hours
16.Target Specific Recommendations
   If the array model is 8200/8400 and target version is 3.3.1.MU5 or above, please be informed of the
   below:
      If you have features in an unsupported combination before installing P170 which has P155
       included, you will still be able to continue using those features after P170 is installed until the
       feature is deleted/removed.
      If you remove a feature that is in an unsupported combination before or after installing P170,
       you will NOT be able to re-enable it after installing P170
      If AFC is configured and used, consider converting AFC from R0 to R1 before P170 is installed to
       address future drive failures.
The Supported Feature Combinations for HPE StoreServ 8200 and 8400 information is available on
Spock (Page 34)- https://h20272.www2.hpe.com/spock/utility/document.aspx?docurl=Shared
%20Documents/hw/3par/3par_support_matrix_3_3_1.pdf