Contents

Dell DD OS 7.5 Operating System Controller Upgrade Guide PDF

1 of 35
1 of 35

Summary of Content for Dell DD OS 7.5 Operating System Controller Upgrade Guide PDF

Dell EMC DDOS Controller Upgrade Guide 7.4 and up

Rev. 02

This document provides instructions for replacing the controller unit with a newer model. This document covers the following topics:

Revision history.........................................................................................................................................................................................................2

Related documentation........................................................................................................................................................................................... 2

Tools and supplies.................................................................................................................................................................................................... 2

Rules and options for upgrading...........................................................................................................................................................................3

Upgrade planning considerations......................................................................................................................................................................... 8

Source controller preparation.............................................................................................................................................................................. 16

Preparing the destination controller.................................................................................................................................................................. 21

System controller upgrade...................................................................................................................................................................................23

Revision history Table 1. Document revision history

Revision Date Description

02 October 2021 IPMI remote access procedure clarification

01 July 2021 Initial publication

Related documentation Dell EMC provides a variety of document types to support our products. End-user documents include user guides, hardware installation guides, administrator guides, software guides, part replacement guides, release notes, and others. Integration documents describe how to integrate Data Domain and PowerProtect systems with third party backup applications, and compatibility matrices show which components are compatible with each other.

This document refers to other documents by title. To locate a referenced document, go to the Online Support website at https://support.emc.com, enter the document title in the search box, and click the search button.

NOTE: Hard copies of a document may be out of date. Always check for the current version of a document before you start

an upgrade or begin a significant configuration change.

Tools and supplies The following are lists of required and recommended materials for performing a controller upgrade on a Data Domain or PowerProtect system:

Required

These tools and supplies are required to complete the installation and setup tasks for a controller upgrade.

Null modem cable (DB-9 female to female) USB-to-DB-9 serial (male connector) converter cable if the laptop does not have a serial port Power adapter, C13 to NEMA 5-15 (if based in North America), or a power cable for your laptop power adapter with a C13

plug, so that you can power your laptop from a rack PDU Antistatic wrist strap and conductive foam pad Screwdrivers:

Phillips #2 with a 12 inches or longer blade and magnetic tip Phillips #2 with a magnetic tip Phillips #1 Flat head 3/16 inch Flat head 1/4 inch Torx T10

4 GB USB flash memory drive

Recommended

These additional tools and supplies might be helpful with completing a controller upgrade.

A spare null modem cable (DB-9 female to female) A spare USB-to-DB-9 serial (male connector) converter cable if the laptop does not have a serial port Flashlight Needle nose pliers Tie wraps (4 and 8 inch) Diagonal wire cutters (for cutting tie wraps) Roll of 5/8 inch Velcro cable tie material (3M Scotchmate SJ-3401 or similar)

2

Rules and options for upgrading

Upgrade rules and restrictions

Be aware of the following rules and restrictions before performing a controller upgrade procedure:

CAUTION: Controller upgrades on HA pairs require a Dell EMC Professional Services engagement.

All-in-one (AIO) systems are ineligible for a controller upgrade. The target system for the upgrade must have more memory than the source system.

More storage can be added to the source system either before or after adding the controller swap but before bringing up the system.

NOTE: If adding storage before the controller swap, the source system must support the new total capacity and

storage type.

Ensure that the source controller is using ELMS licenses. 1. Since ELMS licensing only runs on DDOS 6.x and later, use the system show version CLI command to determine the

current version of DDOS running on the system. 2. If the system is running DDOS 5.7.x or earlier, upgrade to DDOS 7.0. 3. Run the DD eLicense Conversion Tool available from https://support.emc.com before performing the controller upgrade

procedure. If the existing controller uses ELMS licensing, new licenses through ELMS are required for the destination controller. The

DDOS Release Notes provides the most up-to-date licensing information. If the source system is configured with Cloud Tier, the upgrade target system must support Cloud Tier. The DDOS

Administration Guide provides the list of supported Cloud Tier configurations. Extended Retention (ER) systems cannot be upgraded to DDOS 7.0 and cannot upgrade the controller to DD6900, DD9400,

and DD9900 systems. NOTE: ER systems can use MTree replication to replicate data out of the system, and migrate the aged data to the

Cloud Tier.

System upgrade options

The following tables show the upgrade options that are available:

Table 2. DD860 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD860 36 or 72 DD6800 192 Yes Yes

DD860 36 or 72 DD9300 192 or 384 Yes Yes

DD860 36 or 72 DD9800 256 or 768 Yes Yes

ES30 SATA is supported on the new controller after the controller upgrade from an older model.

Table 3. DD890 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD890 96 DD6800 192 Yes Yes

DD890 96 DD9300 192 or 384 Yes Yes

DD890 96 DD9800 256 or 768 Yes Yes

ES30 SATA is supported on the new controller after the controller upgrade from an older model.

3

Table 4. DD990 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD990 128 DD6800 192 Yes Yes

DD990 128 or 256 DD9300 192 or 384 Yes Yes

DD990 with Cloud Tier a

256 DD9300 with Cloud Tier

384 Yes Yes

DD990 128 or 256 DD9800 256 or 768 Yes Yes

DD990 with Cloud Tier

256 DD9800 with Cloud Tier

768 Yes Yes

a. The Cloud Tier feature requires a Cloud Tier license

ES30 SATA is supported on the new controller after the controller upgrade from an older model.

Table 5. DD4200 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD4200 128 DD6800 192 Yes Yes

DD4200 128 DD9300 192 or 384 Yes Yes

DD4200 128 DD9800 256 or 768 Yes Yes

DD4200 with Cloud Tier

128 DD6800 with Cloud Tier

192 Yes Yes

DD4200 with Cloud Tier

128 DD9300 with Cloud Tier

384 Yes Yes

DD4200 with Cloud Tier

128 DD9800 with Cloud Tier

768 Yes Yes

ES30 SATA is supported on the new controller after the controller upgrade from an older model.

Table 6. DD4500 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD4500 192 DD6800 192 Yes Yes

DD4500 192 DD9300 192 or 384 Yes Yes

DD4500 192 DD9800 256 or 768 Yes Yes

DD4500 with Cloud Tier

128 DD6800 with Cloud Tier

192 Yes Yes

DD4500 with Cloud Tier

192 DD9300 with Cloud Tier

384 Yes Yes

DD4500 with Cloud Tier

192 DD9800 with Cloud Tier

768 Yes Yes

ES30 SATA is supported on the new controller after the controller upgrade from an older model.

Table 7. DD6800 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 and ES40 SAS support on DD6900, DD9400 and DD9900

DS60 support on new controller

DD6800 192 DD6900 288 Yes Yes

4

Table 7. DD6800 system upgrade options (continued)

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 and ES40 SAS support on DD6900, DD9400 and DD9900

DS60 support on new controller

DD6800 with HAa 192 DD6900 with HA 288 Yes Yes

DD6800 with Cloud Tier

192 DD6900 With Cloud Tier

288 Yes Yes

DD6800 with Cloud Tier and HAa

192 DD6900 With Cloud Tier and HA

288 Yes Yes

DD6800 192 DD9400 576 Yes Yes

DD6800 with HAa 192 DD9400 with HA 576 Yes Yes

DD6800 with Cloud Tier

192 DD9400 With Cloud Tier

576 Yes Yes

DD6800 with Cloud Tier and HAa

192 DD9400 With Cloud Tier and HA

576 Yes Yes

DD6800 192 DD9900 1152 Yes Yes

DD6800 with HAa 192 DD9900 with HA 1152 Yes Yes

DD6800 with Cloud Tier

192 DD9900 With Cloud Tier

1152 Yes Yes

DD6800 with Cloud Tier and HAa

192 DD9900 With Cloud Tier and HA

1152 Yes Yes

a. Controller upgrades on HA pairs require a Data Domain Professional Services engagement.

NOTE: SATA is not supported on the DD6900, DD9400 and DD9900 controllers. Data must be migrated to SAS.

Table 8. DD6900 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 and ES40 SAS support on DD9400 and DD9900

DS60 support on new controller

DD6900 288 DD9400 576 Yes Yes

DD6900 with HAa 288 DD9400 with HA 576 Yes Yes

DD6900 With Cloud Tier

288 DD9400 With Cloud Tier

576 Yes Yes

DD6900 with Cloud Tier and HAa

288 DD9400 With Cloud Tier and HA

576 Yes Yes

DD6900 288 DD9900 1152 Yes Yes

DD6900 with HAa 288 DD9900 with HA 1152 Yes Yes

DD6900 With Cloud Tier

288 DD9900 With Cloud Tier

1152 Yes Yes

DD6900 with Cloud Tier and HAa

288 DD9900 With Cloud Tier and HA

1152 Yes Yes

a. Controller upgrades on HA pairs require a Data Domain Professional Services engagement.

5

Table 9. DD7200 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD7200 128 DD6800 192 Yes Yes

DD7200 128 or 256 DD9300 192 or 384 Yes Yes

DD7200 128 or 256 DD9800 256 or 768 Yes Yes

DD7200 with Cloud Tier

256 DD9300 with Cloud Tier

384 Yes Yes

DD7200 with Cloud Tier

256 DD9800 with Cloud Tier

768 Yes Yes

ES30 SATA is supported on the new controller after the controller upgrade from an older model.

Table 10. DD9300 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 and ES40 SAS support on DD9400 and DD9900

DS60 support on new controller

DD9300 192 or 384 DD9400 576 Yes Yes

DD9300 with HAa 192 or 384 DD9400 with HA 576 Yes Yes

DD9300 with Cloud Tier

384 DD9400 With Cloud Tier

576 Yes Yes

DD9300 with Cloud Tier and HAa

384 DD9400 With Cloud Tier and HA

576 Yes Yes

DD9300 192 or 384 DD9900 1152 Yes Yes

DD9300 with HAa 192 or 384 DD9900 with HA 1152 Yes Yes

DD9300 with Cloud Tier

384 DD9900 With Cloud Tier

1152 Yes Yes

DD9300 with Cloud Tier and HAa

384 DD9900 With Cloud Tier and HA

1152 Yes Yes

a. Controller upgrades on HA pairs require a Data Domain Professional Services engagement.

NOTE: SATA is not supported on the DD9400 and DD9900 controllers. Data must be migrated to SAS.

Table 11. DD9400 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 and ES40 SAS support on DD9900

DS60 support on new controller

DD9400 192 or 384 DD9900 1152 Yes Yes

DD9400 with HAa 192 or 384 DD9900 with HA 1152 Yes Yes

DD9400 with Cloud Tier

384 DD9900 With Cloud Tier

1152 Yes Yes

DD9400 with Cloud Tier and HAa

384 DD9900 With Cloud Tier and HA

1152 Yes Yes

a. Controller upgrades on HA pairs require a Data Domain Professional Services engagement.

6

Table 12. DD9500 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 SAS support on new controller

DS60 support on new controller

DD9500 256 or 512 DD9800 256 or 768 Yes Yes

DD9500 with HAa 256 or 512 DD9800 with HA 256 or 768 Yes Yes

DD9500 with Cloud Tier

512 DD9800 with Cloud Tier

768 Yes Yes

DD9500 with Cloud Tier and HAa

512 DD9800 with Cloud Tier and HA

768 Yes Yes

a. Controller upgrades on HA pairs require a Data Domain Professional Services engagement.

NOTE: ES30 SATA is supported on the new controller after the controller upgrade from an older model.

Table 13. DD9800 system upgrade options

Source system Memory (GB) Upgrade target system

Memory (GB) ES30 and ES40 SAS support on DD9900

DS60 support on new controller

DD9800 512 or 768 DD9900 1152 Yes Yes

DD9800 with HAa 512 or 768 DD9900 with HA 1152 Yes Yes

DD9800 with Cloud Tier

768 DD9900 With Cloud Tier

1152 Yes Yes

DD9800 with Cloud Tier and HAa

768 DD9900 With Cloud Tier and HA

1152 Yes Yes

a. Controller upgrades on HA pairs require a Data Domain Professional Services engagement.

NOTE: SATA is not supported on the DD9900 controller. Data must be migrated to SAS.

Table 14. System upgrade compatibility to DD6900, DD9400, and DD9900 systems

System Memory (GB) Max Cap (TB) DD6900 - 288 GB

48<->288 TBu

DD9400 - 576 GB

192<->768 TBu

DD9900 - 1152 GB

576 <-> 1248 TBu/ 1536 TBua

DD4200 128 24-192 Yes Yes Yesb

DD4500 192 48-288 No Yes Yesb

DD6800 192 72-144 Yes Yesb Yesb

DD6800 (expanded)

192 72-288 Yes Yes Yesb

DD6900 288 48-288 No Yes Yes

DD7200 128 72-288 Yes Yes Yesb

DD7200 (expanded)

256 72-432 No Yes Yesb

DD9300 192 108-384 No Yes Yesb

DD9300 (expanded)

384 108-720 No Yes Yes

DD9400 576 192-768 No No Yes

DD9500 256 144-504 No Yes Yesb

DD9500 (expanded)

512 144-1008 No No Yes

7

Table 14. System upgrade compatibility to DD6900, DD9400, and DD9900 systems (continued)

System Memory (GB) Max Cap (TB) DD6900 - 288 GB

48<->288 TBu

DD9400 - 576 GB

192<->768 TBu

DD9900 - 1152 GB

576 <-> 1248 TBu/ 1536 TBua

DD9800 256 384-528 No Yes Yesb

DD9800 (expanded)

768 360-1008 No No Yes

a. With DDOS 7.0 and DDOS 7.1, the DD9900 has a maximum capacity of 1248 TBu. In DDOS 7.2 and later, the DD9900 has a maximum capacity of 1536 TBu.

b. Extra storage must be added as part of this upgrade.

Upgrade planning considerations The following topics describe the guidelines that should be considered before starting a controller upgrade.

Upgrade overview and considerations

Upgrade overview

The controller upgrade procedure requires the following tasks:

CAUTION: The upgrade target system must have at least the same amount of memory as the existing system.

1. Preparing the source controller. a. Ensure that the source and destination controllers are running identical versions of DDOS.

NOTE: If the source controller does not support the target's version of DDOS, the replication must be used.

b. Capture the source controller current system configuration. NOTE: Make note of any applicable, special login names and passwords, such as security username and passwords

and passphrase for encryption.

2. Preparing the destination controller. 3. Shutting down the source controller. 4. Removing the source controller. 5. Installing the destination controller. 6. Reviewing the configuration on the destination controller and verify proper system operation.

NOTE: If a source system does not have Cloud Tier and you want to add it during an upgrade to a DD6900, DD9400, or

DD9900 system, then the nearly equivalent Cloud Tier license and associated hardware should be purchased.

Table 15. Cloud Tier upgrade kit

Target DD Cloud Tier kit, nearly equivalent

DD6900 DD6800

DD9400 DD9300

DD9900 DD9800

No configuration, feature, or hardware changes should be made between the start and finish of this procedure. If any changes are made before the procedure is complete and the saved configurations do not match the new system configuration, the system may become unreliable or fail.

NOTE: Read this entire document before you start a controller upgrade. All the information is important; do not review only

the selected text.

NOTE: The controller upgrade procedure requires one or more reboots of the system.

8

Other considerations

Ensure that the appropriate FC and SFPs are available. Existing fibre, cables, or optical connectors may not be available for reuse. Before performing the installation, verify that all required Fibre Channel and Ethernet components will be on-site. Also verify that all necessary cabling will be completed before installation is scheduled to begin.

Verify that the system has a connection available for management through a copper Ethernet connection to a non-IPMI port such as ethMa. Even if there are no plans to use a copper Ethernet network connection, one sometimes becomes necessary. We recommend a copper connection to the IPMI port as well as to an in-band Ethernet port. Having this in-band copper connection to the system can be especially useful during hardware controller upgrades.

Any failed disks must be replaced and reconstructed before the upgrade procedure starts. Any model running DDOS 5.5.0.x or higher does not support more than one Archive Unit. If the source controller has multiple

archive units, contract Support for assistance in merging them to a single archive unit before the upgrade. Space utilization, system alerts, cleaning, and replication statistics are reset after a controller upgrade. Often, this is the

wanted behavior after upgrading the chassis. If this information must be retained, contact Support before starting the upgrade process.

Download and review the Release Notes of the DDOS you are upgrading to as part of the controller upgrade process. Download Release Notes from the Online Support site at https://support.emc.com.

The same release must be installed on the source and destination controller before you start to uninstall the source controller. The major.minor version must be the same on both controllers.

The source file system usage must be 95% or less (free space 5% or more) to support a controller upgrade. The additional space is required to store temporary files during the upgrade.

Do not change software features while a controller upgrade is in progress. If your upgrade plan includes adding software features such as Replication, make the changes before you start the controller upgrade or after the controller upgrade is complete.

If a storage migration is in progress, do not start a system controller upgrade until after the storage migration finalize process completes. This process can be initiated with the previously mentioned CLI command or DD System Manager.

The migration of PCI and I/O expansion cards from a source system to a destination system in the controller upgrade procedure is not supported. New PCI and I/O expansion cards must be purchased for the new system.

The system controller upgrade process preserves the configuration settings for CIFS, NFS, and DD Boost communication. For WWN migration on systems configured for Fibre Channel, the same DDOS version, which must be version 5.3 or later,

is required on the source and destination Data Domain systems. The upgrade process will preserve the VTL and DDBoost over Fibre Channel configuration from the original system. However, the Fibre Channel ports can differ between the source and destination systems, which requires the migration of the Fibre Channel configuration to the appropriate ports on the destination system. The FC migration procedure is covered later in this document.

After WWN migration and if the source system will be redeployed, the Fibre Channel configuration on the source system must be removed or changed to avoid conflicts with the destination system.

The WWN configuration for DDOS version 5.3 or later is persistent across system reboot, DDOS upgrade, and controller upgrade. However, the WWN configuration is reset to the default values after a DDOS reinstall.

The network configurations of the system controller being replaced may not be applicable to the new upgraded system controller since the network slots/names will change. All network configuration settings of the replaced system controller must be recorded carefully to re-create the network configuration on the new system controller.

If the file system is encrypted, acquire the required passwords or passphrase before beginning the upgrade. If VTL is used on the source system and if the source system is redeployed, there is a possibility of VTL drive serial number

collisions if the same backup software accesses the target system and the redeployed system. Make note of the replication configuration. If DD Boost is supported on the source system and you are planning to upgrade the DDOS, determine if the installed DD

Boost plug-in software is compatible with the target DDOS version by consulting the DD Boost Compatibility Guide. You might need to upgrade the DD Boost plug-in software when you upgrade DDOS.

After the controller upgrade, the system will issue an alert stating that the current elicense is invalid and must be replaced within 30 days. Go to the licensing portal at https://support.emc.com/servicecenter/license/, and select Rehost to move the licenses from the original controller to the new controller. The serial numbers of the original controller and the new controller are required to generate a new license.

NOTE: Before performing a controller upgrade on a High Availability system, verify the system supports offline or online

controller upgrades.

9

Safety considerations

Always use two people or a mechanical lift when moving any Data Domain or PowerProtect system. The systems are heavy. To avoid personnel injury or equipment damage, plan to use two people or a mechanical lift when moving a chassis or shelf.

Power remains for some time after you remove power cords. Do not remove hardware components until all of the LEDs are dark.

Rack considerations

Ensure that the target system rail kit works with the target rack, as the distance between front and rear posts of racks can vary. Some systems are 2U (2 rack units) height, DD6900 and DD9400 are 3U, and other systems are 4U.

The rack mount kits for Data Domain and PowerProtect systems work in the Titan C and later racks. Depending on your source and target systems, the SAS cables may not work with the target. Additional cables may be

necessary. Verify that the rack has four power connections available to support each of the four DD9800 controller power supplies. If the new system controller is installed in a different location than the original system controller, ensure that the existing

shelf SAS cables can reach the new location. If not, order longer cables. Cables 3 meters and longer are intended for rack-to-rack connections. The "U" spanning distance is not applicable in that

case.

NOTE: DD6900, DD9400, and DD9900 systems use 3U as a standard "in rack" cable.

The following are FS15 cabling guidelines: A new FS15 shelf is required when upgrading to a DD9800 system. Vertical span assumes a straight vertical run with the recommended 2.5" (63.5 mm) minimum bend radius. If a cable must skip over a controller (system), the vertical span is reduced by 4 RU.

The following are FS25 cabling guidelines: A new FS25 shelf is required when upgrading to a DD9900 and are required for a DD6900 or DD9400 system with an HA

configuration. Vertical span assumes a straight vertical run with the recommended 2.5" (63.5 mm) minimum bend radius. If a cable must skip over a controller (system), the vertical span is reduced by 4 RU.

Cable options are shown in the following figures and tables.

Figure 1. HD-mini-SAS cable for SAS I/O module to DS60 connections

Table 16. Controller with SAS I/O module to DS60 cable options

Cable Part Number Part Number to Order Cable Length

HD-mini-SAS connector on controller and enclosure

Controllers: DD6800, DD9300, and DD9800

10

Table 16. Controller with SAS I/O module to DS60 cable options (continued)

Cable Part Number Part Number to Order Cable Length

038-004-380-01 XC-DS60-SAS-3M 3 m. (118 in.)

038-000-212-00 XC-DS60-SAS-4M 4 m. (158 in.)

038-000-214-00 XC-DS60-SAS-5M 5 m. (196 in.)

Figure 2. Mini-SAS cable for SAS HBA to ES30 and ES30 to ES30 connections

Table 17. ES30 to ES30 cable options

Cable Part Number Part Number to Order Cable Length

Mini-SAS cable, SFF-8088 connectors on both ends, one end keyed for host ports and the other keyed for expansion ports

038-003-786 X-SAS-MSMS1 1 m (39 in.)

038-003-787 X-SAS-MSMS2 2 m (79 in.)

038-003-751 X-SAS-MSMS3 3 m (118 in.)

038-003-628 X-SAS-MSMS4 4 m (158 in.)

038-003-666 X-SAS-MSMS5 5 m (196 in.)

11

Figure 3. HD-mini-SAS cable for SAS I/O module to ES30 host port connections

Table 18. Controller with SAS I/O module to ES30 host port cable options

Cable Part Number Part Number to Order Cable Length

HD-mini-SAS connector on controller, SFF-8088 connector keyed for host port on ES30

Controllers: , DD6800, DD9300, and DD9800

X-SAS-HDMS1 1 m (39 in.)

038-003-810 X-SAS-HDMS2 2 m (79 in.)

038-003-811 X-SAS-HDMS3 3 m (118 in.)

038-003-813 X-SAS-HDMS5 5 m (196 in.)

12

Figure 4. HD-mini-SAS cable for SAS I/O module to ES30 expansion port connections

Table 19. Controller with SAS I/O module to ES30 expansion port cable options

Cable Part Number Part Number to Order Cable Length

HD-mini-SAS connector on controller, SFF-8088 connector keyed for expansion port on ES30

Controllers: DD6800, DD9300, and DD9800

038-004-108 X-SAS-HDMS2 2 m (79 in.)

038-004-111 X-SAS-HDMS5 5 m (196 in.)

Figure 5. SAS cables for controller upgrades

13

If you have an existing system and are upgrading to a DD6900, DD9400, and DD9900, then some SAS cables must be replaced. The SAS connectors on the back of existing systems always stayed in place. The new systems slide out of the rack for service so extra length is needed. Also, the cable types that are used for the new systems are more flexible to work better with the cable management arms.

If the existing chains of DS60s and ES30s will remain the same, you can use a simple process to determine what cables need to be replaced. Look at the SAS cables where they are attached to the I/O interface. There should be a label relatively near the connector. That label has a part number that begins with "038-". The following table shows the corresponding "038-" part number and the model that can be ordered. Some cables are no available in every length. In those cases, a longer cable may be called out.

NOTE: There is no need to change cables that connect DS60s or ES30s together.

NOTE: 5M cables were used when 3M were too short. Usually, the 5M cable will still work. (There is also a 5M flex.) If it is

too short, you can reconfigure the shelves in the rack so the distances will work. This may affect other cables so everything

needs to be checked.

Existing cable part number New cable part number Model and quantity of cables per model.

038-003-810 (2M HD to mini-SAS) 038-000-150-00 (3M HD to mini flex) DD-SAS-HDMS3 (2 per)

038-003-811 (3M HD to mini-SAS) 038-000-167-00 (6M HD to mini flex) DD-SAS-HDMS6 (2 per)

038-003-813 (5M HD to mini-SAS) 038-000-167-00 (6M HD to mini flex) DD-SAS-HDMS6 (2 per)

038-004-380-01 (3M HD to HD) 038-000-212-00 (4M HD to HD flex) DD-SAS-4M-FLX (2 per)

038-004-212-00 (4M HD to HD) 038-000-214-00 (5M HD to HD flex) DD-SAS-5M-FLX (2 per)

038-004-214-00 (5M HD to HD) 038-000-214-00 (5M HD to HD flex) DD-SAS-5M-FLX (2 per)

038-004-108 (2M HD to mini HA) See note.

038-004-111 (5M HD to mini HA) See note.

NOTE: HA cables are not available in flex or different lengths. Work with your SE to plan cabling for the HA cables.

Time considerations

The system controller upgrade process will take time. Plan accordingly so time is maximized between starting the controller upgrade procedure and the next scheduled backup. Consider the following time-related issues before you start your upgrade.

The upgrade may require rebuilding of the index. Re-indexing may take 4-6 hours to complete, in addition to the time required for preparation and replacing hardware.

The controllers must be using the slot-based Ethernet port naming convention as described later in this document. If an upgrade to this convention is needed, Support assistance is required and must be scheduled before the controller upgrade.

If a DDOS upgrade is required on the source or destination controller, the DDOS upgrade may require over an hour, depending on the amount of data on the system. For instructions on how to upgrade DDOS, refer to the upgrade instructions section of the Release Notes for the target DDOS version.

Storage migration requires a license, must be performed by Professional Services, and requires a significant amount of time. If you plan to migrate storage before or after the upgrade, contact your representative to make arrangements. Storage migration is an independent operation, separate from the controller upgrade.

The file system must be enabled. The amount of time required to enable the file system depends on the system changes that are required.

Restoring the Fibre Channel configuration, re-zoning the SAN, and re-mapping WWNNs and WWPNs on the new controller may take additional time, depending on the size, complexity, and manual effort required to redefine the mappings. Tape device configuration on the systems performing backups may also be required, as some systems use the WWNN and WWPN to identify devices.

Although executing the system headswap command may take less than 30 minutes, the entire controller upgrade procedure may take up to eight hours or more. The upgrade process takes longer when the NVRAM batteries on the new system require charging or when the file system index must be rebuilt.

Charging the NVRAM batteries on a new system can add three hours to the upgrade procedure. If possible, apply power to the destination controller on a day prior to the day you plan to perform the upgrade.

If you are unsure of the time required for this upgrade, contact your account team.

14

NOTE: After executing the headswap command, the system reboots automatically.

Hardware considerations

CAUTION: The upgrade target system must have at least the same amount of memory as the existing system.

Ensure that the power requirements for the new system have been met. Some Data Domain and PowerProtect systems operate on either 100-120 V or 200-240 V whereas other systems require a 200-240 V power source.

The DD9800 controller requires four power connections, one for each power supply. Systems require between 2U and 4U. If the new controller is installed in a different location, ensure that the existing shelf

SAS cables reach the new location. If not, order longer cables as needed. DD6900 and DD9400 systems require 3U, including space for cable management. DD9900 systems require 4U, including space for cable management.

Purchase new optional I/O cards for the target system (either PCI cards or I/O modules as appropriate for new systems). Systems do not support transferring any cards or I/O modules.

Upgrading a controller to a DD9800 or DD6800 with HA require the addition of an FS15 SSD enclosure. Add the enclosures before you start the controller upgrade or after the controller upgrade is complete. Do not change the enclosure configuration while a controller upgrade is in progress.

DD6900 and DD9400 with HA require an FS25 SSD enclosure. DD9900 systems always have an FS25. Before upgrading a controller, verify that the destination controller supports the enclosure configuration on the source

controller. Upgrade the enclosures or migrate data to newer enclosures before upgrading the controller. Consider the following information when evaluating the enclosure configuration: SATA and SAS shelves are not supported on the same chain of enclosures.

When upgrading to DD9800, adding an SSD shelf is required as part of the upgrade process.

DD6900, DD9400, and DD9900 considerations

Not supported:

Shelves with SATA drives Shelves with 2 TB SAS drives

Migrating from unsupported drives:

If the entire system has unsupported drives, use Collection Replication to move data to a new system after upgrading the source system to DDOS 7.0.

If most of the drives are unsupported, use Collection Replication. If only a few are unsupported, use shelf migration on the source system before the controller upgrade to migrate data to the

supported shelves and drives (3 TB SAS, 4 TB SAS). Enough 4 TB and 8 TB shelves are needed to enable shelf migration of existing shelves and to enable required system

growth. For the DD9400, there must be a minimum of 4 4TB and 3TB shelves before adding any 8TB shelves For the DD9900, there must be a minimum of 8 4TB and 3TB shelves before adding any 8TB shelves

Note the shelf count rules (maximum seven SAS ES30s per chain, maximum three DS60s per chain). It may be necessary to add a few target shelves at a time due to chain limitations, if rack space is available close enough

to the source system, or SAS cable length restrictions.

Because of the associated procedures and downtime, plan this process carefully.

Only systems upgraded to DD6900, DD9400, and DD9900 support 3 TB SAS. They cannot be added to a new system.

Notes on 8 TB drive usage on DD9400 and DD9900

To ensure the same or better performance as previous models across usable capacity points, the 8 TB drives are over- provisioned with 30 additional drives and the licensed capacity is distributed across the licensed drives in addition to the 30 extra drives.

Controller upgrades prior to 7.4 will result in this over-provisioning method getting disabled. Once 8 TB drives are added into the file system with this setting, the only way to transition to another system that uses over-provisioning is with Collection Replication to that new system.

Controller upgrades to a system running 7.4 or later will disable the over-provisioning method if the configuration being imported has one or more 8 TB drives and was created with no over-provisioning. In all other cases, it will leave the 8

15

TB over-provisioning method on the new system enabled. This allows a configuration that has smaller drives to use shelf migration to reach an 8 TB only configuration, using over-provisioning like a newly configured system.

Source controller preparation The following topics describe how to prepare the source controller for the upgrade.

NOTE: Because these tasks can be time consuming, it is a good practice to complete these tasks prior to the day of the

upgrade.

Examining and updating the port naming method

If the source controller is not already using the slot-based Ethernet port naming system, the controller must be upgraded to use slot-based port naming before the controller upgrade.

To determine if slot-based port naming is in use, connect a console to the system, log in, and enter the command net show hardware. Ethernet port names of ethMa, eth0a, or eth0b, for example, indicate slot-based port naming. Ethernet port names of eth0 or eth1 indicate legacy port naming, which was the default naming used on systems before DDOS 4.9.

To upgrade the system to the slot-based naming convention, contact Support at https://support.emc.com.

Examining and updating DDOS on the source controller

About this task

The controller upgrade requires that the source and destination controllers must be running the same DDOS version as defined earlier in this document.

Steps

1. Reboot the system and watch for any messages indicating hardware issues during startup. Resolve all issues before you continue.

# system reboot 2. Connect a console to the system and log in.

3. Run the alerts show current command to verify there are no active alerts on the system.

4. Enter system show version.

5. Make a note of the version.

If the version matches the version of the destination controller as described earlier in this document, you do not need to update DDOS on the source controller. You can skip steps 5 and 6 of this procedure.

6. If the source controller requires an upgrade, download the appropriate version of the DDOS Release Notes from the Online Support site at https://support.emc.com.

7. Follow the instructions in the release notes, and upgrade DDOS with the system upgrade command.

NOTE: If your upgrade plan calls for mulitple sucessive upgrades (for example, 4.9->5.1->5.2.3), use the filesys show space command to verify that the file system is available before you start the next upgrade. If the filesys show space command does not display the Active Tier Resource information, or if it displays The filesystem is unavailable at this time, the system is not ready for the next upgrade.

Recording the source controller configuration

About this task

Due to differences in network and fiber channel port numbers in the source and destination controllers, you may need to make some configuration changes after the destination controller is installed. While most configuration information is retained during the upgrade, VTL configuration information might be lost. Dell EMC Support recommends that you record

16

the complete configuration of the existing system before beginning the upgrade procedure so that you can review and repair any configuration issues on the destination controller.

Steps

1. If the existing system is configured with a passphrase, verify that you have access to the passphrase when you install the new controller.

CAUTION: You cannot use the system to display the passphrase. After the headswap, you or the passphrase owner

must reenter this passphrase.

a. Run the filesys encryption status command.

b. If encryption is enabled, run the system passphrase change command and enter the existing passphrase.

c. Enter Ctrl + C to cancel changing the passphrase.

2. Run the command #alerts show current. If there are alerts or issues, resolve them before continuing.

3. Send an autosupport email message to yourself.

# autosupport send username@yourcompany.com If access the source controller is needed, configuration, the autosupport message records key system configuration features and serves as another information source .

4. Configure your terminal emulation program to record the session.

Most of these steps capture the source controller configuration using CLI commands. Recording the session captures the configuration for future reference.

5. Check for multiple archive units. If there are multiple archive units, contact Support for assistance in merging them into one archive unit.

# filesys archive unit list all 6. Check if space reclamation is running, and if so, suspend it.

# archive space-reclamation status # archive space-reclamation suspend

7. For a system enabled with Cloud Tier, ensure that there is no data movement.

NOTE: If the source system is configured with Cloud Tier, the upgrade target system must support Cloud Tier. The

DDOS Administration Guide provides the list of supported Cloud Tier configurations.

# data-movement status # data-movement stop all

8. Check if cloud cleaning is running, and if so, stop it.

# cloud clean status # cloud clean stop

9. Display the license keys. For DDOS versions less than 6.0:

# license show For DDOS 6.0 and greater:

# elicense show all 10. If used, record the VTL configuration.

# vtl show config # vtl option show all

11. If either VTL or DDBoost-over-FC is present, record the scsitarget configuration.

# scsitarget port show detailed # scsitarget group show detailed # scsitarget endpoint show detailed # scsitarget initiator show detailed # scsitarget transport option show all # scsitarget endpoint show list

17

NOTE: WWN migration requires the WWN information that is displayed by the scsitarget endpoint show detailed command. To view this information in the autosupport file, search for SCSITARGET Endpoint Show Detailed

or Enclosure Show All. To view this information in DD System Manager, select Hardware > Fibre Channel > Resources

and expand the Endpoints area.

12. Create a table to map the endpoints to the correct ports on the destination controller as described in the following procedure:

The endpoint name, WWPN, and WWNN are transferred to the destination controller during the upgrade. When the destination controller starts; however, new endpoints are created for each port. After the destination controller starts, use the table that you create here to delete the new endpoints and reassign the migrated endpoints to the correct ports.

a. In the first column of each row, record an endpoint name. b. In the second and third columns for each row, enter the WWPN and WWNN for the endpoint. c. In the fourth column, enter the source controller port number that hosts the endpoint. d. In the fifth column, enter the destination controller port number that you want to host the endpoint.

If you do not have the destination port number, you can get it later after the destination system starts. The endpoints are displayed in the output of the scsitarget endpoint show list command.

The following is an example-mapping table. Notice that this table calls for source ports 4a and 4b to migrate to ports 4a and 4b on the destination system, and source ports 5a and 5b map to slot 6 on the destination system. Because new endpoints with WWPN and WWNN addresses are automatically created for the destination ports, all new endpoints for the ports that are listed in the table must be deleted and each migrated endpoint that is assigned to the correct port.

Table 20. Example: Endpoint to port-mapping table

Endpoint WWPN WWNN Source Port Destination Port

endpoint-fc-0 24:00:00:21:88:2d:e8:a0 20:00:00:21:88:2d:e8:a0 4a 4a

endpoint-fc-1 24:10:00:21:88:2d:e8:a0 20:00:00:21:88:2d:e8:a0 4b 4b

endpoint-fc-2 25:00:00:21:88:2d:e8:a0 20:00:00:21:88:2d:e8:a0 5a 6a

endpoint-fc-3 25:10:00:21:88:2d:e8:a0 20:00:00:21:88:2d:e8:a0 5b 6b

NOTE: To prevent service interruption, there must be a corresponding port on the destination system for every port on

the source system.

13. Record the replication configuration.

# replication show config 14. If used, record the NFS and DD Boost configurations.

Although it should not be necessary to manually reconfigure NFS or DD Boost, Dell Technologies recommends that you record those configuration settings. When you run these commands before and after the controller upgrade, the output should be the same.

# nfs show clients # net hosts show # net show hardware # ddboost user show # ddboost storage-unit show # ddboost fc status # ddboost fc dfc-server-name show # ddboost fc group show list

15. If used, record the CIFS configuration and share definitions.

# cifs share show # cifs show config

16. Record the file system configuration.

Run the mtree retention-lock status and snapshot list mtree commands for each MTree on the system.

# disk show state # disk status # disk show hardware

18

# system show ports # disk multipath status # filesys show space # enclosure show topology # mtree list # mtree retention-lock status mtree mtree-path # filesys encryption show # mtree list # snapshot list mtree mtree-path

NOTE: Use the MTree path that is identified from the mtree list command when running the mtree retention- lock status mtree-path command.

17. For systems with Cloud Tier, record the file system configuration.

# storage show all # cloud unit list # cloud profile show all # cloud clean frequency show # cloud clean throttle show # data-movement policy show # data-movement schedule show # data-movement throttle show

18. Record external network card configuration including card slot, card vendor and model, speed, and number of ports.

# system show hardware # system show port # net show hardware

19. Record network services on the controller including failover, aggregation, VLANs, and IP aliases.

# net show config # net show hostname # net show dns # route show gateway # net show settings

port enabled state DHCP IP address netmask type additional setting /prefix length ------ ------- ------- ---- -------------------------- -------------- ------------ ----------------------------------------------- ethMa yes running no 192.168.10.70 255.255.255.0 n/a fe80::260:16ff:fe5c:90dc** /64 ethMb yes running no 10.5.132.70 255.255.255.0 n/a fe80::260:16ff:fe5c:90dd** /64 ethMc yes running no 192.168.10.56 255.255.255.0 floating fe80::260:16ff:fe5c:90de** /64 ethMd yes running no 10.5.132.209 255.255.255.0 floating fe80::260:16ff:fe5c:90df** /64 eth1a no down no n/a n/a floating eth1b no down no n/a n/a floating eth1c no down no n/a n/a floating eth1d yes running no 192.168.60.56 255.255.255.0 floating fe80::260:16ff:fe51:903f** /64 eth11a yes running n/a n/a n/a interconnect bonded to veth99 eth11b yes running n/a n/a n/a interconnect bonded to veth99

19

eth11c yes running n/a n/a n/a interconnect bonded to veth99 eth11d yes running n/a n/a n/a interconnect bonded to veth99 veth99 yes running no d:d:d:d:d:0060:1652:2888 /80 interconnect lacp hash xor-L3L4: eth11a,eth11b,eth11c,eth11d fe80::260:16ff:fe52:2888** /64 ------ ------- ------- ---- -------------------------- -------------- ------------ ----------------------------------------------- ** auto_generated IPv6 address

20. Save the output of your terminal emulation program with an appropriate file name.

21. Develop a port-mapping table that includes identification of the ports on the source controller, the correlation of the ports on the destination controller, and the IP addresses.

Port Enabled State DHCP IP Address Netmask/ prefix length

Type Additional Setting

eth0a yes running no 192.168.10.71 255.255.255. 0

fe80::260:16f f:fe5c:90dc* *

/64

In the rear of the system, I/O slots are numbered in varying ways. Consult your system's documentation for details. When facing the rear of the system, ports are numbered left to right, and/or top to bottom. Platforms with DDOS 5.4 and later have slot numbering that begins with "0" as opposed to older legacy platforms in which numbering began with "1". For example, a legacy system that is mapped to slot1 would be mapped to slot0 on newer platforms. Also, a legacy system onboard port was named with "0" (for example, eth0a/0b), while newer platforms use "M" (for example, ethMa/Mb).

With the exceptions of the DD2200/2500 models, only management and BMC traffic should pass over onboard NICs.

Consider the following guidelines when working with DD4200, DD4500, and DD7200 systems:

There is only one on-board Ethernet interface (ethMa). Migrating the onboard settings from a legacy system would be applicable to only one interface (eth0a in legacy systems).

Network features such as bonding and aggregation for ethMa are not supported. If the legacy system has any of these features for eth0a, these features are not applicable.

For information relating to I/O module population requirements, go to the Online Support site at https://support.emc.com for system-specific installation and hardware guides.

Assume that the source controller has eth3a and eth3b bonded to veth0 (which means the NIC card is in slot 3). To collect the network port settings and IP address information, use the net show settings command as shown in the following example.

Figure 6. Output: net show settings

If the NIC card in the destination controller is in slot1, the same settings should be applied for eth1a and eth1b in the new system. The mapping is slot3->slot1, and the new settings should be as shown in the following example.

20

Figure 7. Output: net show settings

22. If Compliance Mode is enabled, verify that the IPMI user is present.

23. Record IPMI remote access configuration on the controller.

# ipmi show config # ipmi user list

24. Write down the configuration and user names.

The user password is not carried over to the new controller.

25. Develop a service-mapping table to include information for services on the existing controller and the correlation of these services on the new controller.

Include services such as link failover, link aggregation, VLAN, and IP aliases. It may be preferable that the ports for failover or aggregation group are on different cards but connected to the same subnet.

26. For those systems which include the cache disks in head unit (DD6800, DD6900, DD9300, and DD9400), run the following commands before performing the controller upgrade:

a. Run storage show tier cache.

# storage show tier cache

Cache tier details:

Disk Disks Count Disk Additional Group Size Information ----- ------- ----- ---------- ----------- dg1 1.5-1.8 4 745.2 GiB ----- ------- ----- ---------- -----------

b. Remove all cache disks.

# storage remove disks 1.5 1.6 1.7 1.8 27. For those systems that cannot upgrade to DDOS 7.0 (DD4200, DD4500, DD7200, and DD9500), the current DDOS version

should be 6.1.x or 6.2.x. They can then be upgraded to a DD6900, DD9400, or DD9900 system.

Systems running 6.1.x must upgrade to DDOS 6.2.0.35 prior to upgrading to DDOS 7.0.

Preparing the destination controller The following steps describe how to prepare the destination controller for the upgrade.

Prerequisites

NOTE: Because some of the steps in this procedure are time consuming, it is a good practice to perform this procedure on

a day prior to the upgrade day.

Steps

1. If the destination controller is a new system, write down the serial number on the back of the new system.

The serial number is the default password for the sysadmin account on a new system.

2. If possible, set up the new system on a workbench and connect the power cables. Otherwise, see the instructions in the rack mounting kit and install the system in either a temporary or permanent location in a rack.

Leave the power on until the NVRAM batteries are fully charged. It can take up to 3 hours to fully charge the batteries.

21

NOTE: Do not uninstall the source controller to make room for the destination controller until instructed to uninstall the

source controller. For this procedure, the connection of SAS cables to shelves is not necessary and not recommended.

3. Connect a console to the system.

4. Power up the system.

The boot up sequence for a new system may take an extended amount of time.

5. When the login prompt is displayed, log in using the sysadmin username and the configured password.

NOTE: The default password for new system is the serial number on the back of the system.

6. When the configuration utility starts up, enter Ctrl-C to exit the utility. Do not change the sysadmin account password now.

The initial message in the configuration utility is the welcome message.

Welcome to Data Domain OS To improve security, Data Domain recommends that you change the 'sysadmin' password before continuing with the system configuration. Change the 'sysadmin' password at this time? (yes|no) [yes]: no

7. View the licenses. For DDOS versions lower than 6.0:

# license show For DDOS 6.0 and greater:

# elicense show all 8. Record the license key codes.

You must add these license keys after the upgrade.

9. Run the storage show tier cache command to check for the presence of Cache Tier storage on the destination controller.

10. If the Cache Tier exists, uninstall it before continuing with the controller upgrade.

CAUTION: Do not re-create the Cache Tier on the destination controller until after the system headswap command runs and completes successfully, as directed later in this procedure.

# storage remove enclosure 2

Removing enclosure 2...Enclosure 2 successfully removed.

Updating system information...done

Successfuly removed: 2 done 11. Enter the system show version command.

12. Make a note of the version.

13. Collect an autosupport report to document the starting configuration of the destination system.

14. If the destination controller has a newer major version of DDOS than the source controller, do the following:

a. Enter system show nvram.

22

b. When the NVRAM batteries are fully charged (80% minimum), use the system poweroff command to shut down the destination controller.

NOTE: It can take up to three hours to charge the NVRAM batteries. Never shut-down the system by pressing

the power button (if present on your system). The NVRAM batteries in an improperly shut down system can drain

quickly.

c. Stop this procedure because the DDOS versions do not match.

You must either upgrade the source controller to the same DDOS version that is used on the destination controller, or you must replace the DDOS on the destination controller with the version that is used on the source controller.

15. If the destination controller DDOS version is older than that on the source controller, do the following:

a. If the destination controller you are installing already has DDOS software that is installed, follow the upgrade instructions in the release notes, and upgrade DDOS with the system upgrade command.

NOTE: If your upgrade plan calls for multiple successive upgrades (for example, 4.9->5.1->5.2.3), use the filesys show space command to verify that the file system is available before you start the next upgrade. If the filesys show space command does not display the Active Tier Resource information, or if it displays The filesystem is unavailable at this time, the system is not ready for the next upgrade.

b. When the NVRAM batteries are fully charged (80% minimum), use the system poweroff command to shut down the destination controller.

NOTE: Never shutdown the system by pressing the power button (if present on your system). The NVRAM batteries

in an improperly shut down system can drain quickly. Recharging those batteries can add two to three hours to this

upgrade procedure.

16. If the destination controller does not have DDOS installed, install it from a USB device.

System controller upgrade The following topics describe the tasks required to perform the controller upgrade.

Taking the source controller out of service

Steps

1. Before taking the source controller out of service, shut down all backup applications that are writing data to the system using the guidelines for those applications.

2. Check for unresolved alerts. If there are alerts or issues, resolve them before you continue this procedure.

# alerts show current 3. Check for multiple archive units. If there are multiple archive units, contact Support for assistance in merging them into one

archive unit.

# filesys archive unit list all

NOTE: Source systems with archive units or Extended Retention cannot be upgraded to DD6900, DD9400, and

DD9900. See the Upgrade rules and restrictionssection for more information.

4. For a system with the Extended Retention software option, ensure that there is no data-movement or space-reclamation running.

# archive data-movement status # archive data-movement stop # archive space-reclamation status # archive space-reclamation suspend

5. For a system with Cloud Tier, ensure that there is no data-movement or cloud cleaning running.

# data-movement status # data-movement stop

23

# cloud clean status # cloud clean stop

6. Disable VTL:

a. Determine whether VTL is enabled. To determine if VTL is enabled, enter:

# vtl status b. If VTL is enabled, disable VTL:

# vtl disable 7. If SCSI Target is enabled (scsitarget status), disable SCSI Target:

# scsitarget disable 8. If there are DD Boost connections, stop the backup applications on the connected systems. To identify those systems, enter:

# ddboost show connections 9. Determine whether replication is enabled. If it is enabled, disable replication.

a. To check replication status, enter:

# replication status b. If replication is enabled, disable it. Enter:

# replication disable all c. Verify that replication is disabled on all contexts. Enter:

# replication status

NOTE: Do not proceed until all replication contexts are disabled. This may take up to 15 minutes.

10. Verify that there is no traffic on the Ethernet NICs. Enter:

# iostat 2 Run for approximately 60 seconds before exiting.

11. Verify that there is no traffic on the FC ports. Enter:

# scsitarget initiator show stats interval 2 count 60 12. Determine whether there are NFS connections and unmount the file systems on the remote systems. To identify the remote

systems that have mounted the storage, enter:

# nfs show active a. On those remote systems, use the native operating system commands to dismount NFS exports from the system, usually

using the umount command.

b. After all NFS clients are dismounted, disable NFS:

# nfs disable 13. Check for any active CIFS client connections and stop them. To check for CIFS activity, enter:

# cifs show active a. If there are active CIFS clients, disconnect them and disable CIFS by running:

# cifs disable 14. Disable the file system:

NOTE: For HA pairs, run this command on the primary node.

# filesys disable

NOTE: Disabling the file system also disables CIFS and NFS.

15. If a system has a cache tier (such as DD9500 or DD9800), remove it.

24

a. Run storage show tier cache.

# storage show tier cache

Disk Disks Count Disk Additional Group Size Information ----- ------- ----- ---------- ----------- dg2 2.1-2.5 5 745.2 GiB ----- ------- ----- ---------- -----------

b. Remove all cache disks.

# storage remove enclosure 2

Changing controllers

About this task

CAUTION: Once the new system controller has been installed and booted, the upgrade process cannot be stopped. If

this happens, the integrity of the data can be affected.

Steps

1. Before powering off the system, entering the command filesys disable is recommended.

2. Power off the existing system by using the system poweroff command. Do not power off the shelves.

3. Label all cables on the source controller, and then disconnect them.

CAUTION:

Power remains for some time after you remove the power cables. Wait until all of the LEDs are dark. Always use two people or a mechanical lift when moving any Data Domain or PowerProtect system.

4. Remove the existing system controller.

NOTE: The IP and Fibre Channel configurations remain on the source system and will cause conflicts if that system

is later reconnected to the same LAN or SAN to which the destination system is connected. Before reconnecting the

removed system to the original LAN or SAN, remove the configuration from all ports that can be connected to a LAN or

SAN.

5. Install the replacement system controller in the rack.

NOTE: HBA and I/O Modules may not align from one controller model to another. Validate the HBA/Module placement

with the Installation Procedure documents. Certain HBA/Modules cannot be moved to align with the controller being

replaced. This necessitates manually reconfiguring settings for affected interfaces.

6. Excluding the cables for the Fibre Channel ports, connect all system interface cables and any KVM switch for the system console.

Refer to the map tables that you created when preparing for the upgrade. If you did not create configuration tables, stop now and create them using information in the Autosupport report you generated on the source controller.

7. Connect the SAS cables to all storage shelves, ensuring that the shelves are cabled appropriately. For more information, see the installation and setup guide for the new system and the appropriate expansion shelf hardware guides.

8. Power on the new controller.

NOTE: The power for attached shelves should have remained on.

9. When the login prompt is displayed, log in using the serial number as the password for the sysadmin account.

10. Verify the status of the NVRAM batteries by entering the system show nvram command.

If the batteries are charged 60% or less, enable them to charge as described in the chassis replacement or upgrade guide for the new system. Recharging the NVRAM batteries can take up to three hours.

If the NVRAM batteries have been partially discharged, an error light (amber in color) at the back of the NVRAM card begins blinking when the system is powered on, and an error message similar to the following is displayed on the console.

25

NVRAM0: Size 5224288 KB, Battery 1 Enabled (FAILURE), Battery 2 Enabled (FAILURE)

Let the batteries charge before continuing.

Configuring the destination controller

About this task

Many steps in the following procedure require that you compare command output from the destination controller with command output from the source controller to verify that all hardware and features are available on the destination controller. This information should have been recorded as described earlier in this guide. You can also find this information in the ASUP that you were instructed to create.

For HA pairs, run all filesys commands from the first (active) node.

Steps

1. Check for alerts.

# alerts show current 2. Check system topology.

# enclosure show topology 3. Test the topology for each port that connects to an enclosure.

# enclosure test topology duration 1 Review the port assignments and any listed errors, and resolve any issues before you proceed.

4. Verify that the destination controller lists all the enclosures and disks that were installed on the source controller.

The output should indicate that the data storage is not configured and is foreign storage.

NOTE: User data becomes available in the final steps of the procedure.

# storage show all Active tier details: Disk Disks Count Disk Additional Group Size Information ------- ----- ----- --------- ----------- (spare) 2.1 1 931.5 GiB (spare) 3.15 1 931.5 GiB ------- ----- ----- --------- ----------- Current active tier size: 21.8 TiB Active tier maximum capacity: 261.9 TiB Storage addable disks: Disk Disks Count Disk Enclosure Shelf Capacity Additional Type Size Model License Needed Information --------- -------- ----- ---------- --------- -------------- --------- (foreign) 2.2-2.15 14 931.5 GiB ES30 N/A Host SN: 0F41722499 (foreign) 3.1-3.14 14 931.5 GiB ES30 N/A Host SN: 0F41722499 --------- -------- ----- ---------- --------- -------------- ---------

# disk show state Enclosure Disk 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 --------- ---------------------------------------------- 1 . . . . s . . - - - - - . - 2 O O O O O O O O O O O O O O s 3 O O O O O O O O O O O O O O s 4 O O O O O O O O O O O O O O s --------- ----------------------------------------------

26

Legend State Count ------ ------------------- ----- . In Use Disks 7 s Spare Disks 4 O Foreign Disks 42 - Not Installed Disks 6 ------ ------------------- -----

5. NOTE: Before performing a headswap on DD4200, DD4500, DD7200, and DD9500 systems that are upgrading to a

DD6900, DD9400, or DD9900 system, run the following command in bash mode on the new system.

# reg_cmd set group.upgrade.iknowheadswaprisk 1

The system responds:

This command returns the system back to its prior operational conditions. The system will be rebooted before resuming normal operations.

** If system passphrase was set on the old head, you will need to do one of the following after headswap completes: - unlock the filesystem if you have encrypted data, or - set the system passphrase if you don't have encrypted data Are you sure? (yes|no|?) [no]: yes ok, proceeding. Please enter sysadmin password to confirm 'system headswap': Restoring the system configuration, do not power off / interrupt process ...

This command may take several minutes to complete; please wait.

Success deployment chassis replacement..........[start] Reassemble external shelf...............[OK] Save external storage UUID to Head .....[OK] Save head serial to external storage ...[OK] Mount management volume.................[OK] deployment chassis replacement..........[done]

This command might take 30 minutes or longer to complete, depending on the number of shelves and the amount of

data on the shelves. The system should reboot automatically to complete the headswap process.

After verifying that the disks are visible to the destination system as foreign devices, enter the following command:

# system headswap 6. Log in with the original source system password (not the serial number).

7. Verify the health of the file system by entering the following command.

# filesys status If the file system is encrypted and locked, the system prints the following message.

The filesystem is locked: Use 'filesys encryption unlock' to enter the passphrase. Do the following to unlock the encrypted file system:

a. Contact the customer security officer and, when that person arrives, enter the following command:

# filesys encryption unlock b. Have the security officer enter the passphrase.

WARNING: Although the command output recommends that you enable the file system, do not enable the file

system until asked to do so. The system must be rebooted to ensure that the headswap process completes, and

the headswap process must complete before you enable the file system.

NOTE: The system should reboot automatically. A reboot is necessary to ensure initiation and completion of an index

rebuild. Wait until the indexing process is complete before proceeding.

27

NOTE: The index rebuild process can take up to 4 to 8 hours. Do not power off or reboot the system during the index

rebuild.

8. To check on the status of the upgrade and the index rebuild, enter the system upgrade status command, which shows the estimated time remaining for the upgrade.

# system upgrade status Thu May 2 15:22:50 PDT 2013 Current upgrade status: upgrading file system Time remaining: 240 minutes Start: 14:20:54 Stage: index_rebuild Status: running Reason: none

# system upgrade status Current Upgrade Status: DDOS upgrade succeeded End time: 2016.01.07:17:19

9. If the destination controller includes cache tier SSDs in the head unit, or an FS15 or FS25 SSD shelf, an electronic license for cache tier storage is required. The Release Notes provide the most up-to-date information about electronic licensing.

10. Run one of the following commands to add the SSDs to the cache tier:

For systems where the SSD is located in the head unit:

# storage add disks 1.5,1.6 tier cache Checking storage requirements...done Adding disk 1.5 to the cache tier...done

Updating system information...done

Disk 1.5 successfully added to the cache tier.

Checking storage requirements... done Adding disk 1.6 to the cache tier...done

Updating system information...done

Disk 1.6 successfully added to the cache tier.

For systems where the SSDs are located in an FS15 SSD or FS25 shelf:

# storage add enclosure 2 tier cache Checking storage requirements...done Adding enclosure to the cache tier...done

Updating system information...done

Enclosure 2 successfully added to the cache tier.

11. For DD6900, DD9400, and DD9900 systems, run the following command to configure and enable Retention Lock Compliance and to add an iDRAC user:

a. Configure Retention Lock Compliance:

# system retention-lock compliance configure b. Create an iDRAC user:

# user idrac create c. Enable Retention Lock Compliance:

# system retention-lock compliance enable d. Reboot the system.

12. When the upgrade is complete, enter the filesys status command.

# filesys status The filesystem is disabled and shutdown. [filesys disable]

28

13. If the filesys status output shows that the file system is disabled or locked, enter the following command.

# filesys enable Please wait..................... The filesystem is now enabled.

14. Use the config setup or net config command to reconfigure the network settings according to the map tables recorded earlier.

For information about using the config setup command to launch the utility that helps you configure a system, see the Data Domain Operating System Initial Configuration Guide.

For information about using the net config command to add and remove interface configurations, see the DD Command Reference Guide.

NOTE: All network interface connections that are no longer valid (such as configurations for unused slots or slots with a

different type of card) must be removed.

15. Validate that basic services such as NFS and CIFS were enabled when the file system comes online. If they are not online, enable them.

# NFS status # NFS enable # CIFS status # CIFS enable

16. Use the following command to confirm that the correct amount of storage is connected.

# filesys show space The output of this command should show approximately the original amount of storage space. The output from a filesys show space command may vary from original source system as there are minor differences in the space needed forDDOS from release to release.

17. If it is necessary to change the hostname, do so now. Also make the necessary configuration changes in the backup applications that use the old hostname.

If the existing hostname contains the model number of the old controller, you might want to change the hostname. The hostname may be defined in DHCP or with net set hostname.

NOTE: Some backup applications do not tolerate hostname changes. Verify that all backup applications tolerate the

update before changing the hostname.

18. Reset the system passphrase using the following steps:

a. Run the command system passphrase set to reset the passphrase to the same passphrase as before the headswap.

b. To use the imported host certificate, restart the web server using the following commands.

# adminaccess disable http/https # adminaccess enable http/https

NOTE: CSR generated in a DD or PowerProtect system, the "basic-constraint CA:FALSE", needs to be manually

requested for when creating the CSR from the CLI.

19. Verify that all licenses installed on the source controller that is migrated to the destination controller.

NOTE: If the existing controller uses legacy licensing and the new DDOS version is 6.0 or higher, the system alerts

that the current elicense is invalid and must be replaced within thirty days. Go to the licensing portal at https://

support.emc.com/servicecenter/license/. The serial numbers of the original controller and the new controller are

required to generate a new license.

For DDOS versions lower than 6.0:

# license show For DDOS 6.0:

# elicense show all There are a couple circumstances where issues with the SSH keys may occur. A configuration with Avamar may need to re-create and re-add the SSH keys. With Networker, the SSH keys in the known.hosts file need to be removed.

29

Verifying the Cloud Tier configuration

If Cloud Tier was configured on the source controller, use this procedure to verify that it is properly configured on the destination controller.

Steps

1. Check that the following matches the original settings recorded while preparing for the upgrade.

# storage show all # cloud unit list # cloud profile show all # cloud clean frequency show # cloud clean throttle show # data-movement policy show # data-movement schedule show # data-movement throttle show

2. If cloud cleaning was stopped earlier, it can be resumed here.

# cloud clean start 3. Confirm that cloud cleaning is now running with:

# cloud clean status 4. Verify same passphrase set as before the headswap by running command system passphrase set.

# system passphrase set

Restoring and verifying the SCSI Target configuration

If VTL or DDBoost over Fibre Channel services were configured on the source controller, use this procedure to restore and verify the configuration on the destination controller.

Steps

1. Enable SCSITARGET.

a. Determine if SCSITARGET is enabled.

# scsitarget status b. If SCSITARGET is not enabled, enable SCSITARGET.

# scsitarget enable 2. If VTL was used on the source system, enable VTL.

a. Determine if VTL is enabled.

# vtl status b. If VTL is not enabled, enable VTL.

# vtl enable 3. Display the Fibre Channel port information on the destination system.

# scsitarget endpoint show detailed To view this information in the autosupport file, search for SCSITARGET Endpoint Show Detailed or Enclosure Show All. To view this information in DD System Manager, select Hardware > Fibre Channel > Resources and expand the Endpoints area.

4. If you did not complete the Fibre Channel port mapping table when you collected information from the source system, use the information from the source system autosupport file to complete the table.

To view this information in the autosupport file, search for SCSITARGET Endpoint Show Detailed or Enclosure Show All.

5. Update the endpoints on the destination controller as follows:

a. Display the current endpoint list.

30

# scsitarget endpoint show list b. Disable all endpoints.

# scsitarget endpoint disable all c. Delete the endpoints.

For systems running DDOS 5.7.X, delete each endpoint that existed on the source controller.

# scsitarget endpoint del ep-old

For all other DDOS versions, delete each automatically created endpoint.

# scsitarget endpoint del ep-new

The command syntax is scsitarget endpoint del endpoint-spec. Repeat this command using the name of each endpoint to be deleted.

d. Modify the remaining endpoints.

For systems running DDOS 5.7.X, rename the automatically created endpoints to match the names of the endpoints from the source controller, and change the WWPNs of the automatically created endpoints to match the WWPNs of the endpoints from the source controller.

# scsitarget endpoint rename ep-new ep-old

# scsitarget endpoint modify wwpn old-ep-wwpn

For all other DDOS versions, reassign each of the migrated endpoints to the correct port.

# scsitarget endpoint modify endpointfc-0 system-address 4a

The command syntax is scsitarget endpoint rename src-endpoint-name dst-endpoint-name . Replace src- endpoint-name with the current endpoint name and dst-endpoint-name with the endpoint name from the source controller.

The command syntax is scsitarget endpoint modify wwpn wwpn. Replace wwpn with the WWPN of the endpoint from the source controller.

The command syntax is scsitarget endpoint modify endpoint-spec system-address address. Replace endpoint-spec with the endpoint name and address with the new port number.

e. Verify that the endpoints match the table.

# scsitarget endpoint show detailed f. Enable the updated endpoints.

# scsitarget endpoint enable all 6. Connect the SAN cables to the Fibre Channel ports on the destination system.

To avoid switch zoning and initiator configuration issues, the best practice is to reconnect newly mapped ports in the destination system to the same SAN port based on the Fibre Channel port mapping. For example, if source port 5a is mapped to destination port 6a, connect 6a to the same switch port (assuming it is in a fabric configuration) to which port 5a was connected.

7. Verify that all of the devices are visible in the group with the new port numbers.

# vtl group show groupname 8. Verify that all VTL option settings are set correctly.

# vtl option show all 9. Update access groups.

# scsitarget group modify Use this command to update the primary and secondary port configured for all access groups based on the Fibre Channel port mapping.

31

Verifying the network configuration and connectivity

About this task

Network connectivity is required for system access and for alert and autosupport report forwarding. It is a good practice to verify the configuration and network connectivity after a controller upgrade.

Steps

1. To view the current network configuration, use the following commands.

# net show settings # net show config

2. Compare the current network configuration to the map tables recorded earlier.

Table 21. Example Destination Controller Network Settings

Port Enabled State DHCP IP Address Netmask/ prefix length

Type Additional Setting

ethMa yes running no 192.168.10.71 255.255.255. 0

n/a

fe80::260:16f f:fe5c:90dc* *

/64

NOTE: The replacement system has a different host and BMC MAC address, so it might receive a different IP address

from a DHCP server.

3. Use the net ping command to verify connectivity to the default gateway.

If the net ping command is unsuccessful, resolve the problem before continuing.

4. Use the net ping command to verify connectivity to DNS and all external hosts as appropriate.

5. Test all interface failovers as appropriate.

Verifying CIFS, NFS, and DD Boost connectivity

About this task

Follow the steps that apply to the communication protocols that the system uses.

Steps

1. If CIFS is configured in active-directory authentication mode, check for domain connectivity by running:

# cifs troubleshooting domaininfo

32

2. Examine the command output. If lsa-activedirectory-provider status is unknown, rejoin the domain by running the command:

# cifs set authentication active-directory realm

NOTE: An extra step is required if CIFS was in Workgroup mode prior to headswap. In this case, to join an Active

Directory domain after the headswap, run the cifs reset authentication command prior to joining the domain.

3. Verify CIFS connectivity by remounting any CIFS shares, enable CIFS, and verify the operation of the backup application by running these commands:

# cifs status # cifs enable # cifs show active If you are not able to reestablish CIFS connectivity, contact Support for assistance.

4. Verify NFS connectivity by remounting any NFS shares, enable NFS, and verify the operation of the backup application by running these commands:

# nfs status # nfs enable # nfs show clients

5. Verify DD Boost connectivity by running these commands:

# ddboost show user-name # ddboost storage-unit show # ddboost status # ddboost fc status

6. Enable DD Boost by running:

# ddboost enable

Verifying the file system status

About this task

The file system should be enabled at least once with all the external shelves attached to complete the upgrade.

Steps

1. Run the following commands again and compare their output with the output prior to the head swap. The outputs should be identical, except for cases where port numbers have changed (the system show ports command, for example). If other aspects of the output do not match, contact customer support.

# disk show state # disk status # disk show hardware # system show ports # disk multipath status # filesys show space # enclosure show topology # replication show config

2. Check that the encryption, snapshot, and retention lock settings match the original configuration.

Verifying the Replication status

Steps

1. Enable replication.

# replication enable

33

2. Run the following command and compare the output with the output prior to the head swap. Verify that the number of contexts and the configuration is the same as before.

# replication show config 3. Run the following command, verify the active status of each context, and ensure all contexts are enabled and connected.

This verification requires that all replica systems are in an active state. If a replica system is not active at the time, re-verify when it is.

# replication status all 4. Since the hostname and IP are now changed, login to the source system and modify the replication configuration.

# replication modify connection-host [port ]

Updating IPMI remote access

Steps

1. Enter the following commands.

# ipmi show config # ipmi config port {dhcp | ipaddress ipaddr netmask mask gateway ipaddr}

2. Add users back with the following command.

# ipmi user add user {password password}

NOTE: A system reboot is recommended after completing any configuration changes to ensure they remain after a

system reboot is performed.

Generating an Autosupport report

When the upgrade and the destination controller configuration are complete, it is a good practice to generate and save an autosupport report with the complete system configuration.

Prerequisites

Verify the network configuration and connectivity to the default gateway before starting this procedure.

Steps

1. To verify that the system can send email messages, use the autosupport test email youremail command to send a message to yourself and verify that you received the email.

If the test email is unsuccessful, resolve the problem before continuing to the next step.

2. To capture the destination controller configuration, generate an autosupport report and send it in an email message to yourself.

# autosupport send username@yourcompany.com

34

Notes, cautions, and warnings

NOTE: A NOTE indicates important information that helps you make better use of your product.

CAUTION: A CAUTION indicates either potential damage to hardware or loss of data and tells you how to avoid the

problem.

WARNING: A WARNING indicates a potential for property damage, personal injury, or death.

Manualsnet FAQs

If you want to find out how the DD OS 7.5 Dell works, you can view and download the Dell DD OS 7.5 Operating System Controller Upgrade Guide on the Manualsnet website.

Yes, we have the Controller Upgrade Guide for Dell DD OS 7.5 as well as other Dell manuals. All you need to do is to use our search bar and find the user manual that you are looking for.

The Controller Upgrade Guide should include all the details that are needed to use a Dell DD OS 7.5. Full manuals and user guide PDFs can be downloaded from Manualsnet.com.

The best way to navigate the Dell DD OS 7.5 Operating System Controller Upgrade Guide is by checking the Table of Contents at the top of the page where available. This allows you to navigate a manual by jumping to the section you are looking for.

This Dell DD OS 7.5 Operating System Controller Upgrade Guide consists of sections like Table of Contents, to name a few. For easier navigation, use the Table of Contents in the upper left corner.

You can download Dell DD OS 7.5 Operating System Controller Upgrade Guide free of charge simply by clicking the “download” button in the upper right corner of any manuals page. This feature allows you to download any manual in a couple of seconds and is generally in PDF format. You can also save a manual for later by adding it to your saved documents in the user profile.

To be able to print Dell DD OS 7.5 Operating System Controller Upgrade Guide, simply download the document to your computer. Once downloaded, open the PDF file and print the Dell DD OS 7.5 Operating System Controller Upgrade Guide as you would any other document. This can usually be achieved by clicking on “File” and then “Print” from the menu bar.