Host modes and HMOs for VSP E series

HMOs specific to this host mode: 68, 88, 91, 122, 131

HMOs available to this host mode: 2, 7, 13, 22, 25, 39, 68, 71, 78, 80, 88, 91, 96, 97, 113, 122, 131

HMO best practice: 2, 22, 25, 68

HMOs specific to this host mode: 54, 63, 68, 88, 110, 114, 122

HMOs available to this host mode: 2, 7, 13, 22, 25, 39, 54, 63, 68, 71, 78, 80, 88, 96, 97, 110, 113, 114, 122

HMO best practice: See host mode 21 [VMware Extension] in this table.

HMOs specific to this host mode: 12, 33, 43, 60, 88

HMOs available to this host mode: 2, 7, 12, 13, 22, 25, 33, 39, 43, 60, 71, 78, 80, 88, 96, 97, 113

HMO best practice: 2, 12, 22, 25, 60

HMOs specific to this host mode: 33

HMOs available to this host mode: 2, 7, 13, 22, 25, 33, 39, 71, 78, 80, 96, 97, 113

HMO best practice: 2, 22, 25

HMOs available to this host mode: 2, 7, 13, 14, 22, 25, 39, 71, 78, 80, 96, 97, 113

HMO best practice: 2, 22, 25

HMO specific to this host mode: 88

HMOs available to this host mode: 2, 7, 13, 22, 25, 39, 71, 78, 80, 88, 96, 97, 98, 113

HMO best practice: 2, 22, 25

HMOs available to this host mode: 2, 7, 13, 22, 25, 39, 71, 78, 80, 96, 97, 113

HMO best practice: 2, 22, 25

HMOs specific to this host mode: 6, 40, 73, 88, 105, 110, 114, 122

HMOs available to this host mode: 2, 6, 7, 13, 22, 25, 39, 40, 71, 73, 78, 80, 88, 96, 97, 105, 110, 113, 114, 122

HMO best practice: See host mode 2C [Windows Extension] in this table.

HMOs specific to this host mode: 15, 88

HMOs available to this host mode: 2, 7, 13, 15, 22, 25, 39, 71, 78, 80, 88, 96, 97, 113, 124

HMO best practice: 2, 15, 22, 25

When host mode 0F AIX ^® is set, NACA ( Normal Auto Contingent Allegiance) is enabled automatically. To ensure that the host responds correctly to a port with host mode 0F AIX enabled:

1. Shut down the host to be connected to the port.
2. Set the host mode on the port to 0F AIX, and enable the desired HMOs.
3. Connect the cables from the host to the port.
4. Boot the host.

Caution: If the host mode of a port is changed to 0F AIX and the host had already recognized the Inquiry response for an LDEV on the port before host mode 0F AIX was set, the host might experience loss of device access (LDEV or LU). The host must be rebooted after the host mode is changed to 0F AIX to recognize the change in the response to the standard inquiry.

HMOs specific to this host mode: 54, 63, 68, 82, 83, 88, 110, 114, 122

HMOs available to this host mode: 2, 7, 13, 22, 25, 39, 54, 63, 68, 71, 78, 80, 82, 83, 88, 96, 97, 110, 113, 114, 122

HMO best practice: 2, 22, 25, 54, 63, 68, 110

HMOs specific to this host mode: 6, 40, 73, 88, 105, 110, 114, 122

HMOs available to this host mode: 2, 6, 7, 13, 22, 25, 39, 40, 71, 73, 78, 80, 88, 96, 97, 105, 110, 113, 114, 122

HMO best practice: 2, 22, 25, 40, 110

Unreliable access, errors, and performance problems will result if vCenter maps an RDM disk to a virtual machine that is part of a VMware host group with VMware host mode configured. This is because Windows and Linux use SCSI-3 version protocol and VMware uses SCSI-4 version protocol.

Linux-based OS (such as Red Hat Linux) or Windows Virtual Machine (VM) server hosts that use virtual disks and not RDM do not need any host group mapping because they are dependent on the VMware Hypervisor host for creating a virtual disk (VMDK file) from the VMFS file system residing on a Hypervisor-only data stores, because the VM disk files are ultimately VMware VMFS file system files and not actual disks.

To add VMware Hypervisor (not virtual machine) hosts to one host group, you must set this host mode and these HMOs for the host group:

• Host mode: 21 VMware Extension
• Host mode options: 2, 22, 25, 54, 63, 68, and 110 (for details, contact customer support)

Restrictions for VMware server hosts added to the host group:

Do not attempt to set Windows or Linux HMO functions on a VMware Hypervisor, they cannot be used by VMware:

• Thin Provisioning
• Offloaded Data Transfer (ODX)

The above functions are handled by the HMO 54 and 63 in the VMware host group instead.

In the following conditions, Good Status is expected.

When this HMO is enabled, the storage system will switch Test Unit Ready response to Good Status from Reservation Conflict.

Use this HMO when any of the following conditions is met:

• Windows Server Failover Clustering (WSFC) is used.
• Microsoft Failover Cluster (MSFC) is used.
• Linux with cluster configuration is used.
• Symantec Cluster Server (previously known as Veritas Cluster Server (VCS)) is used.
• The response for SPC-3 (Good Status) is required.

When this HMO is enabled, the storage system will switch the TPRLO response to LOGO from PRLO.

Use this HMO when all these conditions are satisfied:

• An Emulex host bus adapter is used in a Windows environment.
• The mini-port driver is used.
• TPRLO=2 is specified for the mini-port driver parameter of the host bus adapter.

However, in some cases, the host might expect Unit Attention.

When this HMO is enabled, the storage system will return the Unit Attention response.

Use this HMO when any of the following conditions is met:

• SUN StorEdge SAN Foundation Software Version 4.2 or later is used.
• You want to enable the automatic recognition of the device's increase or decrease when connecting a genuine SUN HBA.
• You want the Unit Attention (UA) response to be returned to the host at SCSI path change (sense code: REPORTED LUNS DATA HAS CHANGED).

Note: If the UA response occurs frequently and the load on the host becomes high, the data transfer cannot be started on the host side and timeout might occur.

Note: Enable HMO 13 only when requested to do so. HMO 13 is enabled for each port. Set HMO 13 for the host group 00 of the target port.

* HACMP 4.5 version 4.5.0.13 or later, HACMP 5.1 version 5.1.0.4 or later, or HACMP 5.2 or later.

In the following condition, GOOD status is expected when a reserved volume receives the MODE SENSE command from a node that is not reserving this volume.

When this HMO is enabled, the storage system will switch the MODE SENSE command response to Good status from Reservation Conflict.

Use this HMO when any of these conditions is satisfied:

• Veritas Cluster Server is used.
• An OS or middleware that supports SPC-4 is used.
• A single-site or GAD dual-site Clustered VMDK configuration is used.

In the following conditions, Good Status (SPC-3 response) is expected.

When this HMO is enabled, the storage system will switch this command response to Good Status from Reservation Conflict.

Use this HMO when one of these conditions is satisfied:

• Windows Server Failover Clustering (WSFC) is used.
• Microsoft Failover Cluster (MSFC) is used.
• Symantec Cluster Server (previously known as Veritas Cluster Server (VCS)) is used.
• There is no registered key to be deleted when running the PERSISTENT RESERVE OUT command.

Note: Host types other than above listed in Special Direction expect the response when the option is set to OFF.

• Windows Server Failover Clustering (WSFC) is used.
• Microsoft Failover Cluster (MSFC) is used.
• Symantec Cluster Server (previously known as Veritas Cluster Server (VCS)) is used.
• There is no registered key to be deleted when running the PERSISTENT RESERVE OUT command.

• Range of job resetting.
• Range of UAs (Unit Attentions) defined.

When a VMware, Linux, or Windows server is added to a host group, select VMware Extension for the host mode and then set the option to ON.

For further information, see the section below this table.

When this HMO is enabled, the VENDOR SPECIFIC EXTENDED COPY command that the VAAI (vStorage APIs for Array Integration) function of VMware ESX/ESXi 4.1 or later issues to the storage system is available.

Notes:

1. The HMO is set to ON only when a VMware server of ESX/ESXi 4.1 or later using the VAAI (vStorage APIs for Array Integration) function is connected.
2. If the HMO is not applied, the VMware support function, Cloning file blocks, cannot be used.
3. When the VMware ESX/ESXi 4.x server is connected, set HMO 54 to ON. HMO 63 is not necessary.
4. When the ESXi 5.x, ESXi 6.x, or later server is connected, set both HMO 54 and HMO 63 to ON.
5. After setting HMO 54 to ON, perform an operation (such as a server reboot) leading to reissuing the INQUIRY command from the host.

Use this HMO when you connect the storage system to VMware ESXi 5.0 or later and use the VAAI function for iSCSI T10.

Notes:

1. After setting this HMO, perform the operation, such as the server reboot that reissues the INQUIRY command on the host side.
2. When ESXi 5.x or later is connected, set both HMO 54 and HMO 63 to ON.

Note: After setting this HMO, perform the operation, such as the server reboot that reissues the INQUIRY command on the host side.

Use this HMO when using the Page Reclamation function with a Linux host.

Notes:

1. The option is applied when Dynamic Provisioning is used by Linux 2.6.33 or higher.
2. After setting HMO 68 to ON, perform the operation (such as a server reboot) that reissues the INQUIRY command from the host.

• Thin Provisioning function
• OS Offloaded Data Transfer (ODX) function

Use a combination of this HMO and HMO 114 when a VXIO driver with Veritas InfoScale 8.0 or later is used on the Windows OS host.

Notes:

1. After setting HMO 73 to ON, perform an operation (such as a server reboot) that reissues the INQUIRY command on the host side.
2. This HMO is related to HMO 110. For behaviors when both of the options are set to ON, contact customer support.

   When both options are set to ON, the setting of HMO 73 is prioritized over that of HMO 110.

• Global-active device is used in the configuration with the data centers (Metro configuration).
• Hitachi Dynamic Link Manager (HDLM) is used as the alternative path software.
• The host group is on the non-optimized path of HDLM.
• The performance deterioration of I/O responses can be avoided without I/O using the non-optimized path of HDLM.

Caution:

If this HMO is set on the wrong host group, load balance of HDLM with the option cannot be performed intended operations, which might cause a performance issue.

For instance: When the CHAP authentication is performed at the time of the discovery login in the iSCSI environment of the VMware host and storage system

• Configuring redundant paths between the VMware host and storage system.
• Waiting for replying of the target information from the ports other than ports of discovery login.

Notes:

• When this HMO is enabled, the same iSCSI target name and the same target number as the iSCSI target that receives the Send Target request are reported as the targets. For the target information that requires a portal list report, make sure that the iSCSI target name and the target number match with those of the iSCSI target under the port that receives the Send Target request.
• Set this HMO on all targets that require reporting.

• Enabled: LUN path definition is enabled.
• Disabled: LUN path definition is disabled.

This HMO setting is not supported when the HDLM or VxVM DMP multipath software is used.

Notes:

1. Apply this HMO when all the following conditions are met:
   • You are using the GAD function to migrate volumes in multiple migration source storage systems that use the same server.
   • You need to converge target ports used on the migration target storage system.
   • Host mode 00, 01 03, 09, 0C, 0F, 21, or 2C is used.
2. Do not apply this HMO to host groups without the specified host modes enabled. Applying this HMO to host groups with other host modes can cause the following problems:
   • Path addition from the server to the migration target storage system might fail.
   • Display of devices that the server recognizes might be invalid.
3. When using VMware NMP, use the same LUN# for the migration source and migration target.
4. If a LUN path is defined from the host group to an LDEV defined in a virtual storage machine different from the one to which the host group belongs, this HMO cannot be set to OFF.

For example, this HMO is used to configure a SAN Boot system by directly connecting the Fibre Channel channel board on the storage system to the Broadcom (Emulex) 16-Gbps or 32-Gbps Fibre Channel HBA that supports the connection to the storage system.

Note: For details about the conditions of the supported connection configuration, contact customer support. After you set this option, do not use the port for an external storage connection or a remote copy connection.

Note: This HMO option setting can be specified for each port. Set this HMO to host group 0 of the target port.

• Range of job resetting.
• Range of UAs (Unit Attentions) defined.

Note: This HMO was intended to be used only for NAS but was never implemented in NAS software, and therefore it is currently not supported by NAS and should not be used.

Note: For details about the conditions of the supported connection configuration, contact customer support. After you set this option, do not use the port for external storage connection or remote copy connection.

Note: This HMO option setting can be specified for each port. Set this HMO to host group 0 of the target port.

Notes:

• After setting this HMO, perform the operation, such as the server reboot that reissues the INQUIRY command on the host side.
• If both HMO 73 and HMO 110 are set, HMO 73 takes precedence over HMO 110.

Notes:

1. This HMO option setting can be specified for each port. Set this HMO on the host group 0 of the target port.
2. The setting of this HMO applies to all iSCSI targets on the port. Therefore, set this HMO only in iSCSI target 00 of the port.

• Use a combination of this HMO and HMO 63 when connecting to VMware ESXi 6.5 or later, and when using the zero data page reclamation function, which is performed automatically if files on the VMFS (Virtual Machine File System) are deleted.
• Use a combination of this HMO and HMO 73 when a VXIO driver with Veritas InfoScale 8.0 or later is used on the Windows OS host.

Note: After setting this HMO, perform the operation, such as the server reboot, that reissues the INQUIRY command on the host side.

Note: If this option is set for a host other than a Windows/Linux/VMware host, an I/O might not be issued from the host.

Note: If this option is set for a configuration other than the AIX™ host and HDLM-EX, an I/O might not be issued from the host.