Host modes for VSP 5000 series
Host mode | When to select this host mode |
---|---|
00 [Standard] | When registering Red Hat Linux server hosts or IRIX server hosts in the host group.* 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 |
01 [(Deprecated) VMware] | Do not select this host mode. Select host mode 21 VMware Extension instead. 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. |
03 [HP] | When registering HP-UX server hosts in the host group. 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 |
05 [OpenVMS] | When registering OpenVMS server hosts in the host group. 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 |
07 [Tru64] | When registering Tru64 server hosts in the host group. 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 |
09 [Solaris] | When registering Solaris server hosts in the host group. 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 |
0A [NetWare] | When registering NetWare server hosts in the host group. HMOs available to this host mode: 2, 7, 13, 22, 25, 39, 71, 78, 80, 96, 97, 113 HMO best practice: 2, 22, 25 |
0C [(Deprecated) Windows] | Do not select this host mode. Select host mode 2C Windows Extension instead. 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. |
0F [AIX] | When registering AIX ® server hosts in the host group. 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:
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. |
21 [VMware Extension] | When registering VMware server hosts in the host group.* 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 |
2C [Windows Extension] | When registering Windows server hosts in the host group.* 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 |
*Use host mode 21 only for VMware server hosts (Hypervisor host accessing VMFS formatted data stores, not Raw Device Mapping (RDM) disks. But when the (Linux-based OS or Windows) Virtual Machine host directly recognizes and manages the I/O stack on LUs by RDM (Raw Device Mapping), then the disk should only be assigned a host mode that is the same as the operating system of the virtual machine, even when running on VMware Hypervisor. 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:
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:
The above functions are handled by the HMO 54 and 63 in the VMware host group instead. |
Host mode options for VSP 5000 series
HMO | Function | Host mode | Description |
---|---|---|---|
2 | VERITAS Database Edition/Advanced Cluster | Any | Purpose: By default, Reservation Conflict is returned for Test Unit Ready run from host without the PERSISTENT GROUP RESERVATION key setting. 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:
|
6 | TPRLO | 0C [(Deprecated) Windows] or 2C [Windows Extension] | Purpose: By default, PRLO is returned for TPRLO, however, LOGO might be expected under the following conditions. 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:
|
7 | Automatic recognition function of LUN | Any | Purpose: By default, Unit Attention response is not returned when adding a LUN using a host group. 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:
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. |
12 | No display for ghost LUN | 03 [HP] | Use this HMO when you want to suppress creation of device files for devices to which paths are not defined. |
13 | SIM report at link failure | Any | Use this HMO when you want SIMs (service information messages) to be issued if the number of link failures detected between the ports exceeds the threshold. 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. |
14 | HP TruCluster with TrueCopy function | 07 [Tru64] | When all of the following conditions are satisfied:
|
15 | HACMP | 0F [AIX] | Use this HMO when HACMP* or PowerHA® is used. * 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. |
22 | Veritas Cluster Server | Any | Purpose: By default, Reservation Conflict is returned to the MODE SENSE command from the host without the PERSISTENT GROUP RESERVATION key setting. 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:
|
25 | Support SPC-3 behavior on Persistent Reservation | Any | Purpose: By default, Reservation Conflict is returned for PERSISTENT RESERVE OUT (Service Action = REGISTER AND IGNORE EXISTING KEY) command if there is no registered key to be deleted. 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:
Note: Host types other than above listed in Special Direction expect the response when the option is set to OFF. |
33 | Set/Report Device Identifier enable | 03 [HP] or 05 [OpenVMS] | Use this HMO when all these conditions are satisfied:
Note: When this HMO is enabled in the host mode 05 [OpenVMS], make sure to set a UUID. |
39 | Change the nexus specified in the SCSI Target Reset | Any | Use this HMO when you want to enable the host to set these ranges on all initiators in the host group, not only the initiator that issues the job resetting when receiving Target Reset, such as in the IBM SVC connection environment:
|
40 | V-VOL expansion | 0C [(Deprecated) Windows] or 2C [Windows Extension] | Purpose: This HMO is used to control whether UA is returned to the host when a command from the host is received for the LU whose capacity has been expanded. Use this HMO when you want to automate recognition of the V-VOL capacity after increasing the V-VOL capacity. Note: This option is applied when UA returning to the host is required after the V-VOL capacity is expanded. |
43 | Queue Full Response | 03 [HP] | Use this HMO when the command queue is full in your storage system connected to an HP-UX host, and you want to respond Queue Full (instead of Busy) from the storage system to the host. |
51 | Round Trip Set Up Option | Any | Use this HMO if you want to adjust the response time of the host I/O, for example when the distance between the primary and secondary storage systems in the TrueCopy or global-active device pair is long (approximately 100 km) and point-to-point topology is used. For further information, see the section below this table. |
54 | (VAAI) Support Option for the EXTENDED COPY command | 01 [(Deprecated) VMware] or 21 [VMware Extension] | Purpose: This HMO enables the XCOPY command to be used. 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:
|
60 | LUN0 Change Guard | 03 [HP] | Use this HMO when HP-UX 11.31 is used, and you want to suppress the addition or deletion of LUN 0. |
63 | (VAAI) Support Option for vStorage APIs based on T10 standards | 01 [(Deprecated) VMware] or 21 [VMware Extension] | Purpose: This HMO enables the use of VAAI, which complies with SCSI T10. 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:
|
68 | Support Page Reclamation for Linux | 00 [Standard], 01 [(Deprecated) VMware], or 21 [VMware Extension] | Purpose: When this HMO is enabled, the storage system will change the response that Linux OS can issue the WriteSame command to use the Page Reclamation function. Use this HMO when using the Page Reclamation function with a Linux host. Notes:
|
71 | Change the Unit Attention for Blocked Pool-VOLs | Any | Purpose: There is a possibility to prevent a device file from being blocked while a DP pool is blocked. Use this HMO when a device file is blocked while a DP pool is blocked. Note: This option is applied if switching sense key/sense code returned as a response to Check Condition when a read/write I/O is received while a DP pool is blocked can prevent a device file from being blocked, and therefore the extent of impact can be reduced on host side. |
73 | Support Option for WS2012 | 0C [(Deprecated) Windows] or 2C [Windows Extension] | Purpose: Use this HMO if all the following functions provided by Windows Server 2012 (WS2012) or later are used with a Windows OS host running WS2012 or later:
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:
|
78 | The non-preferred path option | Any | Purpose: To prevent the decrease of the response time of the host I/O using Hitachi Dynamic Link Manager (HDLM). Use this HMO if the following conditions are met:
Notes:
|
80 | Multi Text OFF Mode | Any | By using the iSCSI interface, if the storage system connects with the host of which OS is not supported of the Multi Text function. For instance, connecting the storage system and the host of RHEL5.0 which does not support the Multi Text-function. |
81 | NOP-In Suppress Mode | 0A [NetWare] | In the environment with iSCSI connection, the delayed reply of the Delayed Acknowledgment function, which is located on the upper layer, is restrained by sending NOP-IN for executing sense commands such as Inquiry, Test unit ready, or Mode sense. However, select this option when connecting the storage system and the host that does not need to send the NOP-IN.
|
82 | Discovery CHAP Mode | 21 [VMware Extension] | Select this option when the CHAP authentication is performed at the time of the discovery login in the iSCSI connection environment. 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 |
83 | Report iSCSI Full Portal List Mode | 21 [VMware Extension] | Use this HMO when all these conditions are met:
Notes:
|
88 | Port consolidation | 00 [Linux], 01 [(Deprecated) VMware], 03 [HP], 09 [Solaris], 0C [(Deprecated) Windows], 0F [AIX], 21 [VMware extension], 2C [Windows extension] | Use this HMO when converging multiple host-target ports used in the migration source storage system on one host group of the migration target storage system by enabling path definition from the host group to LDEVs of multiple virtual storage machines.
This HMO setting is not supported when the HDLM or VxVM DMP multipath software is used. Notes:
|
91 | Disable I/O wait for OpenStack Option | 00 [Standard] | Use this HMO when manually creating host groups or iSCSI targets that are used as the I/O data paths for OpenStack. |
96 | Change the nexus specified in the SCSI Logical Unit Reset | Any | Use this HMO when you want to enable the host to set these ranges on all initiators in the host group, not only the initiator that issues the job resetting when receiving an LU Reset, such as in the IBM SVC connection environment:
|
97 | Proprietary ANCHOR command support | 00 [Standard] | Do not enable HMO 97 when connecting with Hitachi NAS Platform (HNAS). This setting does not enable any HNAS functionality at this time. 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. |
105 | Task Set Full response in the event of I/O overload | 0C [(Deprecated) Windows] or 2C [Windows Extension] | Use this HMO when you want to return Task Set Full response from the storage system to the host when an overload of I/Os occurs on the storage system. |
110 | ODX support for WS2012 | 01 [(Deprecated) [VMware], 21 [VMware Extension], 0C [(Deprecated) Windows] or 2C [Windows Extension] | Purpose: This HMO is used to enable the Windows Server 2012 Offloaded Data Transfer (ODX) function to work. Use this HMO when ODX provided by Windows Server 2012 or later is used, in the environment with which the WS2012 or later host connects. Notes:
|
113 | iSCSI CHAP Authentication Log | Any | Use this HMO when the result of the CHAP authentication is output to the audit log (DKC). Notes:
|
114 | The automatic asynchronous reclamation on ESXi 6.5 or later | 01 [(Deprecated) [VMware], 21 [VMware Extension], 0C [(Deprecated) Windows], or 2C [Windows Extension] | If either of the following conditions is met:
Notes:
|
122 | Task Set Full response after reaching QoS upper limit | 01 [(Deprecated) VMware], 21 [VMware Extension], 0C [(Deprecated) Windows], 2C [Windows Extension], or 00 [Standard] | Use this HMO when a Windows/Linux/VMware host is connected, and when the QoS upper limit is reached, if you want to return a TASK SET FULL response to the host in order not to retain an I/O inside the storage system. 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. |
124 | Guaranteed response during controller failure | 0F [AIX] | Use this HMO when connecting to an AIX™ host and using HDLM-EX as the alternative path software. 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. |
131 | WCE bit OFF mode | 00 [Standard] | Purpose: This HMO is used to change the WCE (Write Cache Enable) bit in Cache Mode page (08h) of the MODE SENSE command to OFF "Write Cache Disabled", which is returned to the host. Use this HMO when an I/O performance problem occurs while Oracle ASM is used in Linux environment. Notes:
|
HMO 51 details
- When the HMO 51 setting is changed to ON or OFF, a link-down and link-up occur on the target port, and an SSB log of link down is output on RCU (R-DKC).
- This HMO can work only when the microcode supporting this function is installed on both MCU (M-DKC) and RCU (R-DKC).
- The HMO setting is only applied to bi-directional ports when the 4HF32R PCB is used on RCU/MCU.
- When removing the 4HF32R PCB, the operation must be performed after setting HMO 51 to OFF.
- If this option is used, Point to Point setting is necessary.
- If HMO 51 is set to ON while SOM 769 is ON, path creation might fail after automatic port switching.
- Make sure to set HMO 51 from OFF to ON or from ON to OFF after the pair is suspended or when the load is low.
- The RCU Target, which is connected with the MCU where this option is set to ON, cannot be used for Universal Replicator.
- Path attribute change (Initiator Port - RCU-Target Port, RCU-Target Port - Initiator Port) accompanied with Hyperswap is enabled after setting HMO 51 to ON. If HMO 51 is already set to ON on the both ports, HMO 51 continues to be applied on the paths even after execution of Hyperswap.
- This function is available in the following configuration:
Supported version 4HF32R: All versions:
Supported configuration: Point to Point setting with HMO 49 and HMO 50 set to ON (In the case of 4HF32R, setting HMO 49 and HMO 50 is not necessary as it does not support the options) while the path distance between MCU and RCU meets requirements of each program product.
- This function is prepared for long distance data transfer. Therefore, if HMO 51 is set to ON with distance of less than 1 km, a data transfer error might occur on RCU side and the performance improvement might not reach the expected level.
-
When setting the option on a port where LUN security is enabled, set the HMO to ON only for a host group where WWN of the remote storage system is registered. In addition, to delete the host group, first set the option to OFF.
- The option cannot be applied to a port connected with multiple MCUs.