HP StorageWorks XP Disk Array Configuration Guide (T5278-96047, May 2011)
Table Of Contents
- XP Disk Array Configuration Guide
- Contents
- 1 Overview
- 2 HP-UX
- Installation roadmap
- Installing and configuring the disk array
- Installing and configuring the host
- Connecting the disk array
- Configuring disk array devices
- Verifying the device files and drivers
- Creating the device files
- Creating the physical volumes
- Creating new volume groups
- Creating logical volumes
- Creating the file systems
- Setting the I/O timeout parameter
- Creating the mount directories
- Mounting and verifying the file systems
- Setting and verifying the auto-mount parameters
- 3 Windows
- 4 Novell NetWare
- 5 NonStop
- 6 OpenVMS
- 7 VMware
- 8 Linux
- 9 Solaris
- Installation roadmap
- Installing and configuring the disk array
- Installing and configuring the host
- Connecting the disk array
- Configuring disk array devices
- Configuring for use with Veritas Volume Manager 4.x and later
- 10 IBM AIX
- 11 Citrix XenServer Enterprise
- 12 Troubleshooting
- 13 Support and other resources
- A Path worksheet
- B Path worksheet (NonStop)
- C Disk array supported emulations
- D Using Veritas Cluster Server to prevent data corruption
- E Reference information for the HP System Administration Manager (SAM)
- F HP Clustered Gateway deployments
- Glossary
- Index
Table 32 LUSE device parameters (HP-UX) (continued)
Max physical extent size
(MPE)
Physical extent size (PE)Device type
382058n = 22
399428n = 23
416798n = 24
434158n = 25
451528n = 26
468898n = 27
486258n = 28
503628n = 29
520988n = 30
538358n = 31
555728n = 32
573088n = 33
590458n = 34
607828n = 35
625188n = 36
defaultdefaultn = 2 to 3OPEN-L*n
defaultdefaultOPEN-8/9/E-CVS OPEN-V
default870 to 119731(MB) × N1OPEN-8/9/E*n-CVS
OPEN-V*n (n = 2 to 36)
N28119732 to (MB) × N1
N1 = [ VCS volume capacity (in MB) from Remote Console PC ] × nN2 = N1 / PE ( means round up to next
integer) Example: CVS volume capacity is 6000 MB for OPEN-9*22-CVS volume:N1 = 6000 × 22 = 132000N2 =
132000/8 = 16500
SCSI TID map for Fibre Channel adapters
When an arbitrated loop (AL) is established or reestablished, the port addresses are assigned
automatically to prevent duplicate TIDs. With the SCSI over Fibre Channel protocol (FCP), there is
no longer a need for target IDs in the traditional sense.
SCSI is a bus-oriented protocol requiring each device to have a unique address because all
commands go to all devices. For Fibre Channel, the AL-PA is used instead of the TID to direct
packets to the desired destination.
Unlike traditional SCSI, when control of the loop is acquired, a point-to-point connection is
established from initiator to target. To enable transparent use of FCP, the operating system maps
a TID to each AL-PA.
The host maps SCSI protocol to Fibre Channel protocol and detects and accesses Fibre
Channel-connected devices using device files (/dev/dsk/c*t*d* and /dev/rdsk/c*t*d*)
in the same way as for SCSI-connected devices. The device files for Fibre Channel-connected
devices are configured in a different way from SCSI-connected devices, because Fibre Channel
supports 126 addresses per path while SCSI supports 16 TIDs per path.
The following table identifies the fixed mappings between the TID (drive) values assigned by the
operating system and the Fibre Channel native addresses (AL_PA/SEL_ID) for Fibre Channel
138 Disk array supported emulations