Home > Workload Solutions > SAP > Guides > SAP HANA TDI Guides > Dell Validated Design for SAP HANA TDI with Dell PowerStore Storage > Host connection and setup using FC SAN
When you connect an SAP HANA host to a PowerStore system using the FC protocol, you must connect two host bus adapter (HBA) ports supporting 16 Gb/s or 32 Gb/s link speed to the PowerStore system. Although it is possible to use a single dual-port HBA, Dell Technologies highly recommends using two HBAs for HA and redundancy.
Connect and zone each port over the FC SAN network to two FC FE ports per storage processor on the PowerStore system. This configuration produces two active and two standby paths per LUN, as shown in the following figure:
After the HBAs of the SAP HANA server have been zoned to the PowerStore FC FE ports, use the PowerStore Manager UI to create a host entry in the PowerStore system and add volumes to the host.
The Add Host dialog is displayed:
Click Next. The following page is displayed:
A list of automatically discovered initiators is displayed:
A Host Summary page opens:
After the host is created, you can create and add (map) the volumes. Each host requires a data and a log volume for the SAP HANA persistence.
The Create Volumes page opens:
PowerStore can group volumes together into a volume group. A volume group is useful for operations that affect multiple volumes simultaneously, for example, when the same volumes are added (mapped) to multiple hosts.
To create a volume group:
After you have created the SAP HANA data and log volumes using the PowerStore Manager UI, format the volumes on the SAP HANA host.
Follow these steps:
rescan-scsi-bus.sh
fdisk -l | grep -B1 -A4 PowerStore
The output should look like this example:
hana01:~ # fdisk -l | grep -B1 -A3 Power
Disk /dev/sdr: 512 GiB, 549755813888 bytes, 1073741824 sectors
Disk model: PowerStore
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
--
Disk /dev/sdq: 1 TiB, 1099511627776 bytes, 2147483648 sectors
Disk model: PowerStore
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
The /etc/multipath.conf file controls multipathing. PowerStore default multipath values are distributed with both SLES and Red Hat Enterprise Linux and should not need to be configured. This file does not exist by default, however. If you want to view or edit the multipath.conf settings, create the file by running the following command:
multipath -t > /etc/multipath.conf
device {
vendor "DellEMC"
product "PowerStore"
path_selector "queue-length 0"
path_grouping_policy "group_by_prio"
path_checker "tur"
detect_prio "yes"
failback "immediate"
no_path_retry 3
rr_min_io_rq 1
fast_io_fail_tmo 15
max_sectors_kb 1024
}
Note: For the latest PowerStore MPIO configuration settings, see the Dell EMC PowerStore Host Configuration Guide.
systemctl restart multipathd
multipath -ll
The following output shows two devices:
368ccf09800dce784009f298f053f1e03 dm-4 DellEMC,PowerStore
size=1.0T features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
|-+- policy='queue-length 0' prio=50 status=active
| |- 18:0:0:3 sdp 8:240 active ready running
| `- 20:0:1:3 sdgh 131:208 active ready running
`-+- policy='queue-length 0' prio=10 status=enabled
|- 18:0:1:3 sdt 65:48 active ready running
`- 20:0:0:3 sdgd 131:144 active ready running
--
368ccf0980058a0131a024b9cfd171c03 dm-2 DellEMC,PowerStore
size=512G features='1 queue_if_no_path' hwhandler='1 alua' wp=rw
|-+- policy='queue-length 0' prio=50 status=active
| |- 18:0:1:1 sds 65:32 active ready running
| `- 20:0:0:1 sdgc 131:128 active ready running
`-+- policy='queue-length 0' prio=10 status=enabled
|- 18:0:0:1 sdo 8:224 active ready running
`- 20:0:1:1 sdgg 131:192 active ready running
When the block devices are under multipath control, you can format them by using XFS and then mount them as required.
mkfs.xfs -K /dev/mapper/368ccf09800dce784009f298f053f1e03
/dev/mapper/368ccf09800dce784009f298f053f1e03 /hana/data xfs rw,relatime,noquota 0 0
/dev/mapper/368ccf0980058a0131a024b9cfd171c03 /hana/log xfs rw,relatime,noquota 0 0
When you install an SAP HANA multimode scale-out cluster, the SAP HANA storage connector (fcClient) mounts the devices during SAP HANA startup.
The default queue depth settings for HBAs and LUNs work well with SAP HANA devices in most situations. For recommendations for queue depth to achieve optimal performance and maximum scalability, see the latest Dell PowerStore Host Configuration Guide.
To display the current queue depth values, run the commands for your HBA type:
Emulex HBAs:
cat /sys/class/scsi_host/host20/lpfc_hba_queue_depth
cat /sys/class/scsi_host/host20/lpfc_lun_queue_depth
Determine the host number (in this example, host20) by running the multipath -ll command, as shown in step 6 of Scan LUNs on the host and create file systems. Linux assigns a specific number to every connected HBA port. The example uses host18 and host20.
QLogic HBAs:
cat /sys/module/qla2xxx/parameters/ql2xmaxqdepth
To change the LUN queue depth to a new value:
cd /boot
sync
dracut -f
In an SAP HANA scale-out cluster, a single shared file system (/hana/shared) is required and must be mounted on every node. Most SAP HANA installations use a Network File System (NFS) for this purpose. PowerStore unified storage systems can provide this file system with the NAS option. For more information, see Creating the SAP HANA shared file system.