Home > Storage > PowerStore > Databases and Data Analytics > Dell PowerStore: Oracle Best Practices > LUN partitioning
A PowerStore volume intended for ASM can be configured as a raw LUN or as a single partitioned LUN. If ASMLib is used to manage the LUN, the LUN must be configured as a single partitioned LUN. If ASMLib is not used, the LUN can either be configured as raw or as a single partitioned LUN by ASM. Choosing the type of LUN to use depends on the environment, infrastructure design, and daily operations.
Note: For this paper, raw unpartitioned LUNs were used for ASM.
If ASMLib or ASMFD are intended to manage ASM disks created from NVMe volumes, review the NVMe information available on Oracle Support.
When partitioning a LUN, aligning the partition on the 1M boundary is recommended. Use either fdisk or parted to create partitions smaller than 2 TB. To create partitions 2 TB or larger, use parted.
Before creating the partition, label the device as GPT. Then, specify the partition offset at 2048 sector (1M). The following commands add a label to the device and create a single partition that takes up the entire LUN.
# parted /dev/mapper/ora-asm-data-001
GNU Parted 3.1
Using /dev/mapper/ora-asm-data-001
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) mklabel gpt
(parted) quit
Information: You may need to update /etc/fstab.
# parted /dev/mapper/ora-asm-data-001 mkpart primary 2048s 100%
After the partition is created, use the partition file /dev/mapper/ora-asm-data-001p1 to create the ASMLib volume.
When creating a file system, create the file system on a properly aligned partitioned device.
In Oracle Linux 7 and earlier, block devices supported single-queue I/O schedulers. As of Oracle Linux 8, single-queue I/O schedulers are no longer available. Only the following multi-queue schedulers are available:
For a description and use cases for the schedulers, see the Red Hat documentation:
Red Hat Enterprise Linux 8 Monitoring and managing system status and performance
Note: The effectiveness of any I/O scheduler can depend on specific workloads, hardware configurations, and Linux kernels. Dell Technologies recommends consulting the latest documentation from both the Linux kernel and Oracle, and performing thorough testing to determine which scheduler best suits the Oracle database workload.
To verify the current I/O scheduler for a block device, run the following command:
# egrep "*" /sys/block/sd*/queue/scheduler
/sys/block/sda/queue/scheduler:[mq-deadline] kyber bfq none
/sys/block/sdaa/queue/scheduler:[none] mq-deadline kyber bfq
/sys/block/sdab/queue/scheduler:[none] mq-deadline kyber bfq
/sys/block/sdac/queue/scheduler:[none] mq-deadline kyber bfq
To set the I/O schedule persistently, use TuneD, or udev. For more information, see: Red Hat Enterprise Linux 8 Monitoring and managing system status and performance