How to set reserve_lock attribute of IBM hdisk with shared array environment.

How to set the reserve_lock attribute of IBM hdisk with shared array environment to avoid "the disk operation error" whenever the node rebooting.

We can see a number of messages in the errpt log.

IDENTIFIER TIMESTAMP T C RESOURCE_NAME DESCRIPTION
A2205861 0326132409 P S SYSPROC Excessive interrupt disablement time
A39F8A49 0326132409 T S syserrlg ERROR LOGGING BUFFER OVERFLOW
DCB47997 0326125509 T H hdisk1024 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk1024 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk597 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk597 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk206 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk206 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk2986 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk2986 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk2595 DISK OPERATION ERROR
DCB47997 0326125509 T H hdisk2595 DISK OPERATION ERROR

Let's check the attribute of the hdisk in AIX side.
i.e.
[aix1:/]lsattr -El hdisk1024 -a reserve_lock
reserve_lock yes Reserve Device on open True

Further information about this can be found in the Dynamic Multi-Pathing Administrator's Guide.

https://docs.infoscale.com/

Disk Restrictions:
When using VCS in combination with VIO servers and their client partitions, you need to ensure that no reservations are placed on the shared storage. This enables client partitions on different systems to access and use the same shared storage.
- If the shared storage is under MPIO control, set the reserve_policy attribute of the disk to no_reserve.
- If the shared storage is not under MPIO control, look up the array documentation to locate a similar attribute to set on the disk.

Internal testing on EMC disks shows that this field maps as the reserve_lock attribute for EMC disks. In this case, setting it to NO achieves the same result.

For EMC Luns this attribute reserve_policy maps as reserve_lock, and should be set to NO.

IBM's website also states the SSA (Shark) disks also use the reserve_lock attribute.

https://inetsd01.boulder.ibm.com/doc_link/en_US/a_doc_lib/libs/ktechrf2/ssadisk.htm
reserve_lock :- Specifies whether the SSA disk device driver locks the device with a reservation when it is opened for an hdisk.

https://www-01.ibm.com/support/docview.wss?uid=ssg1S1002541
The Persistent Reserve design establishes a method and interface through a reserve_policy attribute for SCSI disks which specifies the type of reservation (if any) that the device driver will establish before accessing data on the disk. For devices that do not support the reserve_policy attribute, the drivers will use the value of the reserve_lock attribute to determine the type of reservation to use for the disk.

Veritas VxVM doesn't use storage in the same way that IBM/HACMP does. We use other mechanisms to protect against data corruption like I/O Fencing and the fencing driver. So any locks that aren't being set by vxfen, should be disabled for the storage managed by InfoScale.

Because, In IBM AIX, the reserve_lock attribute is set "yes" by default. We can change the attribute as the follow.

#chdev -l hdisk1024 -a reserve_lock=no

After we set the attribute on shared disks with HA solution, we no longer receive the error messages.

How to set reserve_lock attribute of IBM hdisk with shared array environment.