Hot-swap may cause data loss - LSI Logic MegaRAID 8480 SAS Controller

Resolving The Problem

Source

RETAIN tip: H19286

Symptom

In rare cases, after removing a hard disk drive under heavy I/O without first marking the drive dead, data may become corrupted with no error events logged.

Affected configurations

This tip is not hardware specific.

The system is configured with one or more of the following IBM Options:

The system is configured with at least one of the following:

Note: This does not imply that the network operating system will work under all combinations of hardware and software.

Please see the compatibility page for more information:

• http://www.ibm.com/servers/eserver/serverproven/compat/us/

Solution

It is strongly recommended that users upgrade to the new MegaRAID 8480 firmware image version 1.03.20-0400.

The file is available at the following URL to resolve this issue.

The Storage Support Matrix can be reached at the following URL:

• http://www.IBM.com/support/docview.wss?uid=psg1MIGR-65723

Workaround

Prior to removing a hard disk drive from an operational state, use the Storage Manager software to mark the drive dead and then physically remove the disk.

Additional information

Users are encouraged to employ frequent and secure means of backing up their data.

Removing a hard disk drive, during operation, causes the MegaRAID controller to begin error recovery routines that test the presence of other known devices on the channel.

Under heavy stress, a successful presence check for another device (such as the hard drive backplane), immediately after a hard disk removal, could result in an incorrectly acknowledged "successful I/O operation" for the hard disk device that became missing.

To avoid this situation, devices should never be removed from an active channel during operation.