IBM Support

Possible undetected data corruption may be experienced due to a controller hardware problem

Troubleshooting


Problem

A hardware error related to the password reset signal logic on the storage controller causes platform dependent responses, including possible undetected data corruption.

Symptom

The symptoms related to this password reset related hardware error are as follows:

DS3500 and DCS3700 without Performance Module:
- The storage system may exhibit controller resets, frequent LUN failovers, LUNs not on preferred path, and/or decreased performance.
- The Major Event Log (MEL) will contain "0x1400 - Password Reset to Default" entries occurring every two minutes.
- Systems running 07.84.xx.xx controller firmware using RAID 6 with write cache enabled may experience Data Parity Mismatches (DPMs) and/or undetected data corruption following the controller resets. The DPMs are reported in the Major Event Log on the storage subsystem as 0x200A (DPM event), 0x2046 (unsuccessful isolation of DPM), and 0x2042 (Successful isolation of DPM) MEL events following an excessive amount of 0x1400 (password reset) MEL events and controller reset sequences..

DCS3700 with Performance Module and DCS3860:


- Controller may halt due to a startup error with “SE 88” showing on the 7-segment display following a controller reboot or power on.

Cause

The root cause of the issue is a low resistance short (typically 5-15 ohms) between the password reset switch and ground. The low resistance short causes the password reset signal to be observed as active by the interrupt logic on the controller.

This problem has a low probability of being encountered.

Resolving The Problem

A hardware change to the controller has been made to resolve this problem in manufacturing.

The following actions should be taken by current users:

DS3500 and DCS3700 without Performance Module:
- A firmware fix has been made to the controller firmware to mask the password reset interrupt which prevents this problem from occurring. This fix is available in controller firmware 7.86.51.00, and later releases, and is available on the download page for the referenced products. Customers should update their controller firmware to a level which contains this fix as soon as possible.
Note: With 07.86.51.00 installed on the referenced controllers, the password reset button interrupt is ignored. Button depressions or malfunctions will have no impact on controller operations. Please contact IBM Support for assistance with resetting the password when controller firmware 7.86.51.00, and later, is installed.
- If the firmware fix has not been installed, then any controller that encounters this problem should be replaced.

DCS3700 with Performance Module and DCS3860:
- No controller firmware fix was needed for these products.
- Controllers that encounter this problem should be replaced.

[{"Product":{"code":"HW28S","label":"Disk systems->DS3500 (DS3512, DS3524)"},"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Component":"--","Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"Version Independent","Edition":"","Line of Business":{"code":"","label":""}},{"Product":{"code":"HW28U","label":"Disk systems->DCS3700"},"Business Unit":{"code":"BU054","label":"Systems w\/TPS"},"Component":" ","Platform":[{"code":"","label":""}],"Version":"","Edition":"","Line of Business":{"code":"","label":""}},{"Product":{"code":"SSUUKF","label":"IBM System Storage DCS3860"},"Business Unit":{"code":"BU058","label":"IBM Infrastructure w\/TPS"},"Component":" ","Platform":[{"code":"","label":""}],"Version":"","Edition":"","Line of Business":{"code":"LOB26","label":"Storage"}}]

Document Information

Modified date:
16 September 2022

UID

ssg1S1005142