Sun Storage 33x0/351x Arrays: How to resolve "DRAM parity error detected"

Asset ID:	1-71-1010810.1
Update Date:	2018-05-17
Keywords:

Solution Type Technical Instruction Sure

Solution 1010810.1 : Sun Storage 33x0/351x Arrays: How to resolve "DRAM parity error detected"

Applies to:

Sun Storage 3511 SATA Array - Version Not Applicable to Not Applicable [Release N/A]
Sun Storage 3320 SCSI Array - Version Not Applicable to Not Applicable [Release N/A]
Sun Storage 3310 Array - Version Not Applicable to Not Applicable [Release N/A]
Sun Storage 3510 FC Array - Version Not Applicable to Not Applicable [Release N/A]
All Platforms

Goal

Memory errors on Sun Storage 3000 Arrays with two hardware RAID controllers are reported in the event log, and possibly elsewhere, without an indication of which controller logged the error. This document pertains to the following products:

Sun Storage 3310 Array
Sun Storage 3320 Array
Sun Storage 3510 Array
Sun Storage 3511 Array

You can view DRAM parity error messages in the event log using the sccli show events command.

[0104] #4287: StorEdge Array SN#xxxxxxx Controller ALERT: DRAM parity error detected

This document explains how to use the RAID firmware menu to determine whether the primary or secondary controller is experiencing the error.

Solution

An occasional DRAM parity error does not require controller replacement. If you consistently see multiple errors logged daily, replace the controller.

If you use the sccli show events command, there is no indication of which controller is generating the error. To identify the correct controller, use a serial or telnet connection to view the RAID firmware menu. To read about how to connect to the array by telnet, refer to the section titled "Setting up Out-of-Band Management Over Ethernet" in Chapter 4 of the Installation, Operation and Service Manual for your array. You can find this manual by going to the Oracle Disk Storage Systems Documentation page on the Oracle Technology Network.

From the main menu, choose "View and Edit Event Logs" and you will see an event similar to the following.

[0104] Controller SDRAM ECC Single-bit Error Detected
---Wed Feb 12 14:47:46 2003------------------------------------<S>---

The "S" in the second line indicates that this error was generated by the secondary controller. If there is a "P" instead, this means the primary controller reported the error.

Care must be taken to identify the correct controller. The event is visible in the RAID firmware menu and shows whether the primary or secondary experienced the error -- at the time of the error. However, it is possible that some later event may have reversed the roles of the two controllers. Manually failing a primary controller or power cycling the array are examples of actions that may result in switching the assignment of primary and secondary controller roles.

Viewing Array Events

The RAID firmware menu and sccli output from the show events command are the two most common methods to view DRAM parity error messages, but they may also appear in messages from Diagnostic Reporter or in /var/adm/messages.

From Diagnostic Reporter:

************************************************************
hostname=x.x.x.x dummyhost
timestamp=02/12/2003 14:48:17
device=HBA 3[Ch0Id2] SUN StorEdge 3510
priority=Critical
error_code=010b1d0d
message=Controller Event, SDRAM Error. Likely controller error. If error
persists, replace defective controller. (Wed Feb 12 14:47:46 2003)
************************************************************

From /var/adm/messages:

Feb 12 14:48:40 dummyhost SUNWscsdMonitor[374]: [ID 677437 daemon.error]
[SUNWscsd 0x10B1D0D: Critical] Controller Event, SDRAM Error.
Likely controller error. If error persists, replace defective controller.
(Wed Feb 12 14:47:44 2003)

Attachments

This solution has no attachment