Asset ID: |
1-75-1147963.1 |
Update Date: | 2012-05-21 |
Keywords: | |
Solution Type
Troubleshooting Sure
Solution
1147963.1
:
Sun StorEdge T3, T3+, and 6120 Arrays: Troubleshooting Battery Failures
Related Items |
- Sun Storage 6120 Array
- Sun Storage T3 Array
- Sun Storage T3+ Array
|
Related Categories |
- PLA-Support>Sun Systems>DISK>Arrays>SN-DK: T3_6020_6120
- .Old GCS Categories>Sun Microsystems>Storage - Disk>Modular Disk - Other
|
In this Document
Applies to:
Sun Storage T3 Array - Version Not Applicable and later
Sun Storage 6120 Array - Version Not Applicable and later
Sun Storage T3+ Array - Version Not Applicable and later
Information in this document applies to any platform.
Purpose
This document addresses the identification of failed or failing batteries in the array, via the symptoms provided.
Symptoms:
- Battery not refreshing after replacement
- StorADE reports an alarm in Browser or email
- Messages in logs for bat#[1]: E: Battery u#b# needs to be replaced.
- Messages in logs for BATW[1]: E: u#b# idle life exceeded
- Performance degraded,
- PCU fault LED lit
- Global fault LED lit.
Troubleshooting Steps
Note : Please validate that each troubleshooting step below is true for your environment. The steps will provide instructions or a link to a document, for validating the step and taking corrective action as necessary. The steps are ordered in the most appropriate sequence to isolate the issue and identify the proper resolution. Please do not skip a step, unless otherwise directed to do so.
1)Verify that you can log into your array using "telnet <array IP address or host name>".
- If you cannot log into your array via telnet, Refer <Document: 1131324.1> Troubleshooting Sun StorEdge[TM] T3, T3+ and 6120 Access Problems
- Otherwise continue to Step 2.
2) Verify battery status using "refresh -s"
Example:
t3:/:<6>refresh -s
Current Date and Time: Wed Mar 17 15:35:11 GMT 2010
Next Scheduled Refresh: Mon Apr 19 02:00:16 GMT 2010
Battery State Last Health Check Warranty Expiration
------- ------- ---------------------------- ----------------------------
u1b1 normal Tue Mar 16 15:45:16 GMT 2010 Tue Nov 15 14:49:08 GMT 2011
u1b2 normal Tue Mar 16 19:45:31 GMT 2010 Tue Nov 15 14:49:08 GMT 2011
u2b1 normal Mon Jan 18 05:54:03 GMT 2010 Tue Nov 15 14:49:23 GMT 2011
u2b2 failed Mon Jan 18 05:57:19 GMT 2010 Tue Nov 15 14:49:08 GMT 2011
Following Table Explains the battery status and their definition.
STATE
|
Definition of State
|
off
|
The PCU is off. The PCU was either turned off by the switch or disabled by the firmware.
|
missing
|
The PCU and/or battery is missing.
|
normal
|
The battery is healthy and operating normally. The PCU is running from AC input.
|
active
|
The battery is supplying power to the PCU indicating an AC input failure.
|
charge
|
The battery is charging and not being tested (health check).
|
in test
|
The battery is being tested (health check) and is either discharging or charging.
|
failed
|
The battery failed testing.
|
fault
|
A fault in the PCU and/or battery was detected.
|
disabled
|
The PCU has been disabled and is not supplying power.
|
margin
|
None of the above states apply and the PCU has been margined. This should only be seen in manufacturing.
|
unknown
|
Internal software error.
|
Shelf Life Expired |
Battery Warranty has expired |
- If the battery status is normal, you have verified that the battery is healthy, and does not require any further action.
- If the battery status is normal, and your are receiving battery messages for "idle life exceeded" or "needs to be replaced", go to Step 6.
- If the battery status is Shelf Life Expired, the battery has passed its warranty date, go to Step 6.
- If any number of battery states are listed as failed, go to Step 3.
- If the battery state is off, missing, active, or disabled, verify whether there is a Power Cooling Unit problem. Reference <Document: 1123923.1> Troubleshooting Sun StorEdge[TM] T3, T3+, and 6120 Power Cooling Unit Failures.
- If the battery state is in test, or charge, check the battery state again in 10 hours. The recharge time for a battery is typically from 3 to 10 hours.
- If the battery state is in test or charge and it has been longer than 10 hours in that state:
Collect the array data from a Solaris host by running: /opt/SUNWexplo/explorer -w !default,t3extended
Open a Service Request with Oracle and supply the extractor.
- If the battery status is listed as fault, margin, or unknown:
Collect the array data from a Solaris host by running: /opt/SUNWexplo/explorer -w !default,t3extended
Open a Service Request with Oracle and supply the extractor.
3. Verify array firmware revision and array type using the "ver" command.
Example:
6120:/:<1>ver
6120 Release 3.1.6 Thu Feb 3 16:48:03 PST 2005 (10.16.10.131)
Copyright (C) 1997-2003 Sun Microsystems, Inc., All Rights Reserved
The firmware is 3.1.6 in this output. And the array type is 6120.
The array type can be:
Value from ver output |
Common Array Names for this model |
6120 |
6120, 6020 |
T3B |
T3+, T3B |
T3 |
T3, T3A |
Your firmware revision and array type will determine how to handle the battery failure
- If your battery is failed AND your array model is a T3, or a T3+ and the firmware is below 3.01.02, go to Step 5.
- If the battery is failed AND your array model is a 6120 below 3.01.01, go to Step 4.
- If the battery is failed AND your array model is a T3+ above 3.01.02, or a 6120 with version 3.01.01, go to Step 5.
- If the battery is normal AND your array model is a T3+ above 3.01.02, or a 6120 with version 3.01.01, go to Step 6.
4. Validate that this is the only battery failure.
Use the same refresh -s output from Step 2. Due to CR 4814235 on 6120 arrays running firmware lower than 3.01.01, the batteries may fail prematurely after replacement.
- If there is only one battery failed, continue to Step 5
- If you have more than one battery failure and the batteries had been previously replaced in the unit, you will need to update the array firmware to prevent this from happening again. Correcting the issue can be done by either
- Replacing the failed batteries. Sun's policy is to only replace the array batteries at the customer's expense.
- Allowing remote access to the array to clear the failed state.
In either case upgrade to f/w 3.2.7 using Patch "116931-23" is recommended.
5. Contact Oracle to have the battery replaced.
Please supply:
- refresh -s output
- id read uNbN (the location of your battery from refresh -s)
- ver output
- fru stat
- fru list
Or Collect the array data from a Solaris host by running: /opt/SUNWexplo/explorer -w !default,t3extended
6. Decide whether you want the battery age reset, or the battery replaced.
By reaching this step, you have validated that the battery is operating normally or failed because its shelf life expired. Shelf life expiration is due to a battery that is beyond the warranty date on the system. Battery replacements due to expiration are replaced at customer expense, and are not covered under the array service contract or warranty.
In the case of any system running a 3.x version, the battery refresh cycle will judge whether the battery is failed, and at that time, should be replaced. In these cases, the message from the array for
Messages in logs for bat#[1]: E: Battery u#b# needs to be replaced.
OR
Messages in logs for BATW[1]: E: u#b# idle life exceeded
should be ignored by the administrator.
In the case of a T3 or T3+ array with firmware less than 03.01.02 firmware, the battery management software on the system does have expiration code built in. The same applies in that the battery will only be replaced when it fails to hold a charge on during a refresh cycle. Again, battery replacements are not covered under warranty or service contract, and are replaced at the customer's expense.
You can update the battery life in these cases by running through the following sequence, on the array(This does not work on any array running 3.x firmware!!):
- Check the date on the system and set it properly
- .bat -n u[1-2]pcu[1-2]
- .bat -n u[1-2]pcu[1-2]
- check refresh -s
- If the batteries are not in a discharging state, issue a refresh -c. The batteries must go through a full refresh cycle to initialize the install date and remove the messages and status on the array.
If you would like your batteries replaced, please contact Oracle with:
- refresh -s output
- id read uNbN (the location of your battery from refresh -s)
- ver output
- fru stat
- fru list
Or Collect the array data from a Solaris host by running: /opt/SUNWexplo/explorer -w !default,t3extended
Attachments
This solution has no attachment