Sun Storage 2510 Array
 
Sun Storage 2540 Array
 
Sun Storage 2540-M2 Array
 
Sun Storage 6780 Array
 
Sun Storage 6130 Array
 
Sun Storage 6180 Array
 
Sun Storage Flexline 380 Array
 
Sun Storage 2530-M2 Array
 
Sun Storage 2530 Array
 
Sun Storage 6580 Array
 
Sun Storage 6140 Array
 
Sun Storage 6540 Array
 

PLA-Support>Sun Systems>DISK>Arrays>SN-DK: 6140_6180
 

In this Document

Symptoms

Changes

Cause

Solution

References

Created from <SR 3-8443779981>

Applies to:

Symptoms

A performance problem has been isolated to a Sun Storage 2500, 2500-M2, or 6000 array.  Anything which accesses data on the array experiences an obvious latency, far exceeding normal operation. Monitoring utilities (such as iostat) will report the disks up to 100% busy with high service times. Typical jobs are taking up to 3 or 4 times longer to complete. All evidence clearly points to a problem on the array. However, at first glance, Common Array Manager (CAM) reports no problems related to performance.

• There are no failed, expired or charging batteries.
• There are no volumes not on preferred path.
• There are no degraded or reconstructing volumes.
• Common Array Manager (CAM) reports read and write cache active for all volumes.
• If there are any alarms, they are unrelated to performance (Such as firmware not at baseline)

Changes

There have been no obvious changes to the array. There may have been some scheduled maintenance, repairs, or resets, but nothing which would identify a true cause and effect.

Cause

The problem is with read and write cache on the array. Even though CAM is reporting read and write cache as active, it is actually inactive. The only way to validate the current cache state in this circumstance is to evaluate the stateCaptureData.dmp file from the supportData. See <Document 1002514.1> Collecting Sun Storage Common Array Manager Support Data for Arrays.

To correctly interpret the file, several acronyms need to be expanded and some terms explained.

• RCE - Read Cache Enabled
• RCA - Read Cache Active
• RCI - Read Cache Inactive
• WCE - Write Cache Enabled
• WCA - Write Cache Active
• WCI - Write Cache Inactive
• CME - Cache Mirroring Enabled
• CMA - Cache Mirroring Active
• CMI - Cache Mirroring Inactive
• CWOB - Cache Without Batteries
• Dev x - These are the volumes on the array
• (Un)owned - The controller which owns (or does not own) the volume.

Once you unzip the supportdata.zip bundle, view the stateCaptureData.dmp file and look for the command ccmShowState. The command should show up twice, as it is run from each controller. In the example output, we can see the actual cache state of each volume. Notice how the A controller reports all cache states active and enabled. The problem is with the B controller. The cache states on B have been internally disabled, thus turning off all write cache for all volumes.

Solution

The solution to the problem is to reset the controller whose cache states have been internally disabled. This may be done from the CAM GUI or with the service command.

From the example above, we would tailor the service command to reset the B controller.

From the CAM GUI, you would

This solution has no attachment