Sun Storage 3310 Array
 
Sun Storage 3320 SCSI Array
 

GCS>Sun Microsystems>Sun Alert>Criteria Category>Data Loss
 
GCS>Sun Microsystems>Sun Alert>Release Phase>Resolved
 

When the connection between the SE33x0 array and the host has degraded to the point that WRITE requests cannot be completed due to connectivity issues, persistent SCSI parity errors may be generated between the host and the SE33X0 array and data inconsistencies may occur.

This issue can occur on the following platforms:

• Sun StorEdge 3310 SCSI array without firmware 4.15F (as delivered in patch 113722-15)
• Sun StorEdge 3320 SCSI array without firmware 4.15G (as delivered in patch 113730-01)

SCSI parity errors can cause invalid data to get written into the array's cache. Prior to firmware version 4.15, this data eventually gets flushed to the disk media, permanently storing this invalid data on the volume. Firmware version 4.15 was modified to discard this corrupted data rather than write it to disk media. This reduces the probability of corrupting the volume. However, in the rare case where the write command overlapped a prior write command's data that still resided in cache, that data will also be discarded.

Single Path Configurations

Configurations in which a host has only one path to one or more logical units on the array are exposed to this problem. This is because there is no redundant path between the host and the SE33x0 array. This lack of redundancy does not allow for a retry using a second path to the SE33x0 array.

When using firmware version 4.15 in this configuration, if any write commands failed due to parity errors, there is a possibility of lost write data in cache if the application or file system issued writes to overlapping LBAs.

When using older firmware in this configuration, the data for LBAs of any WRITE request that cannot be completed as a result of a PARITY ERROR returned by the SE33x0 should be considered to have invalid data.

Multi Path/High Availability Configurations

The exposure for a properly configured High Availability configuration using a host multi-pathing driver and and multiple separate connections between the host(s) and the SE33x0 array is very small. In this configuration, the multi-pathing driver in the host will utilize the second, non-compromised path to the array controller to retry the WRITE request. A successful retry will successfully write the intended data to the correct LBAs with the following exceptions:

1. If the SE33x0 array or the host experiences a power failure between the failed WRITE request and the successful completion of the retry down the second path, the data for the failed WRITE request should be considered invalid.

2. If the Host OS experiences a crash or a multi-path driver error between the failed WRITE request and the successful completion of the retry down the second path, the data for the failed WRITE request should be considered invalid.

Should the described issue occur, persistent SCSI parity errors between the host and the SE33x0 array will be generated. The SE33x0 array will return a SCSI status of "Parity Error" to the host SCSI Host Bus Adapter (HBA). Typically, the host SCSI HBA will retry the WRITE request some number of times (most drivers attempt between 2 to 6 retries) before returning the WRITE request to the application with a FAILURE status.

There is no workaround for this issue. Please see the resolution section below.

• Sun StorEdge 3310 SCSI array with firmware revision 4.15F (as delivered in patch 113722-15 or later)
• Sun StorEdge 3320 SCSI array with firmware revision 4.15G (as delivered in patch 113730-01 or later)

Note: Insure that SCSI connections are reliable and properly configured to minimize the probability of parity errors and use multiple SCSI connections with failover drivers.

Because the nature of the changes would require a major redesign, the issue described in BugID 6378796 was closed as "will not fix."

This Sun Alert notification is being provided to you on an "AS IS" basis. This Sun Alert notification may contain information provided by third parties. The issues described in this Sun Alert notification may or may not impact your system(s). Sun makes no representations, warranties, or guarantees as to the information contained herein. ANY AND ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, OR NON-INFRINGEMENT, ARE HEREBY DISCLAIMED. BY ACCESSING THIS DOCUMENT YOU ACKNOWLEDGE THAT SUN SHALL IN NO EVENT BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, PUNITIVE, OR CONSEQUENTIAL DAMAGES THAT ARISE OUT OF YOUR USE OR FAILURE TO USE THE INFORMATION CONTAINED HEREIN. This Sun Alert notification contains Sun proprietary and confidential information. It is being provided to you pursuant to the provisions of your agreement to purchase services from Sun, or, if you do not have such an agreement, the Sun.com Terms of Use. This Sun Alert notification may only be used for the purposes contemplated by these agreements.

Copyright 2000-2008 Sun Microsystems, Inc., 4150 Network Circle, Santa Clara, CA 95054 U.S.A. All rights reserved.

References

102011 - Sun StorEdge 33x0/3510 Arrays May Report a Higher Incidence of Drive Failures With Firmware 4.1x SMART Feature Enabled

102067 - Sun Cluster 3.x Nodes May Panic Upon Controller Failure/Replacement Within Sun StorEdge 3510/3511 Arrays

102086 - Failed Controller Condition May Cause Data Integrity Issues

102098 - Insufficient Information for Recovery From Double Drive Failure for Sun StorEdge 33x0/35xx Arrays

102126 - Recovery Behavior From Fatal Drive Failure May Lead to Data Integrity Issues

102127 - Performance Degradation Reported in Controller Firmware Releases 4.1x on Sun StorEdge 3310/351x Arrays for All RAID Types and Certain Patterns of I/O

102128 - Data Inconsistencies May Occur When Persistent SCSI Parity Errors are Generated Between the Host and the SE33x0 Array

102129 - Disks May be Marked as Bad Without Explanation After "Drive Failure," "Media Scan Failed" or "Clone Failed" Events

Note: One or more of the above Sun Alerts may require a Sun Spectrum Support Contract to login to a SunSolve Online account.

As referenced in bug 6363490, this issue may occur with a faulty cable where possibly a toggled bit on the upper 8-bit of the cable has occurred.

A firmware release is scheduled that will ensure that data that has been compromised by the host SCSI HBA to the SE33X0 array controller will not be flushed to media. When this fix is delivered, there are cases in which stale data may remain in the SE33X0 array for LBAs that correspond to the failed WRITE Request.

Subsequent firmware release is scheduled that will ensure that the data that has been compromised by the host SCSI HBA to the SE33X0 array will not be flushed to media AND EITHER:

A. The most current version of data successfully written to the SE33X0 array by the host may be read from the SE33X0 array

OR:

B. A host READ request for data that could not be recovered will be returned with an error indicating MEDIA ERROR for the LBAs that can not be recovered.

A successful WRITE request to the LBAs corresponding to the MEDIA ERROR will allow these LBAs to be read correctly again.

References

This solution has no attachment