Tape - Troubleshooting Read/Write Errors

Asset ID:	1-75-1019811.1
Update Date:	2017-06-22
Keywords:

Solution Type Troubleshooting Sure

Solution 1019811.1 : Tape - Troubleshooting Read/Write Errors

Applies to:

Sun StorageTek SDLT Tape Media - Version Not Applicable and later
Sun StorageTek 9840B Tape Drive - Version Not Applicable and later
Sun StorageTek 9840 Data Cartridge - Version Not Applicable and later
Sun StorageTek 9940 Data Cartridge - Version Not Applicable and later
Sun StorageTek 9840D Tape Drive - Version Not Applicable and later
All Platforms

Purpose

How to Troubleshoot Read Write issues on Tape Drives.

If after going through this document, it's determined to be a tape drive hardware issue, please open a MOS Service Request (SR) against the drive tray serial number.

Troubleshooting Steps

Use the following steps to diagnose read/write issues:

Read Errors:

Drive read errors will usually fall into one of the following categories:

- The tape was created with a permanent write error and an attempt is being made to read into the area of tape that had the permanent error.
    Look for FSC37A5 for 9x40 devices.
    There is no FSC for this in T10000 devices as the defects are handled differently.

- Physical damage occurred on the tape after it was written.
    Tape received rough handling causing edge damage.
    Internal problem in the cartridge. (ie: Someone spilled a beverage that got into the cartridge)

- The drive that created the tape may have a problem that was not detected during creation.
    Possibly it was writing data off-track/skewed.
       Writing off-track can damage data on an adjacent track.
    Any other timing or alignment issue that would cause interchange-ability problems between drives.
    The creating drive may be the only drive that can read the tape back.
       Keep the creating drive until all tapes have been read back.

Read Errors - Dump Analysis

-Oracle Technical Support will need a drive dump in order to assist when an unreadable tape is encountered.
    The customer/operator needs to read down the tape until the failure occurs.
    Vary offline/down the drive after the failure occurs so that no other tape gets mounted.
    After the tape is unloaded, trigger a manual dump on the drive.
       Refer to the following solutions for pulling dumps on different drives:
          Document 1014247.1 for T10000 and 9x40 drives
          Document 1013260.1 for LTO drives
          Document 1013576.1 for SDLT drives

When the dump completes, retrieve the dump file and open a case with Oracle Technical Support for analysis.

Duplexing and Data Recovery

- If you are not able to read the tape and no other copy of the data exists, then this tape is a candidate for data recovery.

    The data recovery lab will not be able to recover over-written data.
    They can generally recover data past the over-write.
    The customer should copy off any data they can get before the tape is sent in for recovery.

Refer to Document: 1015907.1 for 9x40/T10000 data recovery process.
Refer to Document: 1014880.1 for LTO/SDLT/DLT data recovery process.

- All customers should consider making 2 copies (duplexing) of any tapes where there is no other copy of the data.

Write Errors:

- Permanent Write Errors can be caused by:
    Defective media. (edge damage, hard banding, etc.)
    Defective drive hardware.
    Combination of marginal media in a marginal drive.
    Servo going off-track during a write operation.
    Firmware issue.

Note: The customer should never try to reuse the tape for writing (append to the tape) while there is live data on the tape.

Troubleshooting a Permanent Write Error

    Swap the tape to another drive if swap is available.
    Copy the data off of the tape.
    Have the tape certified. This can be done bu using utilities that are designed for tape certification
    such as FATS/FATAR.
    If a utility is not available, simply fill the tape with dummy data to ensure no error.
    If the tape certification is successful, suspect a marginal drive.

    Consult the PMA Viewer report to see if this tape has failed on multiple drives.
    Check to see if the drive is taking write errors on multiple tapes.
    Open a case in MOS if you are unsure if the drive or tape is at fault.

Note: The following may require on-site support. (These Tools only available to CSE)
    Run the health check tool against the drive using VOP or SCSI/FC Toolbox.

The SCSI/FC Toolbox program can be downloaded from:
   SCSI/FC ToolBox - Obtaining SCSI/Fibre ToolBox or Magma associated Drivers (Doc ID 1016313.1)

VOP program can be downloaded from:
9840/T10000 - Where to Acquire Virtual Operation Panel (VOP) Application (Doc ID 1014070.1)

Run internal drive Read/Write diagnostics to the drive.

Attachments

This solution has no attachment