HP 3PAR OpenVMS Implementation Guide (763874-002)

9 Troubleshooting HP 3PAR 3.1.2 MU2/MU3 and HP 3PAR

3.1.3

1. Disk device medium offline issue

This is related to HP 3PAR OS 3.1.2 MU2/MU3 and 3.1.3.

The first initialization of HP 3PAR LUN after it is presented to OpenVMS host might result in

the following medium offline error message:

%INIT-F-MEDOFL, medium is offline

Solution

This issue is resolved in the VMS84I_FIBRE_SCSI-V0600 patch release.

2. When OpenVMS Volume Shadowing is used between HP 3PAR StoreServ and other HP arrays

(for example, HP P6000 or EVA arrays that support SCSI READL/WRITEL), the shadowing

driver might remove the HP 3PAR LUN from the shadow set when a bad block is reported

and the Operator log displays an error message similar to:

"%SHADOW_SERVER-E-SSRVTERMCPY, terminating copy operation on device

_DSA191: at LBN: 34404414, ID number: 1D1A3F96" and

"%SHADOW_SERVER-E-SSRVTRMSTS, reason for termination of operation

on device _DSA191: ABORT, abort.

Details

OpenVMS uses SCSI READ/WRITEL to signal a forced error on a particular disk block or

LUN. During the OpenVMS volume shadow copy operation, if the source disk supports SCSI

READL/WRITEL (for example, HP P6000 or EVA arrays), and if Forced Error is detected in

the source disk, then the HP 3PAR StoreServ disk array LUN is removed from the shadow set.

Solution

For the shadow copy to proceed towards completion, perform the following procedure:

a. Identify the file corresponding to the blocks with Forced Error enabled. The block numbers

are displayed in the error message by volume shadowing copy operation. Run the

following OpenVMS command to identify all the blocks with Forced Error bit enabled:

$ANALYZE /DISK_STRUCTURE /READ_CHECK

b. Recreate the file to correct it.

c. If you are unable to correct the file, then restore it from a valid backup.

d. If a valid backup is not available, then do one of the following for the shadow copy

operation to proceed:

• Run the $ DELETE/ERASE command to delete the original file.

• Run the BACKUP utility so that the Forced Error flags are not carried forward.

CAUTION: This data must be considered as degraded data as it has error block.

To view more information on Bad Block Replacement and Forced Errors for DSA Disks, click

http://h71000.www7.hp.com/doc/73final/6136/6136pro_006.html.

To view more information on detecting if the LUNs support Forced Error bit, click http://

h71000.www7.hp.com/doc/84final/ba554_90020/index.html.