Hitachi TAGMASTORE MK-90RD011-25 - Manual (Page 356)

338 Chapter 5 Troubleshooting
5.1 General Troubleshooting
If you have a problem with the CCI software, first make sure that the problem is not being
caused by the UNIX/PC server hardware or software, and try restarting the server. Table 5.1
provides operational notes and restrictions for CCI operations.
For maintenance of Hitachi TrueCopy and ShadowImage volumes, if a failure occurs, it is
important to find the failure in the paired volumes, recover the volumes, and continue
operation in the original system. When a CCI (HORCM), Hitachi TrueCopy, or ShadowImage
failure is detected, please collect the data in the error log file and trace data (all files in
HORCM_LOG directory), and report the failure to your Hitachi Data Systems representative.
If you need to call the Hitachi Data Systems Support Center, please refer to section 5.4 for
instructions. Note: Use the FD Dump Tool or FDCOPY function (refer to the Storage
Navigator User’s Guide for the storage system) to copy the Storage Navigator configuration
information onto diskette, and give the diskette(s) to the Hitachi Data Systems service
personnel. The Storage Navigator Error Codes document for the storage system provides a
list of the error codes displayed by Storage Navigator.
Table 5.1 Operational Notes for CCI Operations
Condition Recommended Action
Startup/shutdown restrictions When the server starts up, the secondary volume may be updated by the primary volume’s
server. The secondary volume must not be mounted automatically in the startup sequence. If
the secondary volume is used by the LVM, the volume group of the LVM must be
deactivated. The secondary volume must be mounted in the split state or in the simplex
mode.
When the server starts up, the secondary volume can be activated without confirming, when
can be guaranteed that the secondary volume has been PSUS (R/W enable) or in the SMPL
state by server shutdown sequence.
Hot standby operations Hitachi TrueCopy commands cannot execute hot standby operations between the primary
and secondary volumes. Use the takeover command intended for the HA configuration to
execute the hot standby operation. In hot standby operation, two servers are used, and the
active (primary) and standby (secondary) server programs are run alternately in each server
in case of failure in one server. Follow these precautions:
ƒ Operation across volumes. Since each Hitachi TrueCopy command causes the server
software to handle the volume by volume, a single volume should not be partitioned to
prevent it from being used by some servers.
ƒ Using LVM and Hitachi TrueCopy together. When constructing the LVM on the paired
volume in the mutual hot standby configuration, the LVM logical volumes must be
constructed in units of volume to prevent the volumes from being mounted by the LVM.
Coexistence of LVM mirror and
Hitachi TrueCopy
When the LVM mirror and Hitachi TrueCopy volumes are used together, the LVM mirror
handles write errors and changes the volumes. Thus, the fence level of the volumes used by
the LVM mirror must be set to data.
Using paired volume in a single
host
When constructing paired volume in a single host, it is necessary to activate two or more
CCI instances. To activate two or more CCI instances, instance numbers must be assigned
using the environmental variable HORCMINST. The HORCM and Hitachi
TrueCopy/ShadowImage commands must possess this environmental variable. A
configuration definition file and a log directory must be set for each instance. The command
device described in the configuration definition file must be established in a way to be
following either every instance. If using a command device between different instances on
the same SCSI port, the maximum number of instances per command device is 16. If this
number is exceeded, the use a different SCSI path for each instance.